Advanced Polymer-Based Drug Delivery Systems for Chemotherapy

Home , Carmofur

UCL SCHOOL OF PHARMACY BRUNSWICK SQUARE

Advanced polymer-based drug delivery systems for chemotherapy

BRENDA SANCHEZ VAZQUEZ

A thesis submitted for the degree of

DOCTOR OF PHILOSOPHY

Department of Pharmaceutics UCL School of Pharmacy University College London

September 2018

Declaration

“I, Brenda Sanchez Vazquez confirm that the work presented in this thesis is my own. Where information has been derived from other sources, I confirm that this has been indicated in the thesis.”

(Signature) (Date)

Acknowledgments

I would like to express my sincere gratitude to my supervisor Dr. Gareth Williams who gave me the amazing opportunity to be part of his group, who also gave me his continuous support and always believed in me, I could not have asked for a better supervisor for my PhD.

Lab 301 for being there when I needed someone and office 308 for all those unforgettable moments which I will always remember fondly. To Rita, Felipe, Alex C, Alex K, Maria, Haitham, Heyu and Sebs for all those great laughs and night outs.

I would also like to thank all the lab technicians who made me part of their team by providing me with the opportunity to become a Teaching Assistant, which taught me valuable lessons, as well as Carlos Silva for always having a smile for me.

To Hira Akhtar, who always offered me wise advice and showed me that I can always do better.

To Victor Diran for being my partner throughout this journey, who had always a smile for me, who always made me laugh and who always knew how to cheer me up when I was down.

And last but not least, special thanks to my parents for their unconditional love and support, without them I would not have been where I am today and even though we were kilometres apart, you were always in my heart. To my sister Sandra Aline, who I will always love unconditionally and who has travelled kilometres to be there for me whenever I needed her.

Abstract

The research in this thesis attempts to help reduce the side effects of current cancer treatments via targeted drug delivery. This was achieved by developing advanced polymer-based delivery systems for chemotherapeutic agents by using different techniques such as: electrohydrodynamic techniques, which include electrospinning and electrospraying, in addition to spray drying. The fabrication of fibres and particles was achieved using single fluid electospinning and electrospraying as well as to side-by-side structures and triaxial fibres.

Chapter 1 is a brief introduction of cancer and its current treatments where their benefits and limitations are critically discussed. Then, the importance and advantages of drug delivery systems in cancer therapeutics are explored. Finally, the fabrication of these delivery systems by electrospinning and spray drying is reviewed. After a detailed overview of the research field, the scientific gaps are identified and the overall aims of the experimental work on this thesis are outlined.

In Chapter 2, electrospinning and electrospraying techniques were used to generate drug-loaded Janus fibres and particles made of polyvinylpyrrolidone (PVP) by using a side-by-side spinneret. A photo-chemotherapeutic formulation was created by using rose bengal and carmofur, where the formulation created by electrospraying achieved a high selectivity towards lung cancer cells.

Chapter 3 describes the production of formulations in the form of fibres or particles composed of a polyvinylpyrrolidone (PVP) matrix and tristearin (GTS) by electrohydrodynamic techniques. These precursor formulations were used as templates to fabricate aqueous-dispersible, drug loaded SLNs. Indomethacin (BCS class 2), and 5FU a chemotherapeutic agent were used.

Chapter 4, builds on the work from Chapter 3, and details the scale-up production of the polymer-based templates to create SLNs, it also compares the electrohydrodynamic techniques with spray drying.

Chapter 5 focuses on using triaxial electospinning to create a template which upon hydration self-assembles in snowman Janus nanoparticles. These Janus nanoparticles create a photo chemotherapy formulation, which constitutes of a micelle containing hematoporphyrin and a liposome entrapping carmofur.

Table of contents

Acknowledgments ...... 3 Abstract ...... 4 List of Abbreviations………………………………………………………………………..11 List of Publications………………………………………………………………………….13

List of Figures ...... 14

List of Tables ...... 21 Impact statement ...... 22 1. Introduction ...... 23 1.1 Cancer ...... 23 1.1.1 Tumour physiology ...... 24 1.1.2 Cancer stages ...... 25 1.1.3 Cancer survival rates ...... 26 1.2 Current cancer treatments ...... 28 1.2.1 Chemotherapy ...... 28

a) 5-Fluorouracil ...... 31 b) Carmofur ...... 31 1.2.2 Photodynamic therapy ...... 32 1.2.3 Photo-chemotherapy ...... 36 1.2.4 Cancer treatment costs ...... 39 1.3 Challenges for an efficient treatment ...... 39 1.4 Drug delivery systems ...... 41 1.4.1 Modified release ...... 43 1.4.2 Targeted drug delivery ...... 43

1.4.3 Nanoscale vs macroscale DDS ...... 45 1.4.4 Polymer-based drug delivery systems ...... 46 1.4.5 Lipid based drug delivery systems ...... 47 1.5 Electrohydrodynamic techniques ...... 48 1.5.1 The electrospinning/electrospraying process ...... 48 1.5.2 Solution parameters ...... 50

1.5.3 Processing parameters ...... 51

1.5.4 Environmental parameters ...... 52 1.5.5 EHD techniques and drug delivery...... 53 1.6 Spray drying ...... 60 1.6.1 Spray drying in the pharmaceutical industry ...... 62 1.7 Thesis aims and outline ...... 64 1.8 References ...... 65 2. Janus structures generated by electrohydrodynamic atomization ...... 78 2.1 Introduction ...... 78

Janus structures ...... 78 Multi drug co-delivery...... 79 Targeted drug delivery ...... 80 Theranostics ...... 81 Preparation of Janus structures ...... 81 2.2 Aims and objectives ...... 83 2.3 Materials and Methods ...... 84 2.3.1 Materials ...... 84 2.3.2 Preparation of drug-loaded electrospun fibres and electrosprayed . particles…………………………… ……………………………………….84 2.3.3 Scanning electron microscopy ...... 85 2.3.4 Fluoresence microscopy ...... 85 2.3.5 Fourier-transform infrared spectroscopy (FTIR) ...... 85 2.3.6 Differential scanning calorimetry (DSC) ...... 85 2.3.7 Thermogravimetric analysis (TGA) ...... 86 2.3.8 X-ray diffraction ...... 86 2.3.9 Drug release studies ...... 86 2.3.10 Cell culture ...... 87

2.3.11 Statistical analysis ...... 89 2.4 Results and discussion ...... 89 2.4.1 Janus fibres ...... 89 2.4.2 Proof-of-concept ...... 89 2.5 Photo-chemotherapeutic fibres ...... 92 6

2.5.1 FTIR ...... 93

2.5.2 XRD ...... 95 2.5.3 DSC ...... 97 2.5.4 Drug release studies ...... 99 2.6 Janus particles ...... 100 2.6.1 PVP particles ...... 100 2.6.2 Solid state properties of Janus particles ...... 102 2.6.3 Drug release studies ...... 103 2.6.4 PCL particles ...... 104

2.6.5 In vitro cytotoxicity assays ...... 107 2.7 Conclusions ...... 110 2.8 References ...... 111 3. Electrohydrodynamic production of templates for the self-assembly of solid lipid nanoparticles ...... 115 3.1 Introduction ...... 115 3.2 Aims and objectives: ...... 119 3.3 Materials and methods ...... 119 3.3.1 Synthetic procedures ...... 119

3.3.2 Scanning electron microscopy ...... 121 3.3.3 Transmission electron microscopy ...... 121 3.3.4 Fourier-transform infrared spectroscopy (FTIR) ...... 121 3.3.5 Differential scanning calorimetry ...... 121 3.3.6 X-ray diffraction ...... 122 3.3.7 Dynamic light scattering (DLS) ...... 122 3.3.8 Entrapment efficiency and drug loading ...... 122 3.3.9 Drug release studies ...... 123 3.3.10 Cell culture ...... 123

3.3.11 Statistical analysis ...... 124 3.4 Results and discussion ...... 125 3.4.1 Optimisation ...... 125 3.4.1.1 Electrospinning optimisation ...... 125 3.4.1.2 Electrospraying optimisation ...... 126 7

3.4.2 SEM of optimised products ...... 129

3.4.3 Characterisation of PVP-GTS-5FU and PVP-GTS-IMC formulations ... 130 3.4.3.1 Fourier transform infrared spectroscopy (FTIR) ...... 130 3.4.3.2 DSC ...... 133 3.4.3.3 XRD ...... 134 3.4.4 Solid lipid nanoparticle characterisation ...... 136 3.4.4.1 TEM ...... 136 3.4.4.2 Entrapment efficiency ...... 138 3.4.4.3 In-vitro drug release ...... 139

3.4.4.4 Cell culture studies ...... 142 3.5 Conclusions ...... 144 3.6 References ...... 145 4. Scaling up production of EHD-generated sacrificial templates ...... 148 4.1 Introduction ...... 148 4.2 Aims and objectives ...... 152 4.3 Materials and methods ...... 153 4.3.1 Medium speed electrospinning ...... 153 4.3.2 High speed electrospinning (HSES) ...... 154

4.3.3 Spray drying ...... 155 4.3.4 Physicochemical characterisation ...... 156 4.3.5 Functional performance assays ...... 156 4.3.6 Permeation assays ...... 156 4.3.7 High performance liquid chromatography ...... 157 4.3.8 Statistical analysis ...... 157 4.4 Results and discussion ...... 158 4.4.1 Materials characterisation ...... 158 4.4.2 SEM ...... 158

4.4.2.1 Medium speed and corona electrospinning ...... 158 4.4.2.2 Medium speed electrospraying ...... 161 4.4.2.3 Spray drying ...... 161 4.4.3 FTIR ...... 162 4.4.4 XRD ...... 164 8

4.4.5 DSC ...... 164

4.5 Solid lipid nanoparticle characterisation ...... 165 4.5.1 TEM ...... 165 4.5.2 Entrapment efficiency and drug loading ...... 168 4.5.3 Drug release ...... 170 4.5.4 Permeation studies ...... 175 4.5.5 Cell culture studies ...... 177 4.5.6 Stability studies...... 179 4.6 Conclusions ...... 181

4.7 References ...... 182 5. Triaxial electrospinning for the creation of self-assembled nanostructures ...... 187 5.1 Introduction ...... 187 Liposomal drug delivery systems ...... 187 Micelles as drug delivery systems ...... 188 Formulation background ...... 189 Janus particles assembled from electrospun fibres ...... 190 Multiaxial electrospinning ...... 191

Triaxial electrospinning ...... 191 5.2 Aims and objectives ...... 192 5.3 Materials and methods ...... 193 5.3.1 Formation of polymer micelles ...... 193 5.3.2 Electrospinning solutions and parameters ...... 194 5.4 Results and discussion ...... 197 5.4.1 Micelles preparation ...... 197 5.4.1.1 TEM of self-assembled micelles ...... 197 5.4.2 Electrospinning ...... 198

5.4.3 Triaxial electrospinning ...... 201 5.4.3.1 Entrapment efficiency and hydrodynamic size ...... 205 5.4.3.2 FTIR ...... 207 5.4.3.3 XRD ...... 209 5.4.3.4 DSC ...... 211 9

5.4.3.5 In-vitro drug release ...... 213

5.4.3.6 Cell viability studies ...... 215 5.5 Conclusions ...... 217 5.6 References ...... 218 6. Conclusions and future work ...... 222 6.1 Conclusions ...... 222 6.2 Future work ...... 225 Appendix I: Additional experimental data for Chapter 2 ...... 228 a) Solid state properties of Janus particles with RB and HP ...... 228

b) Physical form characterisation of Janus particles loaded with RB and carmofur ...... 229

List of Abbreviations

DNA – Deoxyribonucleic acid AJCC- American Joint Committee of Cancer TMN- Tumor-Node-Metastasis RNA- Ribonucleic acid 5-FU- 5-Fluorouracil FDA- Food and Drug Administration FdUMP- Fluorodeoxyuridine monophosphate FdUTP- Fluorodeoxyuridine triphosphate TS- Thymidylate synthase (TS) dNTP- Deoxynucleotide dUTP- Deoxyuridine Triphosphate AC- Acid Cerimidase PDT- Photodynamic Therapy PS- Photosensitizer ROS- Reactive Oxygen Species PCI- Photochemical Internalization MDR- Multidrug resistance EPR- Enhance Permeation and Retention DDS- Drug Delivery Systems API- Active Pharmaceutical Ingredients EHD – Electrohydrodynamic PEO- Poly (ethyline Oxide) PVP- Polyvynilpyrrolidone DOX- Doxorrubicin PLGA- Poly (Lactic-c-glycolic) acid CMC- Critical Micelle Concentration PEI- Polyethyleneimine EC- Ethylcellulose SDS- Sodium Dodecyl Sulfate RB- Rose Bengal HP- Hematoporphyrin PCL- Polycopralactone SEM- Scanning Electron Microscopy TEM- Transmission Electron Microscopy 11

FTIR- Fourier Transformed Infrared spectroscopy DSC- Differential Scanning Calorimetry XRD- X-Ray Diffraction TGA- Thermogravimetric Analysis HDF- Human Dermofibroblasts SLN- Solid Lipid Nanoparticles DLS- Dynamic Light Scattering NSAID- Non-Steroidal Anti-inflammatory Drug IMC- Indomethacin GTS- Glyceryl Tristearate DMF- Dimethylformamide EE- Entrapment Efficiency PdI- Polydispersity Index HSES- High Speed Electrospinning MSES- Medium Speed Electrospinning BTES- Benchtop Electrospinning HPLC- High Performance Liquid Chromatography SD- Spray Drying FaSSIF- Fasted State Simulated Intestinal Fluid TEER- Transepithelial Electrical Resistance BCS- Biopharmaceutical Classification System

List of publications

• Sanchez-Vazquez B, Yu D-G, Pasparakis G, Williams GR. Solid lipid nanoparticles self-assembled from spray drying microparticles. Manuscript

• Sanchez-Vazquez B, Yu D-G, Pasparakis G, Williams GR. (2017) Using electrohydrodynamic techniques to template the self-assembly of solid lipid nanoparticles. Poster presentation. EPSRC EHDA Conference, London, United Kingdom.

• Sanchez-Vazquez B, Amaral AJR, Yu D-G, Pasparakis G, Williams GR. Electrosprayed Janus Particles for Combined Photo-Chemotherapy. AAPS PharmSciTech. 2017;18(5):1460-8

• Sanchez-Vazquez B, Pasparakis G, Williams GR. Electrohydrodynamic tecniques used to create compartmentalized structures. (2016) Poster presentation. PhD Research Day UCL School of Pharmacy, London, United Kingdom.

• Sanchez-Vazquez B, Pasparakis G, Williams GR. (2015) Electrohydynamic techniques for targeted drug delivery. Oral presentation. Biomedical Applications Conference, Eger, Budapest.

List of Figures

Figure 1-1. Bowel Cancer, Five-Year Relative Survival by Stage, Adults (Aged 15-99 Years), Former Anglia Cancer Network, 2002-2006...... 27

Figure 1-2. Lung Cancer, Five-Year Relative Survival Rates by Stage, Adults Aged 15- 99 Years, Former Anglia Cancer Network, from 2003-2006. *Stage IV does not have enough data available...... 27

Figure 1-3. Schematic representation of the cell cycle ...... 29

Figure 1-4. The chemical structure of left:5-FU and right: carmofur..... Error! Bookmark not defined.

Figure 1-5. The mechanism of action of PDT. PDT requires three elements: light, a photosensitizer and oxygen. The photosensitizer is activated by light and it turns from a ground state to an excited state. Energy is released when the excited state returns to the ground state, which helps generate reactive oxygen species which in turn mediate cellular toxicity (Dolmans, Fukumura et al. 2003)...... 34

Figure 1-6. Type I and type II reactions in PDT. Adapted from Zhang et al. (Zhang, Jiang et al. 2018)...... 34

Figure 1-7. Chemical structures of rose bengal sodium salt at the left and haematoporphyrin at the right ...... 36

Figure 1-8. Photochemical internalization of a chemotherapeutic agent. The macromolecule/chemotherapeutic agent is internalized in the cell with the help of the photosensitizer (I). Light exposure triggers the activation of the photosensitizer (III) creating ROS (mainly singlet oxygen) which causes oxidative damage and ruptures the vesicular membranes leading to the release of the chemotherapeutic agent in the cytosol (IV).Subsequently, the chemotherapeutic can find its target in the cytoplasm (V) or in the nucleus (VI). Alternatively, the macromolecules may be degraded by hydrolytic enzymes in late endosomes and lysosomes (II). Image adapted from Berg et al. (Berg, Prasmickaite et al. 2003) ...... 37

Figure 1-9. Diagram representing the therapeutic range/therapeutic window ...... 42

Figure 1-10. An illustration of the common electrospraying (left)/electrospinning (right) experimental setups ...... 49

Figure 1-11. Schematic representation of polymer entanglements ...... 50

Figure 1-12. Diagram depicting the influence of electrospinning/electrospraying parameters on the final product...... 53

Figure 1-13 A schaematic diagram of an electrospun/sprayed drug loaded fibre/particle...... 53

Figure 1-14. Schematic representation of the spray drying instrument ...... 61

Figure 2-1. Schematic representations of Janus type particles. Adapted from Perro et.al. (Perro, Reculusa et al. 2005) ...... 79

Figure 2-2. A schematic illustration of side by side electrospinning (left) and electrospraying (right)...... 82

Figure 2-3. Cell Titer Glo luciferase reaction ...... 88

Figure 2-4. SEM images of the monolithic fibres PVP-HP-F, PVP-RB-F, and PVP-F, prepared at 1 ml h-1, 20 cm tip-to-collector distance and a voltage of 15 kV ...... 90

Figure 2-5. A photograph of the Janus Taylor cone obtained in side-by-side spinning of PVP ...... 91

Figure 2-6. A representative SEM image of PVP Janus fibres (PVP-HP-RB-F) and a graph showing their size distribution...... 92

Figure 2-7. A fluorescence microscopy image of Janus fibre loaded with HP (red) and RB (green) ...... 92

Figure 2-8. An SEM image of Janus fibres loaded with hematoporphyrin and carmofur (PVP-RBC-F), and the fibre size distribution ...... 93

Figure 2-9. FTIR spectra of the raw materials used to fabricate Janus structures ...... 94

Figure 2-10. FTIR spectra of single fluid and Janus fibres containing carmofur and RB ...... 95

Figure 2-11. XRD patterns of the raw materials utilised to fabricate electrospun and electrosprayed materials ...... 96

Figure 2-12. X-ray diffraction patterns of the electrospun Janus fibres...... 97

Figure 2-13. DSC thermograms of the raw materials (PVP, RB, Carmofur). The second heating cycle is shown for PVP...... 98

Figure 2-14. DSC thermograms showing the second heating cycle of PVP-RBC-F, PVP-RB-F, PVP-C-F and PVP-F...... 98

Figure 2-15. Dissolution data for the electrospun fibres and raw APIs. Data are presented as mean ± S.D. from three independent experiments...... 99

Figure 2-16. An SEM image of electrosprayed PVP particles and their size distribution ...... 100

Figure 2-17.SEM images and distribution of PVP-RB-P ...... 101

Figure 2-18. PVP-C-P electrosprayed particles and the size distribution ...... 101

Figure 2-19. An SEM image showing PVP-RBC-P (left) and their size distribution (right) ...... 102

Figure 2-20. A fluorescence microscopy image showing the two compartments of PVP- RBC-P loaded with carmofur(red) and RB (green)...... 102

Figure 2-21. Drug release of PVP-RB-P, PVP-C-P and PVP-RB-P. The particles were placed in a capsule and release undertaken using a PBS buffer (pH 7.4) medium at 15

37ºC and 100 rpm. Percentage of drug release was calculated using UV measurements. Data are from three independent experiments and are given as mean ± S.D...... 104

Figure 2-22. SEM image of electrosprayed PCL-P in chloroform and the parcticle size distribution ...... 105

Figure 2-23. An SEM image and the size distribution of electrosprayed PCL-RB-P .. 105

Figure 2-24. SEM images of PCL-C-P microspheres (left), with the size distribution (right)...... 106

Figure 2-25. An SEM image and the size distribution of PCL-RB-HP-P ...... 106

Figure 2-26. SEM images from electrospraying PCL-RB-C-P ...... 107

Figure 2-27. Cell viability studies with a) HDF cells, b) HDF cells exposed to light at 521 nm, c) A549 cells, and d) A549 cells exposed to light at 521 nm. Data are shown from three independent experiments as mean ± S.D. ***p < 0.001; *p < 0.05 with respect to the control (untreated cells) ...... 108

Figure 3-1. An illustration of SLNs, liposomes and polymeric particles...... 116

Figure 3-2. Schematic representation of the self-assembly of SLNS from electrosprayed particles ...... 118

Figure 3-3. SEM images of unloaded fibres (a,b) 5%GTS, 15 Kv,, 1.5 ml/h, collector distance 20 cm (c,d) 2.5%GTS electrospun fibres, 13Kv, flow rate: 1.0 ml/h collector distance 20 cm...... 126

Figure 3-4. SEM image of PVP-GTS-IMC-P using a) 7 kV at 1ml h-1 with a collector distance of 25 cm b) 11 kV at 1ml h-1 with a collector distance of 25 cm...... 127

Figure 3-5. SEM image of PVP-GTS-IMC-P using a) 6 kV at 1 ml h-1 with a collector distance of 25 cm b) 10 kV at 0.5 ml h-1 with a collector distance of 20 cm ...... 128

Figure 3-6. SEM images of the optimal products from a) electrospinning of PVP-GTS- IMC b) electrospraying of PVP-GTS-IMC, c) electrospinning of PVP-GTS-5FU, d) electrospraying of PVP-GTS-5FU. Conditions used for electrospinning 5FU and IMC: 13Kv, flow rate: 1.0 ml h-1 collector distance 20 cm. Conditions for elecrospraying IMC: 15 kV at 0.8 ml h-1 with a collector distance of 25 cm. Conditions used for electrospraying 5FU: 10 kV at 0.5 ml h-1 with a collector distance of 25 cm ...... 129

Figure 3-7. Chemical structures of a) IMC b) GTS c) 5-FU, d) PVP...... 130

Figure 3-8. FTIR spectra of the raw materials used for formulation development ...... 131

Figure 3-9. FTIR spectra of the fabricated materials ...... 132

Figure 3-10. DSC of the raw materials...... 133

Figure 3-11. DSC thermograms for the formulations loaded with IMC and 5FU...... 134

Figure 3-12. XRD of the raw materials used to create polymer / drug formulations. .. 135

Figure 3-13. XRD patterns of the electrospun and electrosprayed materials ...... 136

Figure 3-14. TEM images of SLNs self-assembled from electrospun/electrosprayed materials ...... 136

Figure 3-15. In vitro release from the optimized IMC SLN formulations. Data are shown as mean ± S.D. from three independent experiments ...... 140

Figure 3-16. In vitro release from the optimized 5-FU SLN formulations. Data are shown as mean ± S.D. from three independent experiments...... 141

Figure 3-17. The Alamar blue reaction, involving the reduction of resazurin to resorufin ...... 142

Figure 3-18. Cytotoxicity of the IMC-loaded SLNs on the Caco-2 cell line, as determined using the Alamar Blue assay. Data from three independent experiments and three different formulation batches of materials are reported as mean ± SD. (n=3) * denotes p<0.05 with respect to IMC...... 143

Figure 3-19. Cytotoxicity of the 5-FU-loaded SLNs on the Caco-2 cell line, as determined using the Alamar Blue assay. Data from three independent experiments and three different formulation batches are reported as mean ± SD. * denotes p<0.05 with respect to 5-FU...... 144

Figure 4-1. A schematic illustration of multineedle arrays for high-throughput electrospinning. Reproduced with permission from (Persiano, Camposeo et al. 2013).Copyright 2013 John Wiley & Sons, Inc...... 149

Figure 4-2. Equipment for the large scale production of fibres and particles by electrohydrodynamic techniques. Images reproduced with permission from http://www.yflow.com/services/custom_devices/ ...... 149

Figure 4-3. A photograph of the TL-Pro-BM Nanofibre Electrospinning Unit...... 154

Figure 4-4. Diagram of the Corona HSES system. 1- High Voltage supply, 2- sharp- edged circular electrode, 3- grounded collector, 4- nanofibre formation, 5-lid, 6-solution inflow, 7- collector conveyor. Reproduced from (Molnar and Nagy 2016) ...... 155

Figure 4-5. SEM images of electrospun fibres: a) MSES-IMC-F, b) HSES-IMC-F ..... 159

Figure 4-6. Optimisation of the MSES for the electrospinning of SLN-5FU fibres ...... 160

Figure 4-7. SEM Images for the scaled-up electrosprayed particles a) MSES-IMC-P and b) MSES-5FU-P ...... 161

Figure 4-8. SEM images of formulations prepared by spray drying: a) SD- IMC-P, b) SD-5FU-P ...... 162

Figure 4-9. FTIR spectra of the scaled up formulations...... 163

Figure 4-10. XRD diffraction patterns of IMC and 5-FU-loaded scaled up formulations...... 164

Figure 4-11. DSC thermograms of the scaled-up formulations. Exo up...... 165

Figure 4-12- TEM images SLNs self-assembled SLN from the electrospun, electrosprayed, and spray dried materials. F1 and F4, F7 and F10 are reported in Chapter 3, Figure 3-14...... 167

Figure 4-13. Cumulative drug release from indomethacin loaded SLNs (F2-F5) in PBS. Data are given as mean ± S.D. from three independent experiments...... 171

Figure 4-14. Cumulative drug release from 5-FU loaded SLNs (F8-F12) in PBS. Data are given as mean ± S.D. from three independent experiments...... 172

Figure 4-15 Cumulative IMC release from F2, F5 and F6 in FaSSIF. Data are given as mean ± S.D. from three independent experiments...... 173

Figure 4-16. Cumulative drug release of 5FU from SLN formulations in FASSIF Data are given as mean ± S.D. from three independent experiments...... 174

Figure 4-17. Percentage of drug found in the cell lysate. Data are given as mean ± SD (n=3) ...... 176

Figure 4-18. Cell viability of Caco-2 cells exposed to the IMC loaded SLNs. Data are given as mean ± S.D. from three independent experiments with three replicates in each.* denotes p<0.05 with respect to IMC...... 178

Figure 4-19. Viability vs. PDI for IMC-loaded SLNs ...... 178

Figure 4-20. Cell viability of Caco-2 cells exposed to 5FU loaded SLNs from three independent experiments. Results represent mean ± SD from three independent experiments with three replicates in each. * denotes statistically significant differences (p<0.05) from 5-FU...... 179

Figure 4-21. The results of stability studies performed for 6 months at room temperature. Data are shown from 3 independent experiments as mean ± S.D. * denotes p < 0.05 with respect to t0...... 180

Figure 5-1. The generic chemical structure of poloxamers, in which a=20,b=27 ...... 188

Figure 5-2. The chemical structure of DOPC ...... 189

Figure 5-3. Schematic representation of the proposed self-assembled Janus structure ...... 190

Figure 5-4. Configuration of triaxial spinneret. Left: two inner nozzles concentrically arranged; right: two inner nozzles side-by-side ...... 191

Figure 5-5. Schematic representation of the triaxial spinneret ...... 196

Figure 5-6. Formation of micelles upon water addition. Left: the dried film; right: the micelles after the addition of water...... 197

Figure 5-7. TEM images of a) unloaded and b) HP- loaded micelles...... 198

Figure 5-8. The results of failed electrospinning a micellar suspension with 5% w/v PVP ...... 198

Figure 5-9. Fibres prepared from a micellar suspension with 10% w/v PVP (M_HP- PVP) ...... 199 18

Figure 5-10. TEM images of a) micelles-HP and b) V-micelles ...... 199

Figure 5-11. SEM images of DOPC-loaded electrospun fibres and their diameters: . 200

Figure 5-12. TEM images from the liposomes self-assembled from ...... 201

Figure 5-13. An SEM image of fibres produced from triaxial electrospinning at a flow rate of 1 mL/h for the sheath and 0.3 mL/h for the core fluids...... 202

Figure 5-14. Fibres prepared by triaxial electrospinning using a flow rates of 0.1 ml/h for the core fluids and 1 ml/h for the shell...... 202

Figure 5-15. a) SEM image of V_3A-F; b) size distribution of the fibres in V_3A-F; c,d) TEM images of V_3A-F...... 203

Figure 5-16.Data on the 3A-F fibres: a) an SEM image ; b) a frequency graph showing the fibre diameters; c,d) TEM images ...... 204

Figure 5-17. TEM images of the Janus particles self-assembled from triaxial electrospun fibres...... 205

Figure 5-18. FTIR spectra of the raw materials utilised for the triaxial electrospinning process...... 207

Figure 5-19. FTIR spectra of the triaxial drug-loaded electrospun fibres, M_HP-PVP, 3A-F and DOPC-PVP-C...... 208

Figure 5-20. XRD patterns of the raw materials...... 209

Figure 5-21. XRD patterns of blank fibres: V_M-PVP, V_3A-F, V_DOPC-PVP and drug- loaded fibres M_HP-PVP, 3A-F, DOPC-PVP-C...... 210

Figure 5-22. DSC thermograms of DOPC and P188...... 211

Figure 5-23. DSC thermograms of the triaxial fibres...... 212

Figure 5-24. Drug release from single fluid fibres and triaxial fibres. Mean average shown (n=3) with error bars representing S.D...... 213

Figure 5-25. Cell viability of A549 cells exposed to the self-assembled structures. ... 215

Figure A1-1. XRD patterns for the raw materials (hematoporphyrin & rose bengal) and electrospun fibres PVP+HP, PVP+RB and the Janus fibres...... 228

Figure A1-2. DSC data for PVP fibres loaded with rose bengal and hematoporphyrin...... 228

Figure A1-3. FTIR spectra of single fluid and Janus particles containing carmofur and rose bengal ...... 229

Figure A1-4. XRD patterns of the electrosprayed particles PVP-RB-P, PVP-C-P, PVP- RBC-P and PVP-P...... 229

Figure A1-5. DSC data for carmofur, rose bengal , PVP-RB-P, PVP-C-P, PVP-RBC-P and PVP-P...... 230

List of Tables

Table 1-1. Methods by which DDS can overcome challenges in drug delivery. Adapted from Allen and Cullis (Allen and Cullis 2004)...... 46

Table 1-2. Some representative drugs processed into electrospun/electrosprayed materials. Abbreviations: PLA:polylactic acid, PLGA: poly(lactic-co-glycolic acid), PLLA: polylactid acid, PCL: polycopralactone, PU: polyurethane, PEG: polyethylene glycol, PEO:polyethylene oxide...... 55

Table 1-3. Chemotherapeutic drugs entrapped by EHD techniques ...... 57

Table 1-4. Electrosprayed particles encapsulating cancer therapeutics ...... 58

Table 1-5. Comparison between spray drying and EHD. Adapted from Sosnik and Serementa (Sosnik and Seremeta 2015) ...... 63

Table 2-1. Selectivity index for the formualtions containing RB, Carmofur and PVP-RB- Carmofur ...... 109

Table 3-1. Polymer solutions used for single fluid electrospinning/spraying. PVP (MW 360KDa was used to fabricate the fibres, and PVP MW 40 KDa was used to fabricate particles). Fibres and particles used as vehicle were fabricated using the same ratios, omitting the drug content...... 120

Table 3-2. Particle sizes of the SLNs as determined from TEM and DLS...... 137

Table 3-3. Results of encapsulation efficiency experiments. Values are shown as mean ± standard deviation (n = 3)...... 138

Table 4-1. Identification of the materials and SLNs in Chapter 4...... 158

Table 4-2. The diameters of IMC loaded fibres prepared using different electrospinning variants ...... 159

Table 4-3. Size comparison of the scale up process of 5FU loaded fibres ...... 160

Table 4-4. The sizes of the SLNs, as determined from TEM images and DLS (n=number of particles counted)...... 168

Table 4-5. Entrapment efficiency of the formulations (n=3)...... 168

Table 5-1. Electrospinning fibre composition ...... 195

Table 5-2. DLS size and entrapment efficiency of the self-assembled structures...... 206

Impact statement

The delivery of drugs is of paramount importance: regardless of how potent an active pharmaceutical ingredient is, if it is not delivered effectively to its target it will not achieve the desired therapeutic effect. A failure to achieve targeted delivery of APIs can cause numerous side effects and adverse reactions (e.g. nausea, hair loss, cardiomyopathy).

Only very rarely is a drug is administered in its pure form; most often, a vehicle is needed for each and every drug to improve factors such as solubility, permeability, stability and potency. Previous research suggests that targeting drug delivery to specific parts of the body or on a cellular level has many advantages such as reduction of doses and of side effects.

To achieve an optimal drug delivery system, the drug should be encapsulated in a functional and biocompatible carrier system. The development of methods to improve drug delivery can be described in four categories: (1) routes of delivery; (2) delivery vehicles; (3) cargo (e.g. drug, proteins); and, (4) targeting to specific sites. The research in this thesis attempts to prepare novel delivery vehicles for commercially available drugs, ranging from model drugs to chemotherapeutic and photodynamic therapy agents. Key findings of the work include a photochemotherapy formulation with exceptionally high selectivity for cancer cells over non-cancerous cells. Further, a range of sacrificial templates was prepared from which drug-loaded nanoparticles could be self- assembled. These materials led to enhanced stability and extended drug release.

The new knowledge generated in this research will aid future scientists to improve the delivery of diverse pharmaceutical ingredients by incorporating them into vehicles prepared using techniques such as electrospinning, electrospraying and spray drying: the outcomes of this research will thus be very informative to researchers seeking to optimise the delivery of a particular drug.

1. Introduction

1.1 Cancer

Over the past 6 million years humans have evolved from a single celled organism to a complex multi celled organism. The evolution process was driven by random changes in our deoxyribonucleic acid (DNA), called mutations. DNA is genetic material containing encoded information that our cells need to function and is found inside cells. Our cells divide countless number of times during our lifetime, which means that an exact copy of

3 billion nucleotides need to be transmitted to daughter cells (Pray 2008). While most of the time DNA replicates with high fidelity, sometimes mismatches in the DNA are generated (Chen, Miller et al. 2014); there exist mismatch repair mechanisms, but not all mismatches are corrected. In the latter case, they become permanent mutations which can either be beneficial or disadvantageous (Carlin 2011).

If an individual accumulates enough mutations, the end result could be cancer: this arises when mutated cells do not obey the rules or structural design of normal cells and start to grow excessively and multiply; this growth will eventually lead to a tumour (Knudson and

Strons 1972).It is not clear how many mutations must accumulate for cancer to develop

(Boland and Ricciardiello 1999). Certain agents and factors are known to accelerate the development of cell mutations however; for instance smoking, exposure to harmful chemicals, chronic health conditions (e.g. diabetes, chronic kidney disease), smoking, and alcohol, among others (Tu, Wen et al. 2018).

Cancer begins with a primary tumour, which is where cells start dividing excesively and lead to a cancerous mass. The primary tumour can ultimately lead to a secondary tumour or metastasis, which is the spread of tumour cells to other sites of the body

(Weinberg 2013). The development of cancer can affect various aspects of the body such as the blood circulation, or the immune, hormone and lymphatic systems (Hong

and Hait 2010). In their early stages, cancers are mostly asymptomatic, but when cancer begins to advance it starts to cause symptoms that can be confused with other non- cancerous diseases such as viral infections (e.g. mononucleosis, common cold, flu, etc)

(Silverstein, Silverstein et al. 2006). If cancer is not treated promptly it can be deadly.

According to the the USA National Cancer Institute, cancer is one of the leading causes of death worldwide, accounting for nearly 1 in 6 deaths. There were 14 million new cancer cases in 2012, and this number is expected to rise to 22 million in the next two decades.

The latest data available show that in the United Kingdom there were 356,860 new cases of cancer in 2015 and 164,444 cancer-related deaths in 2014 (Cancer Research UK,

2018).

1.1.1 Tumour physiology

There are differences in physiology between normal tissue and malignant tissue that arise from the tumor vasculature. The vasculature of the tumour is firstly composed of the vessels that come from the normal tissue where it arose (Brown and Giaccia 1998).

In addition, when a tumour proliferates and reaches a size of 2 mm3, the cells enter a hypoxic state as the vasculature fails to provide a sufficient oxygen supply

(Wachsberger, Burd et al. 2003). As a consequence, cells start to secrete growth factors that lead to the formation of new blood vessels from the existing ones; this process is known as angiogenesis (Bergers and Benjamin 2003), and leads to the formation of more vessels that are irregular in shape and defective. As a result, there is irregular and inconsistent blood flow in the vessels. This enhances their vascular permeability, leading to increased leakage of blood plasma components into the tumour tissues (Maeda, Wu et al. 2000). Both types of vessels develop structural and physical abnormalities and do not present any lymphatic system, which causes the extracellular fluid of the tumour tissues to be retained within the tumour (Maeda 2001). These differences between normal and cancerous tissue provide opportunities for drug targeting, as will be discussed later in this chapter. 24

1.1.2 Cancer stages

The American Joint Committee on Cancer (AJCC) developed a system for staging cancer to promote the use of a universal standard for reporting clinical and scientific research (Egner 2010). The different stages reflect how advanced the cancer is and mainly depend on three parameters: the location of the cancer, whether it has spread, and whether it is affecting several parts of the body (Greene, Balch et al. 2002). To determine the specific stage of the cancer, physicians or oncologists often use the

Tumour-Node-Metastasis (TMN) staging system for solid tumours (Greene and Sobin

2002). This system involves:

1) The size and location(s) of the primary tumour (T) – this can lie in the range of 1-4

with 1 being small (0.1-2 cm) and 4 large (<5 cm).

2) Indicates if the tumour has tumour spread to the lymph nodes, and if so, where and

to how many (N). N can vary from 0 (no lymph nodes) to three (large number of

nodes presenting cancer cells).

3) Has the cancer spread to other part of the body; if so, where and how much (M)?

M takes values of 0 (cancer has not spread) or 1 (cancer has spread).

4) Are there any biomarkers linked to the cancer that make it more or less likely to

spread (Strimbu and Tavel 2010)?

After obtaining the results of the TMN staging system, the stage of the cancer can be determined. Most types of cancers have four stages, I – IV, but some also have stage 0, such as breast and skin cancer.

Stage 0- This stage describes that the cancer is “in situ”. The cancer is still located in the place it started and it has not spread to nearby tissues. This stage is highly curable and removing the tumour with surgery often treats it.

Stage I- This stage is where the cancer has not grown deeply into nearby tissues and the lymph nodes and other parts of the body are not affected. It is often referred to as

“early-stage cancer”.

Stage II and Stage III- These stages are an indication of larger tumours that have grown deeply into nearby tissue. They may have spread to the lymph nodes but not to other parts of the body.

Stage IV- This stage is also known as “advanced or metastatic cancer” where the cancer has spread to other organs or parts of the body (Socinski, Morris et al. 2003).

For localized tumours (for instance in Stages 0 or I), treatments such as surgery, radiotherapy or photodynamic therapy are recommended. However, when the cancer has spread, more systemic treatments are needed, these treatments consist of chemotherapy, hormone therapy and targeted cancer drugs. In England and Northern

Ireland, most are diagnosed at an early Stage I or II (54-55%) rather than a later Stage

III or IV (45-46%). Around 25-27% have metastases at the diagnosis Stage IV (Cancer

Research UK 2017).

1.1.3 Cancer survival rates

Survival rates give an overall percentage of people who survive a certain type of cancer for a specific amount of time, often using an overall of five-year survival rate (National

Cancer Institute 2018).

Cancer survival rates can be obtained from Cancer Research UK (Cancer Research UK

2017). The 5-year bowel cancer survival rates in the UK can be seen in Figure 1-1. The data are divided by gender, and although there is not much difference in outcomes for male and female patients there is a clear decline in survival rates from Stage I-IV. Similar data are seen for lung cancer (Figure 1-2).

Figure 1-1. Bowel Cancer, Five-Year Relative Survival by Stage, Adults (Aged 15-99 Years), Former Anglia Cancer Network, 2002-2006.

Figure 1-2. Lung Cancer, Five-Year Relative Survival Rates by Stage, Adults Aged 15-99 Years, Former Anglia Cancer Network, from 2003-2006. *Stage IV does not have enough data available.

It can be clearly seen that even though the survival rates of different types of cancers are very different, the common factor is the decline of relative survival rates as Stage IV is approached.

1.2 Current cancer treatments

To cure or control cancer, or ease cancer symptoms, the common treatments used today are chemotherapy, radiotherapy, and surgery. In England from 2013-2104, 27% of patients diagnosed with cancer had radiotherapy, 44.8% had surgery and 28.4% received chemotherapy (UK 2017). Less common treatments include photodynamic therapy, immunotherapy, hormone therapy and stem cell transplants (Society 2018).

Some people will only have one treatment, but most people have a combination (National

Cancer Institute 2017).

Radiotherapy is a treatment that involves the use of high-radiation energy to kill cancer cells, while surgery consists of the removal of a tumour and surrounding tissue in an operation (DeVita, Lawrence et al. 2010). The principles of chemotherapy and photodynamic therapy are described in more detail below.

1.2.1 Chemotherapy

Chemotherapy is defined as the use of drugs to treat cancer (Boesen and Davis 1969).

The purpose is to prevent cancer cells from multiplying, invading, metastasizing and ultimately resulting in the death of the host. The dosage form of chemotherapies is varied and they can be administered orally, topically or by subcutaneous, intraventricular, intrapleural, intraperitoneal, intra-arterial or intravenous injection (Chabner and Longo

2011).

There are more than 200 molecules used as chemotherapeutics, for example cis-platin, doxorubicin, and paclitaxel, among others (Chabner and Roberts 2005). The treatment regimen depends on the type of cancer, the stage of cancer, previous treatments, and the patient's overall health (e.g. diabetes, heart disease) (Pal and Hurria 2012). The goal 28

is to find an agent that can inhibit growth (cytostatic) and kill (cytotoxic) cancerous cells and at the same time have a minimal effect on the healthy tissues of the patient.

However, most traditional chemotherapies have a major effect on the proliferation of all cells (Eder and Skarin 2010), and since multiplication is a characteristic of healthy cells too, most chemotherapies also have toxic side effects on normal cells (Aldred, Buck et al. 2009). This is particularly problematic for cells with a fast cell division cycle, such as the bone marrow and the mucous membranes (Balducci 2007).

As chemotherapy works on active cells, to have a better understanding it is necessary to understand the cycle of the cell. This is the process in which a cell divides (Cooper

2000). In normal cells, this process is regulated by complex biological pathways, but in cancer cells these pathways malfunction, leading to the overgrowth of cancer cells.

(Aldred, Buck et al. 2009) The cell cycle is divided in four phases (Figure 1-3)

Figure 1-3. Schematic representation of the cell cycle

• G1 – In this phase the cell grows and prepares for DNA synthesis. The G1 phase duration is approximately 18-30 h.

• S phase – The 36 chromosomes are duplicated by the cell. This lasts around 18- 20 h.

• G2 – In this phase the cells grow and produce new proteins. The duration is usually 2-10 h • M phase – The cell divides; this process takes around 30-60 minutes

There are several types of chemotherapeutic agents known as antineoplastic agents, which means that they inhibit or halt the development of a tumour (Eldridge and Davis

2018). There can be classified as alkylating agents, antimycotics, antimetabolites, topoisomerase inhibitors and miscellaneous antineoplastics (Langevin and Atlee 2007).

Alkylating agents alter the DNA structure or function, and they are not-specific towards the cell cycle (Colvin 2002). Antimetabolites induce cell death during phase S, when incorporated into RNA and DNA, or inhibit the enzymes needed for nucleic acid production. Antimycotics induce vasculature damage by preventing blood-vessel formation to the tumour. Topoisomerases stop DNA replication by blocking the topoisomerases enzymes I and II (Tanaka, Matsushima et al. 2009). Meanwhile, there are other type of agents that do not fall in any of the above classifications, such as enzymes, retinoids, among others.

The chemotherapy drugs used in this thesis, 5-fluorouracil (5-FU; Error! Reference source not found.) and its prodrug carmofur (Error! Reference source not found.), lie in the antimetabolites group. These are commonly used as model drugs for drug delivery systems and thus were chosen for this thesis. 5-FU is an Food and Drug Administration

(FDA) approved chemotherapeutic which is currently used to treat cancers including breast, bowel, skin, stomach, oesophageal and pancreatic cancer (Macmillan-Cancer-

Support 2018). 5-FU can be given by intravenous administration or alternatively as a cream to treat skin cancer (Sharquie and Noaimi 2012). There are several side effects which are known to arise when taking 5-FU. Most are related to the fact that it reduces the number of white blood cells (neutropenia), red blood cells (anaemia) and platelets in

the patient’s body; as a consequence, they experience infections, bruising, bleeding and tiredness (Kumar, Kumar et al. 1999).

In some cases, it has been proven that using adjuvant therapy provides a reduction in mortality for cancer patients, as is the case with 5-FU and oxiliplatin (Andre, Boni et al.

2009). The addition of a second chemotherapeutic agent improves the survival rate but also increases the toxicity (Sanoff, Carpenter et al. 2012). It was found that the addition of oxaliplatin increased paresthesia, severe neutropenia, severe nausea, vomiting and diarrhe. (André, Boni et al. 2004).

a) 5-Fluorouracil

5-FU (Error! Reference source not found.) has cytostatic and cytotoxic effects on cells.

5-FU is converted intracellularly in active metabolites: fluorodeoxyuridine monophosphate (FdUMP), fluorodeoxyuridine triphosphate (FdUTP) and fluorouridine triphosphate (FUTP). FdUMP binds to an enzyme called thymidylate synthase (TS), blocking the synthesis and incorporation of thymine nucleotide which is necessary for

DNA replication and repair. This also results in (deoxynucleotide) dNTP imbalances and increased deoxyuridine triphosphate dUTP, both of which cause DNA damage (Longley,

Harkin et al. 2003). FUTP is incorporated into RNA, disrupting normal RNA processing and function (Kufe and Major 1981).

a) b)

Figure 1-4. The chemical structure of a) 5-FU and b) carmofur.

b) Carmofur

Carmofur (1-hexylcarbamoyl-5-fluorouracil, Car) depicted in Error! Reference source not found., is a 5-FU prodrug used as an antineoplastic agent (Malet-Martino, Jolimaitre et al. 2002). It is used as an adjuvant chemotherapy for colorectal cancer (Ito, Yamaguchi et al. 1996, Nakamura, Ohno et al. 2001). The main disadvantage of carmofur is its low solubility and chemical instability: it hydrolyzes relatively easy under physiological conditions (Domracheva, Muhamadejev et al. 2015).

Carmofur is a highly potent acid ceramidase inhibitor (Pizzirani, Pagliuca et al. 2013).

Acid ceramidase (AC) is a cysteine amidase that catalyzes the hydrolysis of the pro- apoptotic lipid ceramide. AC is involved in modulating the ceramide levels in cells, influencing the survival, growth, and death of cells (Mao and Obeid 2008). AC levels are higher in various types of human cancer cells than in healthy cells, and interfering with its activity is a crucial component of carmofur’s anti-neoplastic properties (Realini,

Solorzano et al. 2013).

1.2.2 Photodynamic therapy

Another therapy used to treat cancer is photodynamic therapy (PDT). This is an FDA approved treatment for non-small cell lung cancer and oesophagal cancer (Patrizia,

Kristian et al. 2011). PDI involves injecting a light-sensitive drug called a photosensitizer

(PS) (a molecule capable of producing a chemical change in another molecule upon absorption of light) into the bloodstream or applying it topically (in the case of skin cancer). Approximately 24-72 h after administration, the tumour is exposed to light at a certain wavelength using a lamp or laser being shone on the treatment area for 10-15 min. This is facile in the case of skin cancer, but for internal tumours can be problematic.

For instance, for tumour in the lungs an endoscope must be used (Simone, Friedberg et al. 2012). Upon exposure to light, the photosensitizer produces singlet oxygen and reactive oxygen species that destroy the surrounding cells (Figure 1-5) (Berg, Selbo et al. 2005). 32

The therapy relies on the absorption of a photon of light of a particular wavelength by the

PS; the latter will then be converted to a short-life excited state (PSEs). This can decay back to the ground state, or undergo a rearrangement to a longer-lived state (triplet) state

(PSEt) (Wilson 2002). When it reaches the triplet state, it can undergo two types of reaction. In type I reactions the triplet state reacts with a substrate, such as the cell membrane, and transfers an electron to form reactive oxygen species (ROS) such as superoxide anion radicals, hydroxyl radicals and hydrogen peroxides (Zhang, Jiang et

3 al. 2018). Type II reactions involve exchange energy with molecular oxygen ( O2) to

1 produce highly active singlet oxygen ( O2). These reactions occur simultaneously, a schematic diagram can be seen in Figure 1-6 (Wilson 2002, Dolmans, Fukumura et al.

2003).

Figure 1-6. Type I and type II reactions in PDT. Adapted from Zhang et al. (Zhang, Jiang et al. 2018).

The phototoxicity produced by PDT does not affect the collagen or elastin of the tissue, and hence healing occurs with no long-term scarring or side effects (Webb and Jones

2004). PDT is also a low-cost therapy and can be repeated without cumulative toxicity

(Saini, Lee et al. 2016). When used properly, it can be an effective cancer treatment for patients with localized lesions (Triesscheijn, Baas et al. 2006).

However, there are several disadvantages associated with PDT. The first thing to take into consideration is the delivery of the photosensitizer to the target tissue. One of the main disadvantages is that most sensitizers are retained by both cancerous and normal cells, although they are retained by the former for longer periods of time (Hsi, Rosenthal et al. 1999). This means that non-specific effects can arise. The second consideration is that of how to apply light to the target tissue (Henderson and Dougherty 1992,

Shafirstein, Bellnier et al. 2017).

Three mechanisms are known to describe how PDT acts to destroy a tumour. First, it can destroy the micro blood vessels of the tumour, leading to tumour necrosis as vasculature is essential for tumours to obtain the necessary amount of nutrients

(Wilczewska, Niemirowicz et al. 2012). Second, the reactive oxygen species generated can kill tumour cells directly, although complete tumour eradication is not possible by this mechanism alone (Huang, Xu et al. 2008). One of the reasons for this is that the photosensitizer is not distributed evenly within the tumour, making total tumour destruction unlikely to happen. Oxygen availability will also be varied throughout the tumour tissue. Finally, PDT stimulates anti-tumour immunity, as it triggers and inflammatory response and causes photo-damage to drug efflux pumps (Tang, Zhang et al. 2009).

In summary, PDT offers advantages for the treatment of certain cancers. It is a localized treatment, thus it avoids the side effects caused by the systemic distribution of cytotoxic agents. PDT is an alternative for when the removal of a tumour is not possible by surgery.

Finally, PDT has a lower cost than chemotherapy and it can be repeated for several cycles without damaging the nearby tissues (Kubler, Niziol et al. 2005).

Today, there are two types of commercially available PS for clinical use: aminolevulinic acid and derivatives (Hadjipanayis, Widhalm et al. 2015), commercialized as Levulan®,

MetVix®, and Levulan®, and the chlorin-based PS Photochlor® (Zhang, Jiang et al. 2018).

The two photosensitizers used in this project are rose bengal and haematoporphyrin

(Figure 1-7), the reason being model drugs were needed as the aims of the thesis focus more in proof of concept.

a b

Figure 1-7. Chemical structures of a) rose bengal sodium salt and b) haematoporphyrin

1.2.3 Photo-chemotherapy

Photo-chemotherapy is the combination of two cancer treatments: chemotherapy and photodynamic therapy (Dalla Via and Marciani Magno 2001). There are several reasons why combined photodynamic therapy and chemotherapy can provide a more effective approach to treat cancer.

Some photosensitizers are mainly located intracellularly, in endocytic vesicles (Berg,

Prasmickaite et al. 2003). When the photosensitizer is activated by light it induces the rupture of these vesicles and allows the release of endocytosed molecules into the cytosol. This process is called photochemical internalization (PCI), and it can be used to improve therapeutic efficacy and specificity (Berg, Weyergang et al. 2010). PCI brings substantial advantages to chemotherapy as it involves the drug being localized inside endocytic vesicles, and after the rupture of these vesicles is triggered by the reactive oxygen species (ROS) generated by light exposure, they release the cytotoxic agent in the cytosol as shown in Figure 1-8 (Berg, Weyergang et al. 2010). This mechanism can be applied to facilitate the cytosolic delivery of drugs that do not normally enter cells

(Spring, Rizvi et al. 2015). PCI has also been recently studied as a technique to avoid

neuronal toxicity; it was found that in in vitro models, neurons in culture survived PCI treatment under conditions sufficient to kill tumour cells (O’Rourke, Hopper et al. 2017).

Cancerous cells posses several mechanisms which can render standard treatments ineffective, such as drug-efflux proteins that lead to multi-drug resistance: this is a major factor in the failure of many forms of chemotherapy (Goldman 2003, Alfarouk, Stock et al. 2015). Drug-sensitive cells will be killed, leaving behind drug-resistant cells.

Consequently, as the tumour begins to grow, there is a possibility that chemotherapy will fail because of the high level of heterogeneity of the tumour (Ramakrishna, Fujihara et al. 2006, Burrell and Swanton 2014)(Ramakrishna, Fujihara et al. 2006, Burrell and

Swanton 2014). As mentioned before, PDT casues photo-damage to drug efflux proteins and thereby makes multi-drug resistant (MDR) cells more sensitive to the chemotherapy 37

treatment, as the release of ROS can stimulate proteins that suppress the expression of

MDR genes (Kuo 2009).

The combination of photo- and chemotherapies has been successfully used to treat cancer or cancer related symptoms on several occasions. For instance, in a trial conducted by Kimura et al., 12 patients suffering from advanced lung cancer with airway stenosis were given photo-chemotherapy for local control of the lesions situated within the lumen (e.g. inside an artery. This was successful in improving quality of life and reducing symptoms, in terms of increased opening to the airways and prevention of obstructive pneumonia (Kimura, Miyajima et al. 2015).

It has also been shown that photo-chemotherapy significantly decreases the self-renewal capacity of metastatic melanoma cells (Biteghe and Davids 2017), and Zhou demonstrated that the combination therapy exhibited a synergetic anticancer effect

(Zhou, Zhou et al. 2014). In other work, Canti tested the combinational therapy in murine cancers, concluding that it was able to greatly reduce the effective doses required, thus lowering toxic effects on normal host tissues (Canti, Nicolin et al. 1998).

Generally, PDT and chemotherapy are given separately. However, it would be much more convenient for patients and therapeutically more effective to simultaneously deliver the two active agents to the same target cell using a single drug carrier, in order to provide exposure to synergistic drug combinations and maximize cytotoxicity. However, successful incorporation of ingredients with different properties such as solubility present a challenge when fabricating these drug carriers (Tiancheng, Xiabin, et al, 2012). In addition, careful consideration regarding drug interactions must be taken when formulating dual drug delivery The combination of a photosensitizer and chemotherapeutic in a single formulation has been explored as a potential way to synergise ROS-mediated cancer cell necrosis with the apoptotic events driven by chemotherapeutic agents (Zhou, Zhou et al. 2014). The most common route of

administration (intravenous) is not able to deliver the agents simultaneously at the same tumour tissue, making it difficult for them to accumulate and enter the same cancer cells at the same time (Wolinsky, Colson et al. 2012).

1.2.4 Cancer treatment costs

Even though cancer therapeutics have been on the market for a long time, the prices are often very high. The NHS states that, according to a report from the private healthcare provider BUPA, in 2010 the approximate cost for the diagnosis and treatment of cancer in the UK was around £30,000 per patient: given that there were approximately 318,000 cancer patients, this amounts to a total cost of £9.5 billion. In 2021, the total costs are expected to rise to £15.3 billion, which is equivalent to £40,000 per cancer patient (BUPA

2011). Over a quarter of the expenditure goes to hospital inpatient costs (excluding surgery), 22% goes on the cost of surgery, and 18% on drug treatments: this means only around £1.7 billion is used for the latter (NHSChoices 2011). A newer report by Cancer

Research UK states that new treatments such as immunotherapies, have price tags of more than £100,000 per patient per year (BUPA 2011). It is clear that new, low-cost, anticancer treatments could have major benefits to patients and healthcare providers.

1.3 Challenges for an efficient treatment

Modern medicines promise to introduce novel therapeutic modalities that will transform current chemotherapies, improving the quality of life of patients (Banerjee and Sengupta

2011). Using nanotechnology in the pharmaceutical field can lead to many advantages such as improving the stability, biodistribution and pharmacokinetics of drugs, resulting in improved efficacy (Devalapally, Chakilam et al. 2007). As a consequence, adverse side effects and toxicity are reduced since there is a favourable accumulation at target sites, for instance via the enhanced permeation and retention (EPR) effect (a detailed explanation can be found in Section 1.4.2) (Bamrungsap, Zhao et al. 2012). Thus, nanoscale formulations have the potential to overcome a number of the targeting issues which exist with current therapies. This could be very beneficial because the 39

development of a new drug molecule is expensive and time-consuming, whereas improving the efficacy of “old drugs” is more rapid and cost-effective (Tiwari, Tiwari et al.

2012).

Systemic chemotherapy is the most used therapeutic strategy and involves intravenous administration at the maximum tolerated doses, having as a consequence severe toxicity in many healthy tissues (Kreidieh, Moukadem et al. 2016). Over-dosing in chemotherapy can be lethal; therefore, extreme care should be taken to determine the effective dosage for a particular patient. (Schulmeister 2006) However, to date there is not an established formula to calculate the dose required to achieve an optimal systemic drug exposure in each individual patient. While the conventional cytotoxic chemotherapy regimen has a very important role in patients diagnosed with cancer, the use of these agents has clearly been associated with long-term toxicities in long-term survivors (Duffner 2006). The approach of using more potent but non-specific regimes to improve cancer survival will only bring an increase in the number of patients suffering from these long-term toxicities

(Palumbo, Kavan et al. 2013). One major drawback of current oncology treatments is their low therapeutic index (maximum tolerated dose/curative dose), and to try to overcome this problem there is a need to create efficient delivery systems for the currently available drugs (Vasir and Labhasetwar 2005).

Chemotherapy also encounters biological barriers such as vasculature walls and the interstitial space within the tumour, among others, resulting in only a fraction of the administered dose reaching the target cancer cells (De Souza, Zahedi et al. 2010). New treatments and technologies or personalized medicines could be the answer to reduce treatment costs and improve patient well-being (Mathur and Sutton 2017, Dolgin 2018).

There have in recent years been significant developments in new therapeutics (Savage and Mahmoud 2015), but problems remain since many drugs have unacceptable side effects because of unwanted interactions between the medication and untargeted parts

of the body, and because of the accumulation of the drug in non-cancer tissues (Himri and Guaadaoui 2018). A major challenge lies in developing targeted medicines, and site- specific drug delivery is one of the main goals in contemporary pharmaceutics (Kim and

Nie 2005). To achieve this goal there are several steps to take into consideration: a) localization of the drug into the target organ; b) recognition and interaction between the organ and drug formulation; and, c) delivery of a therapeutic concentration to the target site with, preferably, no uptake by non-targeted cells. Additionally, the carrier must protect the active ingredient from degradation and inactivation during transit to the target, and should also be biodegradable and non-immunogenic (Poste and Kirsh 1983,

Liechty, Kryscio et al. 2010). The past decades have seen increasingly rapid advances in the development of targeted drug delivery methods, including polymer conjugates, nano and microparticles, liposomes, and micelles (Wilczewska, Niemirowicz et al. 2012).

1.4 Drug delivery systems

Drug delivery is a method of administering a pharmaceutical compound to achieve a therapeutic effect. To obtain the maximum benefit of the active pharmaceutical ingredient

(API), the drug should be delivered at the right concentration at the correct rate, in the target site, at the appropriate time. This is a goal that is rarely satisfactorily achieved

(Tiwari, Tiwari et al. 2012).

Conventional drug delivery formulations (e.g. tablets) usually provide an immediate release. Immediate release tablets disintegrate rapidly and release the API, but usually do not prolong the rate of drug release/absorption (Ghosh, Bhuiyan et al. 2012). This results in a rapid increase in the concentration of the drug level in the bloodstream, but no long term control of this. As a result, to maintain the desired therapeutic concentration over a prolonged time period, the drug needs to be administered at a particular frequency. The frequency of administration depends on the half-life and therapeutic index of the drug; a need for frequent administration leads to poor patient compliance, increasing the chances of missed doses. In addition, the fluctuation of the concentration 41

in the plasma may lead to under- or overmedication (Figure 1-9), which could be dangerous with drugs with low therapeutic indices such as anti-cancer drugs. To avoid side effects some medication can be administered locally (topical administration), but many can only be applied systemically.

Figure 1-9. Diagram representing the therapeutic range/therapeutic window

In conventional drug delivery the way the API is distributed into the body is mainly dependent on its physicochemical properties (e.g. partitioning coefficient). Conventional oral delivery systems such as tablets, capsules or solutions do not protect the molecule from degrading factors, but these systems can be modified to improve the delivery of the

API, for example in the case of enteric coated tablets. In contrast, advanced drug delivery systems release the drug in a dosage form in a more precisely controlled manner by using carrier systems (Muller and Keck 2004).

The prospect of controlling the pharmacokinetics, pharmacodynamics, toxicity, immunogenicity and efficacy of an API has led to much effort being applied in the design of novel drug delivery systems (DDS). DDS can conveniently be divided into two major groups: modified release systems, and targeted drug delivery systems. These will be discussed in more detail below. 42

1.4.1 Modified release

There are several methods to modify the rate at which the drug is released from its carrier. These include:

Delayed release - The drug is not released immediately after administration, but will stay in the carrier until desired (e.g. enteric coated tablets are known to release their drug loading only when they reach the small intestine) (Manasa, Vanitha et al. 2016).

Controlled release - The concept of controlled release relates to the rate at which the drug is being freed from its carrier (Honey, Rijo et al. 2014). The use of controlled release polymer-based systems can improve the duration of action and effectiveness of a drug molecule by maintaining its systemic concentration within the therapeutic window over a prolonged time, and can also increase its stability (Liechty, Kryscio et al. 2010).

For example, the basis of injectable delivery systems lies in their providing controlled drug delivery by embedding the drug in a polymer matrix from which it is released by diffusion and/or polymer degradation. There are a range of such DDSs approved by the

FDA, for instance Lupron, Zoladex and Decapeptyl. These give release over one to four months and have been successfully used by thousands of patients (Moses, Brem et al.

2003). Another example are Solid Lipid Nanoparticles, often known as SLN, these are particulate systems with a size of 50-1000 nm and have been studied for controlled drug delivery systems and different administration routes (Muller, Mader et al. 2000).

1.4.2 Targeted drug delivery

The concept of drug targeting was conceived and first published by Ehrlich, who proposed a system termed the “magic bullet”: the system consists of drug carriers that provide exclusive delivery to preselected target cells in a specific manner (Strebhardt and Ullrich 2008). Targeted drug delivery systems aim to predominantly accumulate the

drug at or near its site of action, improving the local concentration without the risk of undesired delivery to other sites (Singh and Lillard 2009). Ideally, this would be achieved independently of the method and route of administration. Several key requirements have to be met for targeted drug delivery to be effective: retain, evade, target and release

(Mills and Needham 1999). For example, in intravenous administration this means efficient drug loading into a vehicle, sufficient time for the vehicle to be in circulation until it reaches its target, retention by intended target sites, and drug release at that site within a time that allows the drug to perform its function (Bae and Park 2011). Targeting can often be divided into passive and active targeting, and has been widely explored in the context of cancer treatment (Torchilin 2010).

Passive targeting is solely related to the circulation half-life of a formulation (the time taken for half to be removed by biological processes) (Bazak, Houri et al. 2014). An increase in circulation time and appropriate size of the drug carrier aids the formulation to accumulate in diseased tissue (Dreaden, Austin et al. 2012). The best known example of passive targeting is the enhanced permeability and retention (EPR) effect (Fang,

Nakamura et al. 2011), which can be very potent in the treatment of cancers. The EPR effect takes advantage of the physiological differences between the tumour and normal tissue (discussed in Section 1.1.1): as there is no drainage and no renewal of fluids in the tumour, there tends to be retention of formulations in the tumour site (Noguchi, Wu et al. 1998).

Active targeting is used to describe specific interactions between the drug carrier and the target cells, and usually exploit receptors that are up-regulated in the cells of interest

(Torchilin 2010). For example, the anticancer drug trastuzumab exploits the overexpression of HER2 receptors in breast cancer cells. (Shirshahi, Shamsipour et al.

2013). The combination of this type of targeting with other selective approaches, such as the use of prodrugs which are only activated in the target organ, could add another level of selectivity (Al-Jamal 2013). 44

It is interesting to notice that most formulations are passively targeted, most probably because passive targeting is easier to achieve: it arises with almost all drug carriers whether such distribution is intended or not (Bae and Park 2011). In contrast, active- targeted drug delivery needs a ligand-receptor interaction and is only effective when they are in close proximity (<0.5 nm) , which is often difficult to achieve (Bae and Park 2011).

1.4.3 Nanoscale vs macroscale DDS

Nanoscale DDS, refers to systems which are approximately in the 1 -1000 nm range

(e.g. polymeric nanoparticles, lipid-based DDS) (Fang, Nakamura et al. 2011),

Macroscales DDS (e.g. implants, inserts) allow the prolonged spatial release of encapsulated drug at the site where the DDS is located. They can overcome common challenges such as obviating the need for high doses in systemic administration, and frequent dosing (Kearney and Mooney, 2013).

When compared to nanoscale DDS, macroscale DDS are more suitable for disease treatment where promoting cellular uptake and systemic distribution are not desired.

There are several implantable macroscale drug delivery systems which have been approved by the FDA; some examples are: Zoladex® (goserilin acetate) to treat advanced prostate cancer, and Promus® (everolimus) for the narrowing of coronary arteries caused by coronary artery disease (Kumar and Pillai 2018).

The DDS properties (size, surface, charge, hydrophobicity, shape, etc.) will affect its performance and distribution in the body (Moghimi and Farhangrazi 2014). Each delivery system has advantages and limitations, but if well designed they can provide an approach to overcome the most common problems that conventional drug delivery presents (See Table 1-1)

Table 1-1. Methods by which DDS can overcome challenges in drug delivery. Adapted from Allen and Cullis (Allen and Cullis 2004).

Problem DDS approach

Solubility Precipitation of Drug solubility is enhanced by hydrophobic drugs in amphiphilic carriers aqueous media.

Tissue Extravasation of cytotoxic Controlled drug release from damage/Selectivity agent due to the lack of the DDS can reduce/eliminate selectivity for target leads tissue damage. DDS can to tissue damage. increase the local concentration in a desired tissue (i.e. EPR effect).

Half-life/ Activity loss of the drug DDS protect the drug from

Pharmacokinetics after administration. premature degradation and (PK) Drugs are cleared from the elimination, leading to system too rapidly sustained release. Altering the PK of the drug can help reduce the clearance

Distribution Distribution of drugs DDS such as lipid throughout the whole body nanoparticles achieve a better can affect normal tissues. biodistribution in target areas. (e.g. cardiac toxicity of doxorubicin) (Chatterjee, Zhang et al. 2010)

1.4.4 Polymer-based drug delivery systems

Polymer matrices prepared from biodegradable or biocompatible natural or synthetic polymers have been explored extensively in the development of drug delivery systems

(Pillai and Panchagnula 2001, Kwon and Furgeson 2007, Liechty, Kryscio et al. 2010).

Polymeric drug delivery systems often aim both to control drug delivery and also to protect the drug from enzymatic or other degradation pathways (Tiwari, Tiwari et al.

2012). For example, carriers can be made which are slowly degradable (e.g. poly(lactic- co-glycolic acid) [PLGA], degradation time in vivo 5-6 weeks, (Gentile, Chiono et al.

2014)), stimuli-reactive (e.g. poly(N-alkyl substituted acrylamides), (Zhang, Cao et al.

2017)) and targeted (e.g. PLGA nanoparticles functionalised with targeting peptides, ligands, or antibodies, (Feng, Yao et al. 2015)). For these reasons, polymers have been thoroughly investigated for biomedical and pharmaceutical applications (Liechty, Kryscio et al. 2010). Polymer selection is critical to the success of the formulation, with appropriate control of drug release (Semwal, Semwal et al. 2015). Polymer-based drug delivery systems can be prepared by various methods, including emulsification-solvent removal, interfacial polymerization, three-dimensional printing, vapour deposition, photolithography and electrohydrodynamic techniques (Naveed, Mora‐Huertas et al.).

1.4.5 Lipid based drug delivery systems

Lipid based drug delivery systems comprise carriers such as liposomes and solid lipid nanoparticles, among others. They are often used to address challenges like the low solubility and poor bioavailability of poorly water soluble drugs (Shrestha, Bala et al.

2014). They can encapsulate small molecules or proteins and can be used for the development of vaccines (Anwekar, Patel et al. 2011). These technologies have been very successful and many lipid-based delivery systems have received regulatory approval and reached the market. Some of the clinically used liposome-based products for cancer therapeutics include Doxil® (active ingredient: doxorubicin), which was the first approved cancer nanomedicine in 1995; DaunoXome® (daunorubicin); Depocyt®

(cytarabine); Myocet® (doxorubicin); Mepact® (mifamurtide); and Onivyde® (irinotecan)

(Ventola 2012). A more detailed introduction of lipid based drug delivery systems is given in Chapter 3.

1.5 Electrohydrodynamic techniques

Electrohydrodynamic (EHD) techniques exploit electricity to generate polymer-based products. There are two main EHD fabrication techniques: electrospraying and electrospinning. The two techniques are fairly similar, and differ only in the final product.

Electrospinning is a technique used to produce fibres with diameters often on the nanoscale, while electrospraying produces particles (Zamani, Prabhakaran et al. 2013).

EHD techniques have gained increasing interest in recent years since they are versatile and have the potential for applications in various fields such as tissue engineering, biosensors, filtration, wound dressings, and drug delivery (Sill and von Recum 2008,

Sridhar and Ramakrishna 2013). Numerous polymers are suitable to use in EHD, either alone or in blends; they can either be natural, synthetic or a mixture of both (Bhardwaj and Kundu 2010).

1.5.1 The electrospinning/electrospraying process

The equipment used for both consists of a syringe pump, a power supply, a collector

(generally covered with foil) and a spinneret (a narrow syringe needle, usually made of stainless steel), (Luana, Andrea et al. 2013); see (Figure 1-10). In EHD, a polymer solution, usually containing a functional component, is ejected through a spinneret at a constant flow rate (1-5 mL/h) while a high voltage (5-25kV) is applied causing rapid evaporation of the solvents.

Many authors have studied the physics and fluid dynamics of EHD processes (Taylor

1964, Lukáš, Sarkar et al. 2009). In both electrospinning and spraying, a polymer solution is expelled from a syringe at a precisely controlled rate, with the help of a syringe pump.

The polymer solution is simultaneous charged with the help of a voltage generator. The charges all have the same sign, leading to repulsion between them in the solution and the formation of a conical shape called the Taylor cone. The resultant Coulombic forces in the charged solution eventually overcome the surface tension forces, causing a jet to

be emitted from its tip. The electrical energy induces a rapid acceleration of the jet from the syringe to the collector leading to rapid evaporation of the solvent and non-volatile components, and in the case of electrospinning the solution solidifying to form fibres on a collector placed at the bottom of the equipment. The difference with electrospraying is that the jet breaks down into droplets, as can be seen in Figure 1-10 (Ciach 2007).

Syringe pump

Metal Spinneret

High voltage supply

Collector

Figure 1-10. An illustration of the common electrospraying (left)/electrospinning (right) experimental setups

Both processes use the same basic experimental setup, and the implementation of one or the other is achieved by changing the processing parameters: to switch from spinning to spraying the polymer concentration is reduced in order to reduce chain entanglement in the solution (see Figure 1-11). The charged droplets generated by the applied high voltage will repel each other, thus, generating a spray (Williams, Chatterton et al. 2012).

The collector is usually a plain metal sheet or a grid; however, in electrospinning, if a particular alignment of the fibres needs to be achieved, a rotating mandrel can be used

(Mitchell and Tojeira 2013). The alignment of the fibres can directly influence the

mechanical properties of the fibre mat (Matthews, Wnek et al. 2002), which can be very powerful in tissue engineering.

A number of operational parameters affect the products produced, and can be divided into three categories: solution, processing and ambient parameters.

1.5.2 Solution parameters

The solution parameters are crucial in the development of an electrospun/sprayed product. Important aspects to consider include concentration, molecular weight, viscosity, surface tension, and conductivity (Haider, Haider et al. 2015). Chain entanglements depend on each polymer and in the concentration of the chosen polymer in the solution. The overlap concentration (c*) is where the chain entanglements begin forming, before these there are no entanglemetns, this concentration is inversely proportional to the intrisitc viscosity (Bohr, P Boetker et al. 2013).(Figure 1-11)

Figure 1-11. Schematic representation of polymer entanglements

Changing the viscosity affects the fibre morphology, with fibres showing a smooth morphology in the optimal range. If the concentration is too high, helix shaped fibres will be observed (Li and Wang 2013). On the other hand, fluids with a lower viscosity tend to

produce beaded nanofibres or particles, due to the low chain entanglement (Figure 1-11).

The optimum viscosity is polymer dependent: for example, uniform fibres were achieved at 40 wt% of 29 KDa poly(vinylpyrrolidone) (PVP viscosity 153.4 cP) in ethanol:deionized water, whereas particles were fabricated at 20% wt% (viscosity 18.5 cP) (Munir,

Suryamas et al. 2009). In contrast, viscosities of 800-4000 cP are needed for the creation of polyethylene oxide (PEO) fibres (Doshi and Reneker 1995).

1.5.3 Processing parameters

The key processing parameters include the voltage, flow rate, and tip to collector distance (Nurwaha, Han et al. 2013). The appropriate processing parameters are typically dependent on the solution parameters. After the solution for spinning/spraying is chosen then the voltage, flow rate, needle diameter and tip-to-collector distance can be optimized to create uniform fibres or monodisperse particles (Haider, Haider et al.

2015).

The applied voltage influences Taylor cone formation and controls the diameter of the produced fibres or particles. There is a critical voltage, which depends on the polymer in use, at which Taylor cone formation and fibres/particles production are achieved (Li and

Wang 2013). Increased voltages lead up to a certain point to a reduction in the diameter of the product, but exceeding this threshold voltage results in the formation of beaded fibres and elongated particles (Haider, Haider et al. 2015). Higher voltages can also lead to incomplete solvent evaporation and the production of merged fibres (Henriques,

Vidinha et al. 2009).

The flow rate determines the fibre/particle morphology and also influences solvent evaporation. A higher flow rate will result in thicker fibres or bigger particles, along with the formation of bead-on-string fibre or ribbon-like fibres (Garg and Bowlin 2011). As with the voltage, there is a critical flow rate where there is sufficient continuous replacement of the polymer solution at the needle tip to replenish the Taylor cone and avoid spinneret

blockage, but the flow is low enough to prevent defects forming in the products (Megelski,

Stephens et al. 2002).

The solvent selected should dissolve all the components in the feeding solution. Also, a certain volatility is required to achieve complete evaporation during the process. Solvent evaporation occurs when the polymer solution travels from the needle tip to the collector

(Leach, Feng et al. 2011). The alteration of this distance influences the time available for the solvent to evaporate. Short distances result in defective fibre/particles, whereas increasing the distance reduces the product diameter, but also has the potential of producing fibres with beads on them. In addition, longer distances lead to the accumulation of material outside the collector, resulting in lower yields of the fabricated material (Henriques, Vidinha et al. 2009).

The inner diameter of the needle should be taken into consideration when developing a process: a wider-bore needle results in increased product diameter, but a higher voltage must be applied for smaller needle diameters in order to overcome the surface tension

(Macossay, Marruffo et al. 2007).

1.5.4 Environmental parameters

Humidity and temperature also have an impact on the EHD technique. Product diameter has been shown to decrease as the relative humidity is increased up to a threshold value; if this value is exceeded, beaded fibres are created (De Vrieze, Van Camp et al. 2009).

Increasing the temperature decreases product diameter as the viscosity of the polymer solution decreases, and there is also an increased evaporation rate. Highly volatile solvents face the issue of spinneret occlusion at high rates of solvent evaporation, however (Bhardwaj and Kundu 2010).

A schematic illustration of the influence of the EHD parameters on the products produced is shown in Figure 1-12.

Figure 1-12. Diagram depicting the influence of electrospinning/electrospraying parameters on the final product.

1.5.5 EHD techniques and drug delivery

For drug delivery systems, materials produced by EHD have great potential. The fabrication of particles or fibres appropriate for drug delivery is possible since the loading of therapeutic molecules in polymer carriers can be achieved relatively easily (Figure

1-13) (Chakraborty, Liao et al. 2009).

Figure 1-13 A schematic diagram of an electrospun/sprayed drug loaded fibre/particle.

EHD offers multiple advantages over other techniques for making polymer/drug composites, involving no use of elevated temperatures, no separate drying step, and

often yielding products with high mono-dispersity through simple and accurate control of the processing parameters (Ramakrishna, Fujihara et al. 2006). EHD is a one-step process, fast, relatively cheap, and provides the opportunity for scale up to bulk production (Sridhar and Ramakrishna 2013).

EHD techniques can also help to control the physical properties of the API, for instance changing the crystalline form of a drug into its amorphous state (Ribeiro, Sencadas et al.

2011). This is useful as therapeutic compounds are most stable when they are in a crystalline form, but a crystalline compound often has poor aqueous solubility and low dissolution rates. The production of a stable amorphous form of the drug has great potential for solubility enhancement, thus improving drug dissolution and bioavailability

(Williams, Chatterton et al. 2012).

The incorporation of drugs into nanofibres via electrospinning was first reported in 2002.

Since then the incorporation of a wide range of drugs has been explored because of the fact that electrospun nanofibres allow ease of drug incorporation (including of fragile biomolecules), have a large surface area to volume ratio, and highly porous and interconnected architectures resembling the extracellular matrix. They are also highly tunable materials as the polymer, fibre diameter, morphology and porosity can be changed. Due to the versatility of the EHD techniques, a wide range of solvents and polymers can be used to entrap diverse pharmaceutical ingredients such as antibiotics, anticancer drugs, proteins, DNA, RNA, growth factors, and antibodies, among others.

Some examples from the recent literature are given in Table 1-2

Drugs Polymer Reference

Antibiotics 3% w/v PLA (Inherent viscosity = 5.7- Tetracycline 6.5 dL/g) + 12 % w/v PCL (70-80 (Zahedi, Karami et al. 2012) hydrochloride kDa) in Chloroform/ dimethylformamide (DMF) (7:3 v/v)

PLGA (LA/GA=50:50) Mn=49,100 Da Fusidic acid in Tetrahydrofuran (THF)/ DMF (75:25 (Gilchrist, Lange et al. 2013) and rifampicin v/v)

Anticancer 6 wt% PLLA (Mn=13 800 Da) drugs (More Doxorubicin in chloroform/Methanol/DMSO (Liu, Zhou et al. 2013) examples in (80/14/6 v/v/v) Table 1-3) 13% w/v Eudragit (ES-100, 125 kDa) 5-fluorouracil (Illangakoon, Yu et al. 2015) in Ethanol/DMF (8:2 v/v)

Antibodies 10% PCL (Mw=80 kDa) in (Angkawinitwong, Awwad et al. Bevacizumab Trifluoroethanol (TFE): deionized 2017) water (90:10 v/v)

CD-34 20 wt% PU (Mw not specified) in Monoclonal Dimethylcacetamide (DMAc) : (Joung, Hwang et al. 2010) antibody Butanone (1:2 v/v)

Proteins (25-100 wt%) PCL (Mn=80 000) (Chakrapani, Gnanamani et al. Collagen in Hexafluoroisopropanol (HFIP) 2012) /Acetic acid

15 wt% PEO (4x105 Da) in Deionised (Kowalczyk, Nowicka et al. BSA water 2008)

DNA 35% w/v PLA–PEG–PLA triblock DNA (pCMVβ Copolymer/PLGA (LA/GA=75:25) (Luu, Kim et al. 2003) plasmid) in DMF

RNA 14 wt% PCL (65 kDa) SiRNA (Cao, Jiang et al. 2010) in RNase-free water

Growth Human β- 12 wt% PCL-co-PEEP with a 15 factors nerve growth molar percent of EEP (70 760 Da) in (Chew, Wen et al. 2005) factor (NGF) DCM

1.5.5 EHD for cancer therapeutics

All the properties mentioned above can be advantageous for the delivery of cancer therapeutics; for example, by using a local delivery the minimum required dose decreases, reducing unwanted side effects. In vivo studies showed that when using electrospun fibres fabricated with PLLA and encapsulating 5-FU and oxaliplatin for the treatment of colon cancer, a larger necrotic region was seen in tumours when compared with the free drugs (Zhang, Wang et al. 2016). Another study demonstrated that when

PLGA was combined with paclitaxel to create nanofibres, smaller tumours were observed in animals treated by these nanofibers when compared with the commercial formulation of paclitaxel (Taxol) (Xie and Wang 2006). Hydroxylcamptothecin-loaded nanofibres reported by Luo had a 20-fold higher inhibitory effect than free hydroxylcamptothecin after 72 hours in vitro (Luo, Xie et al. 2012).

Several other studies using different combinations of polymer and drugs have been prepared for anti-cancer applications; some examples can be seen in Table 1-3.

Electrospraying has also been investigated to create micro- and nanoparticles for anticancer applications (Sridhar and Ramakrishna 2013). Polymer-based micro and nanoparticles have a number of advantages and can be administered as oral, parenteral, inhalation, topical and local drug delivery systems (Pridgen, Alexis et al. 2015). However, electrosprayed formulations have attracted significantly less attention than the products of electrospraying.

Table 1-3. Chemotherapeutic drugs entrapped by EHD techniques. *Mw not stated

Chemotherapeutic Targeted Polymer solution Reference Drug cancer

Cis-platin 18% w/v PCL (45 kDa) in Cervical cancer (Aggarwal, DCM/DMF (6:4) and 1% Goyal et al. chitosan (310 kDa) in TFA 2017) Doxorrubicin 0-30 w/w% PLA (75 kDa) in Not stated (Doustgani Chloroform 2017) Paclitaxel 2-30% w/v PLGA Glioma (Xie and Wang (LA/GA=50:50) in DCM 2006) 5-flourouracil 13% w/v Eudragit (125 kDa) in Not stated (Illangakoon, Yu Ethanol/DMF (8:2 v/v) et al. 2015) Daunorubicin 10% wt of Poly(N- Leukemia (Song, Guo et isopropylacrylamide)-co- al. 2008) polysterene in DMF* Doxorubicin and 16% w/v PLGA (LA/GA=50:50, Liver (Wei, Hu et al. camptothecin 10 kDa) Gelatin in HFP 2014) Paclitaxel and 6 wt% PEG‐PLA in Chloroform Glioma (Xu, Chen et al. doxorubicin 2009) 5‐Aminolevulinic acid 13.5% w/v PVA (13-23 kDa) in Cholangiocarcin (Yoo, Kim et al. water oma 2012) Hydroxycamptothecin 20% of PELA containing 10% Hepatomacarcin (Xie, Li et al. of PEG (PEG Mw= 6 kDa) oma 2010) PELA-PEG (Mw= 46.4 kDa) in DCM Titanocene dichloride 10% w/v PLLA (100 kDa) in Lung cancer (Chen, Wu et al. DCM 2010) Temozolomide 5-100% wt/wt PLGA (50:50, Glioma (Ramachandran Mw=45 kDa, or 75:25 , Junnuthula et Mw=75kDa, or 85:15, Mw=75 al. 2017) kDa) / 5-50% w/w PLA (Mw= 150 kDa) / 1-10% wt/wt PCL (Mw=100 kDa) in acetone (PLGA) or chloroform (PLA and PCL) 1,3‐Bis(2‐chloroethyl)‐ 12 wt% PLGA (LA:GA=80:20, Glioma (Xu, Chen et al. 1‐nitrosourea Mn=60 800) and 7.5 wt% 2006) PEG‐PLLA (Mn=84 800) in chloroform Curcumin 40% w/v PLGA (LA:GA=80:20, Skin (Sampath, Lakra 70:30 or 60:40) in chloroform et al. 2014)

Gulfam et al reported electrosprayed nanoparticles with entrapped cyclophosphamide, and found the nanoparticles enabled the release of the drug in a controlled manner

(Gulfam, Kim et al. 2012). In other work, doxorubicin hydrochloride was electrosprayed

together with chitosan, achieving a 68% entrapment efficiency and presenting controlled release over a period of 72h. (Songsurang, Praphairaksit et al. 2011). Additionally, Xie and Wang studied the fabrication of electrospun paclitaxel-loaded PLGA biodegradable fibres, and an encapsulation efficiency of more than 90% and release for more than 60 days was achieved (Xie and Wang 2006). A controlled drug release matrix composed of electrosprayed particles from PCL and paclitaxel was successful in maintaining a sustained release profile over a one month period (Ding, Lee et al. 2005). Table 1-4 shows the cancer therapeutics encapsulated via electrospraying.

Table 1-4. Electrosprayed particles encapsulating cancer therapeutics

Targeted Cancer therapeutic Polymer solution Reference cancer Cyclophosphamide 7% w/v Gliadin in 70% Breast (Gulfam, Kim et al. ethanol and gelatin in 90% 2012) acetic acid Paclitaxel 5% w/v PCL (65 kDa) in Not stated (Ding, Lee et al. DCM 2005) Paclitaxel 5% w/v PLA (100 kDa) in Not stated (Ciach 2006) DCM Paclitaxel 2-16% w/v PLGA (50:50) in Not stated (Xie, Lim et al. ACN 2006) Paclitaxel 2-8% w/v PCL (14 kDa) in Glioma (Xie, Marijnissen et DCM, ACN or THF al. 2006) 6-8 % PLGA (50:50, 90-120 kDa) in DCM, ACN or THF Paclitaxel/Etanidazol 2% w/v ((90%-PLGA Glioma (Kumar e (65:35, 40-75 kDa) and Naraharisetti, Yung 10% PEG (3350 Da)) in Sheng Ong et al. DCM 2007) Tamoxifen 5% w/v Stearic acid in Breast (Trotta, Cavalli et presence of EC (4.5:0.5 al. 2010) w/w) in ethanol, n-propanol, n-butanol, n-pentanol and n-hexanol Paclitaxel, 4:2 – 5:3 wt% PLGA Not stated (Kim and Kim Doxorrubicin (50:50, 20 kDa) :PDLLA 2011) (75 kDa) in TFE Paclitaxel, suramin PLGA/PLLA Brain (Nie, Fu et al. 2010)

Artesunat Core: 1-1.5% w/v PLGA Not stated (Ho, Laidmäe et al. (50:50, 7-15 kDa) in ACN 2017) with or without acetone Shell: 0.37% w/v chitosan (50-190 kDa) in water: acetic acid: EtOH (12.5:37.5:50 v/v/v) Resveratrol A weight ratio of 9:9 of CD44 (Lee, Lee et al. Hyaluronic acid-ceramide receptor 2016) and Soluplus in methanol expressed cancers

In addition, EHD techniques offer the opportunity to load multiple drugs in a single formulation. For example, hydrophilic doxorubicin hydrochloride and lipophilic paclitaxel were successfully entrapped in PEG-PLA nanofibres, creating a multi-drug delivery fibre mat which showed a higher cytotoxicity in the C6 cell line (brain glioma) than a single drug-loaded system (Xu, Chen et al. 2009). The co-encapsulation of daunorubicin and

Fe3O4 nanoparticles in nanofibres was also explored for the treatment of leukemia (Lv,

He et al. 2008). These novel composites of polylactide (PLA) showed enhanced accumulation of daunorubicin on the membranes of cancer cells. The nanofibres facilitated interactions between daunorubicin and leukemia cancer cells, leading to cell death of the K562 (leukemia) cell line. The results demonstrate the magnetic nanoparticles enhance the local drug concentration in the target tumor cells and hence the efficiency of the cancer treatments

The most common type of electrospinning explored by researchers is the single fluid technique, however these fibres face a great disadvantage as they present an initial burst release when they come into contact with a physiological medium.

There are more complex electrohydrodynamic techniques, such as side-by-side, coaxial and triaxial electrospinning. Side-by-side electrospinning/spraying consists in a spinneret with an arrangement of two needles next to each other, it provides an alternative to create side-by-side structures, also called Janus, that can be used to load two different active

ingredients in one formulation (Roh, Martin et al. 2005); coaxial electrospinning can be used as fabrication method to create core-sheath fibres (Sun, Zussman et al. 2003); meanwhile, more complex electrospinning such as triaxial electrospinning, which prevent the burst release of drug, have been explored by different authors (Han and Steckl 2013).

These different types of EHD techniques are widely discussed in Chapter 4.

As seen in Table 1-2, Table 1-3 and Table 1-4, organic solvents are normally used for electrospinning and electrospraying of polymer materials. Any residual organic solvent in the fabricated materials could potentially be toxic to cells; thus, it is extremely important to ensure that any residual solvent inside the matrices is within the limits of exposure regulated by international health authorities (e.g. the International Program on Chemical

Safety, World Health Organisation, Food and Drug Administration) (Xie et al, 2008)

Even though, at the moment, there are no commercialized products fabricated with EHD techniques, there are multiple publications from numerous research groups regarding the scaling up of these processes. Several approaches including the use of multi-needle arrays for high-throughput electrospinning and electrospraying, corona electrospinning,

Nanospider among others, have been described. These scale up processes are further discussed in more detail in Chapter 3.

1.6 Spray drying

Spray drying is a well-established technique for producing a dry powder from a liquid phase (Dyvelkov and Sloth 2014). It involves removing the liquid phase of a solution, suspension or emulsion via evaporation. This process is widely used in many industries, including the pharmaceutical and food industries, and there are several products in the market which are fabricated using this technique (Woo and Bhandari 2013), (Jain,

Ganesh et al. 2012). The process starts with the feed preparation. One of the advantages of the spray drying process is the ability to use slurry range of precursors

(Ho, Truong et al. 2017). The only requirement is that the precursor fluid must be able to flow through the pump and should be free of impurities.

The spray drying apparatus is illustrated in Figure 1-14. The most critical step of the process is the atomization, which involves separating the feed fluid into fine particles.

The feed material enters the dryer vertically through a nozzle with hot air/gas entering the dryer chamber at the same location, at a temperature of 150-300oC. The high velocity of the air instantly atomizes the liquid. This augments the surface to volume ratio allowing spray drying to achieve a fast drying rate (Cal and Sollohub 2009).

In the second stage the droplets meet the heated drying medium, resulting in solvent evaporation beginning immediately. A constant drying rate is maintained as the solvent diffuses from within the droplet to maintain saturated surface conditions. When the solvent contents become too low, saturated surface conditions cannot be achieved, resulting in the formation of a dry layer on the surface of each droplet (Gharsallaoui,

Roudaut et al. 2007). Lastly, the powder is separated from the carrier gas by centrifugal force and collected in a collecting vessel which is attached at the bottom of the cyclone

(Figure 1-14) (Gharsallaoui, Roudaut et al. 2007).

Figure 1-14. Schematic representation of the spray drying instrument

1.6.1 Spray drying in the pharmaceutical industry

Spray drying is often used to generate formulations of APIs, with systems containing etravirine (HIV treatment), ivacaftor (cystic fibrosis), tracrolimus (immunosuppressive), itraconzole (an antifungal), and everolimus (cancer therapeutic), among many others, having been reported (Newman 2015). In addition, laboratory research into spray dried formulations of different types of drugs and proteins has been extensive throughout the years (Mumenthaler, Hsu et al. 1994, Maury, Murphy et al. 2005, Mönckedieck,

Kamplade et al. 2016, Halliwell, Bhardwaj et al. 2017). These studies have found that, for instance, spray-drying an aqueous solution of pure protein produces aggregation and loss of activity, but this can be ameliorated by formulation measures (Maury, Murphy et al. 2005).

Spray drying can be used for a range of purposes including modification of biopharmaceutical properties, or generating formulations of dried emulsions, nanoparticle-loaded microspheres, or biodegradable microspheres, for instance (Ré

2006, Sollohub and Cal 2010). It allows the size distribution, crystallinity, polymorphism and moisture content of a formulation to be tuned (Baaklini 2015). These can have significant consequences in the bioavailability and stability of the fabricated materials

(Babu and Nangia 2011). As with EHD techniques, spray drying allows the API to be encapsulated in the amorphous form, thus increasing the bioavailability (Huntington

2004).

Chemotherapeutic agents have been explored in several studies using spray drying. For example, Santos developed a delivery matrix which consisted of a spray-dried formulation of 5-FU and hydroxyapatite. The material produced had a very fast release rate of 5-FU in phosphate buffer (Santos, Rovath et al. 2009). Meenach et al. studied the delivery of PEGylated phospholipid particles loaded with paclitaxel for the treatment of lung cancer using a spray-dried formulation, and found high encapsulation efficiencies

(43-99%) and an excellent aerosol dispersion performance (Meenach, Anderson et al. 62

2013). Encapsulating doxorubicin in PLGA microparticles through spray-drying provides the opportunity to overcome one of the restrictions of this drug, since its therapeutic potential is restricted by dose-limited cardiotoxicity (Chatterjee, Zhang et al. 2010). The cytotoxicity of doxorubicin-loaded PLGA microspheres prepared by spray drying was investigated in Glioma C6 cancer cells, and it was found that the cytotoxicity of DOX is enhanced through use of the PLGA microparticles (Lin, Shi Ng et al. 2005). A comparison between spray drying and EHD can be seen in Table 1-5.

Spray drying and EHD are in many ways complementary but have different advantages and disadvantages; both processes will be investigated in this thesis.

Table 1-5. Comparison between spray drying and EHD. Adapted from Sosnik and Serementa (Sosnik and Seremeta 2015)

Properties EHD Spray drying

Cost-effective High High, but required large first capital investment

Scalability Moderate Easy

Use of heat No heat Use of heat, thus involved degradation of thermosensitive materials

Adaptable to product Yes Yes specifications

Control over physicochemical Yes Yes parameters

One step process Yes Yes Yields High High

Steps Single-step Single-step

Continuity Continuous Continuous

Speed Medium-Low High Versatility High High

Final drying step Not required Not required

Effective encapsulation of drugs Yes Yes within polymer carriers

1.7 Thesis aims and outline

While most chemotherapy treatments target the cell cycle, PDT works by inducing an acute stress response which leads to mitochondrial damage, cythochrome c release and formation of an apotosome (Castano, Demidova et al. 2005). It has been proven that a combinational therapy of photodynamic and chemotherapy results in significant inhibition of tumour proliferation, increased induction of apoptosis, damage of tumour vasculature, strong antitumour effects and improved animal survival (Zhou, Zhou et al. 2014).

A major part of this project seeks to develop a multi drug co-delivery system for combined photo-chemotherapy, as these therapies have proved to be efficient as cancer treatments. Specifically, polymer particles co-loaded with potent anticancer drugs

(carmofur or 5-FU) and a photosensitizer (rose bengal or haematoporphyrin) will be investigated and the EHD processing parameters systematically varied to develop optimum systems. These will be characterized in detail in terms of their physicochemical properties, and then tested against model cell lines (human dermofibroblasts and different types of cancer cell lines) in vitro. By using drug delivery systems, the therapeutic index could be increased, for instance through enhanced stability of the components, more precisely targeted delivery or increasing the uptake by prolonging the retention time and enhancing dissolution (Tahover, Patil et al. 2015). Thus, encapsulating chemotherapeutic agents into polymer carriers can reduce the adverse effects caused. Specific aims and objectives are outlined further in each chapter.

The overall aims of this thesis are:

1. To explore the use of electro-hydrodynamic techniques to create a

photochemotherapy formulation via the use of Janus nanoparticles and fibres

2. To use EHD and spray-dried materials as precursors for the assembly of solid

lipid nanoparticles encapsulating a model drug and a chemotherapeutic agent

(indomethacin, 5-FU).

3. To use of triaxial electrospinning to create a photo-chemotherapy formulation. 64

1.8 References

Aggarwal, U., A. K. Goyal and G. Rath (2017). "Development and characterization of the cisplatin loaded nanofibers for the treatment of cervical cancer." Mater Sci Eng C Mater Biol Appl 75: 125-132. Al-Jamal, K. T. (2013). "Active drug targeting: Lessons learned and new things to consider." International Journal of Pharmaceutics 454(1): 525-526. Aldred, E. M., C. Buck and K. Vall (2009). Chapter 39 - Chemotherapy. Pharmacology. Edinburgh, Churchill Livingstone: 307-313. Alfarouk, K. O., C.-M. Stock, S. Taylor, M. Walsh, A. K. Muddathir, D. Verduzco, A. H. H. Bashir, O. Y. Mohammed, G. O. Elhassan, S. Harguindey, S. J. Reshkin, M. E. Ibrahim and C. Rauch (2015). "Resistance to cancer chemotherapy: failure in drug response from ADME to P-gp." Cancer Cell International 15: 71. Allen, T. M. and P. R. Cullis (2004). "Drug Delivery Systems: Entering the Mainstream." Science 303(5665): 1818-1822. André, T., C. Boni, L. Mounedji-Boudiaf, M. Navarro, J. Tabernero, T. Hickish, C. Topham, M. Zaninelli, P. Clingan, J. Bridgewater, I. Tabah-Fisch and A. de Gramont (2004). "Oxaliplatin, Fluorouracil, and Leucovorin as Adjuvant Treatment for Colon Cancer." New England Journal of Medicine 350(23): 2343-2351. Andre, T., C. Boni, M. Navarro, J. Tabernero, T. Hickish, C. Topham, A. Bonetti, P. Clingan, J. Bridgewater, F. Rivera and A. de Gramont (2009). "Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial." J Clin Oncol 27(19): 3109-3116. Angkawinitwong, U., S. Awwad, P. T. Khaw, S. Brocchini and G. R. Williams (2017). "Electrospun formulations of bevacizumab for sustained release in the eye." Acta Biomater 64: 126-136. Anwekar, H., S. Patel and A. Singhai (2011). "Liposome-as drug carriers." International journal of pharmacy & life sciences 2(7). Baaklini, G. (2015). The impacts of Spray Drying on the polymorphism of pharmaceutical and organic compounds, Université de Rouen. Babu, N. J. and A. Nangia (2011). "Solubility Advantage of Amorphous Drugs and Pharmaceutical Cocrystals." Crystal Growth & Design 11(7): 2662-2679. Bae, Y. H. and K. Park (2011). "Targeted drug delivery to tumors: Myths, reality and possibility." Journal of Controlled Release 153(3): 198-205. Balducci, L. (2007). CHAPTER 45 - Cancer A2 - Ham, Richard J. Primary Care Geriatrics (Fifth Edition). P. D. Sloane, G. A. Warshaw, M. A. Bernard and E. Flaherty. Philadelphia, Mosby: 533-545. Bamrungsap, S., Z. Zhao, T. Chen, L. Wang, C. Li, T. Fu and W. Tan (2012). "Nanotechnology in therapeutics: a focus on nanoparticles as a drug delivery system." Nanomedicine 7(8): 1253-1271. Banerjee, D. and S. Sengupta (2011). "Nanoparticles in cancer chemotherapy." Prog Mol Biol Transl Sci 104: 489-507. Bazak, R., M. Houri, S. E. Achy, W. Hussein and T. Refaat (2014). "Passive targeting of nanoparticles to cancer: A comprehensive review of the literature." Molecular and Clinical Oncology 2(6): 904-908.

Berg, K., L. Prasmickaite, P. K. Selbo, M. Hellum, A. Bonsted and A. Høgset (2003). "Photochemical internalization (PCI)--a novel technology for release of macromolecules from endocytic vesicles." Oftalmologia (Bucharest, Romania : 1990) 56(1): 67-71. Berg, K., P. K. Selbo, A. Weyergang, A. Dietze, L. Prasmickaite, A. Bonsted, B. O. Engesaeter, E. Angell-Petersen, T. Warloe, N. Frandsen and A. Hogset (2005). "Porphyrin-related photosensitizers for cancer imaging and therapeutic applications." J Microsc 218(Pt 2): 133-147. Berg, K., A. Weyergang, L. Prasmickaite, A. Bonsted, A. Hogset, M. T. Strand, E. Wagner and P. K. Selbo (2010). "Photochemical internalization (PCI): a technology for drug delivery." Methods Mol Biol 635: 133-145. Bergers, G. and L. E. Benjamin (2003). "Tumorigenesis and the angiogenic switch." Nat Rev Cancer 3(6): 401-410. Bhardwaj, N. and S. C. Kundu (2010). "Electrospinning: A fascinating fiber fabrication technique." Biotechnology Advances 28(3): 325-347. Biteghe, F. A. N. and L. M. Davids (2017). "A combination of photodynamic therapy and chemotherapy displays a differential cytotoxic effect on human metastatic melanoma cells." Journal of Photochemistry and Photobiology B: Biology 166: 18-27. Boesen, E. and W. Davis (1969). Cytotoxic drugs in the treatment of cancer. London, Hodder & Stoughton Educational, 1st Edition. Bohr, A., J. P Boetker, T. Rades, J. Rantanen and M. Yang (2013). Application of Spray- Drying and Electrospraying/Electospinning for Poorly Water-Soluble Drugs: A Particle Engineering Approach.Curr Pharm Des 20(3):325-348 Boland, C. R. and L. Ricciardiello (1999). "How many mutations does it take to make a tumor?" Proceedings of the National Academy of Sciences of the United States of America 96(26): 14675-14677. Brown, J. M. and A. J. Giaccia (1998). "The Unique Physiology of Solid Tumors: Opportunities (and Problems) for Cancer Therapy." Cancer Research 58(7): 1408-1416. BUPA. (2011). "Cancer Diagnosis and Treatment: A 2021 Projecion." from http://www.bupa.com/media/354684/cancer_diagnosis_and_treatment_- _a_2021_projection_-_final.pdf. Burrell, R. A. and C. Swanton (2014). "Tumour heterogeneity and the evolution of polyclonal drug resistance." Molecular Oncology 8(6): 1095-1111. Cal, K. and K. Sollohub (2009). "Spray Drying Technique. I: Hardware and Process Parameters." Journal of Pharmaceutical Sciences 99(2): 575-586. Cancer Research UK. (2017). "Cancer incidence by stage at diagnosis." Retrieved 12/09/2018, from https://www.cancerresearchuk.org/health-professional/cancer- statistics/incidence/all-cancers-combined#heading-Two. Cancer Research UK. (2017). "Cancer survival statistics." Retrieved 12/09/2018, from https://www.cancerresearchuk.org/health-professional/cancer-statistics/survival. Canti, G., A. Nicolin, R. Cubeddu, P. Taroni, G. Bandieramonte and G. Valentini (1998). "Antitumor efficacy of the combination of photodynamic therapy and chemotherapy in murine tumors." Cancer Letters 125(1–2): 39-44. Cao, H., X. Jiang, C. Chai and S. Y. Chew (2010). "RNA interference by nanofiber-based siRNA delivery system." J Control Release 144(2): 203-212. Carlin, L. J. (2011). "Mutations Are the Raw Materials of Evolution." Nature Education Knowledge 3 (10): 10.

Castano, A. P., T. N. Demidova and M. R. Hamblin (2005). "Mechanisms in photodynamic therapy: part two—cellular signaling, cell metabolism and modes of cell death." Photodiagnosis and photodynamic therapy 2(1): 1-23. Chabner, B. A. and D. L. Longo (2011). Cancer Chemotherapy and Biotherapy: Principles and Practice, Wolters Kluwer Health. Chabner, B. A. and T. G. Roberts (2005). "Chemotherapy and the war on cancer." Nat Rev Cancer 5(1): 65-72. Chakraborty, S., I. C. Liao, A. Adler and K. W. Leong (2009). "Electrohydrodynamics: A facile technique to fabricate drug delivery systems." Advanced drug delivery reviews 61(12): 1043-1054. Chakrapani, V. Y., A. Gnanamani, V. R. Giridev, M. Madhusoothanan and G. Sekaran (2012). "Electrospinning of type I collagen and PCL nanofibers using acetic acid." Journal of Applied Polymer Science 125(4): 3221-3227. Chatterjee, K., J. Zhang, N. Honbo and J. S. Karliner (2010). "Doxorubicin Cardiomyopathy." Cardiology 115(2): 155-162. Chen, J., B. F. Miller and A. V. Furano (2014). "Repair of naturally occurring mismatches can induce mutations in flanking DNA." eLife 3: e02001. Chen, P., Q.-S. Wu, Y.-P. Ding, M. Chu, Z.-M. Huang and W. Hu (2010). "A controlled release system of titanocene dichloride by electrospun fiber and its antitumor activity in vitro." European Journal of Pharmaceutics and Biopharmaceutics 76(3): 413-420. Chew, S. Y., J. Wen, E. K. F. Yim and K. W. Leong (2005). "Sustained Release of Proteins from Electrospun Biodegradable Fibers." Biomacromolecules 6(4): 2017-2024. Ciach, T. (2006). "Microencapsulation of drugs by electro-hydro-dynamic atomization." International Journal of Pharmaceutics 324(1): 51-55. Ciach, T. (2007). "Application of electro-hydro-dynamic atomization in drug delivery." Journal of Drug Delivery Science and Technology 17(6): 367-375. Colvin, O. M. (2002). Alkylating Agents A2 - Bertino, Joseph R. Encyclopedia of Cancer (Second Edition). New York, Academic Press: 35-42. Cooper, G. (2000). The Cell: A Molecular Approach. The Eukaryotic Cell Cycle. S. Associates, Sunderland Dalla Via, L. and S. Marciani Magno (2001). "Photochemotherapy in the treatment of cancer." Curr Med Chem 8(12): 1405-1418. De Souza, R., P. Zahedi, C. J. Allen and M. Piquette-Miller (2010). "Polymeric drug delivery systems for localized cancer chemotherapy." Drug Delivery 17(6): 365-375. De Vrieze, S., T. Van Camp, A. Nelvig, B. Hagström, P. Westbroek and K. De Clerck (2009). "The effect of temperature and humidity on electrospinning." Journal of materials science 44(5): 1357. Devalapally, H., A. Chakilam and M. M. Amiji (2007). "Role of nanotechnology in pharmaceutical product development." J Pharm Sci 96(10): 2547-2565. DeVita, V. T., T. S. Lawrence and S. A. Rosenberg (2010). Cancer: Principles and Practice of Oncology-Advances in Oncology, Wolters Kluwer/Lippincott Williams & Wilkins Health. Ding, L., T. Lee and C.-H. Wang (2005). "Fabrication of monodispersed Taxol-loaded particles using electrohydrodynamic atomization." Journal of Controlled Release 102(2): 395-413. Dolgin, E. (2018). "Bringing down the cost of cancer treatment." 67

Dolmans, D. E. J. G. J., D. Fukumura and R. K. Jain (2003). "Photodynamic therapy for cancer." Nat Rev Cancer 3(5): 380-387. Domracheva, I., R. Muhamadejev, M. Petrova, E. Liepinsh, A. Gulbe, I. Shestakova, G. Duburs and P. Arsenyan (2015). "1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC) increases Carmofur stability and in vitro antiproliferative effect." Toxicology Reports 2: 377-383. Doshi, J. and D. H. Reneker (1995). "Electrospinning process and applications of electrospun fibers." Journal of Electrostatics 35(2): 151-160. Doustgani, A. (2017). "Doxorubicin release from optimized electrospun polylactic acid nanofibers." Journal of Industrial Textiles 47(1): 71-88. Dreaden, E. C., L. A. Austin, M. A. Mackey and M. A. El-Sayed (2012). "Size matters: gold nanoparticles in targeted cancer drug delivery." Therapeutic delivery 3(4): 457-478. Duffner, P. K. (2006). "The long term effects of chemotherapy on the central nervous system." Journal of Biology 5(7): 21-21. Dyvelkov, K. N. and J. Sloth (2014). Chapter 6 - New Advances in Spray-Drying Processes. Microencapsulation in the Food Industry. A. G. Gaonkar, N. Vasisht, A. R. Khare and R. Sobel. San Diego, Academic Press: 57-63. Eder, J. P. and A. T. Skarin (2010). 21 - Systemic and Mucocutaneous Reactions to Chemotherapy. Atlas of Diagnostic Oncology (Fourth Edition). Philadelphia, Content Repository Only!: 721-736. Egner, J. R. (2010). "AJCC cancer staging manual." JAMA 304(15): 1726-1727. Eldridge, S. and M. Davis (2018). 3.14 - Antineoplastic Agents☆ A2 - McQueen, Charlene A. Comprehensive Toxicology (Third Edition). Oxford, Elsevier: 219-232. Fang, J., H. Nakamura and H. Maeda (2011). "The EPR effect: Unique features of tumor blood vessels for drug delivery, factors involved, and limitations and augmentation of the effect." Advanced Drug Delivery Reviews 63(3): 136-151. Feng, X., J. Yao, X. Gao, Y. Jing, T. Kang, D. Jiang, T. Jiang, J. Feng, Q. Zhu, X. Jiang and J. Chen (2015). "Multi-targeting Peptide-Functionalized Nanoparticles Recognized Vasculogenic Mimicry, Tumor Neovasculature, and Glioma Cells for Enhanced Anti- glioma Therapy." ACS Applied Materials & Interfaces 7(50): 27885-27899. Garg, K. and G. L. Bowlin (2011). "Electrospinning jets and nanofibrous structures." Biomicrofluidics 5(1): 013403. Gentile, P., V. Chiono, I. Carmagnola and P. V. Hatton (2014). "An Overview of Poly(lactic-co-glycolic) Acid (PLGA)-Based Biomaterials for Bone Tissue Engineering." International Journal of Molecular Sciences 15(3): 3640-3659. Gharsallaoui, A., G. Roudaut, O. Chambin, A. Voilley and R. Saurel (2007). "Applications of spray-drying in microencapsulation of food ingredients: An overview." Food Research International 40(9): 1107-1121. Ghosh, R., M. Bhuiyan, I. Dewan, D. Ghosh and M. Asraful Islam (2012). Immediate Release Drug Delivery System (Tablets): An Overview. Gilchrist, S. E., D. Lange, K. Letchford, H. Bach, L. Fazli and H. M. Burt (2013). "Fusidic acid and rifampicin co-loaded PLGA nanofibers for the prevention of orthopedic implant associated infections." J Control Release 170(1): 64-73. Goldman, B. (2003). "Multidrug Resistance: Can New Drugs Help Chemotherapy Score Against Cancer?" Journal of the National Cancer Institute 95(4): 255-257.

Greene, F. L., C. M. Balch, I. D. Fleming, A. Fritz, D. G. Haller, M. Morrow and D. L. Page (2002). AJCC cancer staging handbook: TNM classification of malignant tumors, Springer Science & Business Media. Greene, F. L. and L. H. Sobin (2002). "The TNM system: Our language for cancer care." Journal of Surgical Oncology 80(3): 119-120. Gulfam, M., J.-e. Kim, J. M. Lee, B. Ku, B. H. Chung and B. G. Chung (2012). "Anticancer Drug-Loaded Gliadin Nanoparticles Induce Apoptosis in Breast Cancer Cells." Langmuir 28(21): 8216-8223. Gulfam, M., J. E. Kim, J. M. Lee, B. Ku, B. H. Chung and B. G. Chung (2012). "Anticancer drug-loaded gliadin nanoparticles induce apoptosis in breast cancer cells." Langmuir 28(21): 8216-8223. Hadjipanayis, C. G., G. Widhalm and W. Stummer (2015). "What is the Surgical Benefit of Utilizing 5-ALA for Fluorescence-Guided Surgery of Malignant Gliomas?" Neurosurgery 77(5): 663-673. Haider, A., S. Haider and I.-K. Kang (2015). "A comprehensive review summarizing the effect of electrospinning parameters and potential applications of nanofibers in biomedical and biotechnology." Arabian Journal of Chemistry. Halliwell, R. A., R. M. Bhardwaj, C. J. Brown, N. E. Briggs, J. Dunn, J. Robertson, A. Nordon and A. J. Florence (2017). "Spray drying as a reliable route to produce metastable carbamazepine form IV." Journal of Pharmaceutical Sciences. Han, D. and A. J. Steckl (2013). "Triaxial Electrospun Nanofiber Membranes for Controlled Dual Release of Functional Molecules." ACS Applied Materials & Interfaces 5(16): 8241-8245. Henderson, B. W. and T. J. Dougherty (1992). "How does photodynamic therapy work?" Photochemistry and Photobiology 55(1): 145-157. Henriques, C., R. Vidinha, D. Botequim, J. P. Borges and J. A. Silva (2009). "A systematic study of solution and processing parameters on nanofiber morphology using a new electrospinning apparatus." J Nanosci Nanotechnol 9(6): 3535-3545. Himri, I. and A. Guaadaoui (2018). Chapter 1 - Cell and organ drug targeting: Types of drug delivery systems and advanced targeting strategies. Nanostructures for the Engineering of Cells, Tissues and Organs. A. M. Grumezescu, William Andrew Publishing: 1-66. Ho, H. N., I. Laidmäe, K. Kogermann, A. Lust, A. Meos, C. N. Nguyen and J. Heinämäki (2017). "Development of electrosprayed artesunate-loaded core–shell nanoparticles." Drug development and industrial pharmacy 43(7): 1134-1142. Ho, T. M., T. Truong and B. Bhandari (2017). Chapter 5 - Spray-Drying and Non- Equilibrium States/Glass Transition. Non-Equilibrium States and Glass Transitions in Foods. B. Bhandari and Y. H. Roos, Woodhead Publishing: 111-136. Honey, P. J., J. Rijo, A. Anju and K. R. Anoop (2014). "Smart polymers for the controlled delivery of drugs – a concise overview." Acta Pharmaceutica Sinica B 4(2): 120-127. Hong, W. K. and W. N. Hait (2010). Holland Frei Cancer Medicine Eight, People's Medical Publishing House. American Association for Cancer Research. Hsi, R. A., D. I. Rosenthal and E. Glatstein (1999). "Photodynamic therapy in the treatment of cancer: current state of the art." Drugs 57(5): 725-734. Huang, Z., H. Xu, A. D. Meyers, A. I. Musani, L. Wang, R. Tagg, A. B. Barqawi and Y. K. Chen (2008). "Photodynamic therapy for treatment of solid tumors – potential and technical challenges." Technology in cancer research & treatment 7(4): 309-320. 69

Huntington, D. H. (2004). "The Influence of the Spray Drying Process on Product Properties." Drying Technology 22(6): 1261-1287. Illangakoon, U. E., D. G. Yu, B. S. Ahmad, N. P. Chatterton and G. R. Williams (2015). "5-Fluorouracil loaded Eudragit fibers prepared by electrospinning." Int J Pharm 495(2): 895-902. Ito, K., A. Yamaguchi, K. Miura, T. Kato, S. Baba, S. Matsumoto, M. Ishii and H. Takagi (1996). "Oral adjuvant chemotherapy with carmofur (HCFU) for colorectal cancer: five- year follow-up. Tokai HCFU Study Group--third study on colorectal cancer." J Surg Oncol 63(2): 107-111. Jain, M., L. Ganesh, B. Manoj, C. Randhir, B. Shashikant and S. Chirag (2012). Spray Drying in Pharmaceutical Industry: A Review. Joung, Y. K., I. K. Hwang, K. D. Park and C. W. Lee (2010). "CD34 monoclonal antibody- immobilized electrospun polyurethane for the endothelialization of vascular grafts." Macromolecular Research 18(9): 904-912. Kim, G. J. and S. Nie (2005). "Targeted cancer nanotherapy." Materials Today 8(8, Supplement): 28-33. Kim, W. and S. S. Kim (2011). "Synthesis of biodegradable triple-layered capsules using a triaxial electrospray method." Polymer 52(15): 3325-3336. Kimura, M., K. Miyajima, M. Kojika, T. Kono and H. Kato (2015). "Photodynamic Therapy (PDT) with Chemotherapy for Advanced Lung Cancer with Airway Stenosis." International Journal of Molecular Sciences 16(10): 25466-25475. Knudson, J. A. G. and L. C. Strons (1972). "Mutation and Cancer: A Model for Wilms' Tumor of the Kidney2." JNCI: Journal of the National Cancer Institute 48(2): 313-324. Kowalczyk, T., A. Nowicka, D. Elbaum and T. A. Kowalewski (2008). "Electrospinning of Bovine Serum Albumin. Optimization and the Use for Production of Biosensors." Biomacromolecules 9(7): 2087-2090. Kreidieh, F. Y., H. A. Moukadem and N. S. El Saghir (2016). "Overview, prevention and management of chemotherapy extravasation." World Journal of Clinical Oncology 7(1): 87-97. Kubler, A., C. Niziol, M. Sidhu, A. Dunne and J. A. Werner (2005). "[Analysis of cost effectiveness of photodynamic therapy with Foscan (Foscan-PDT) in comparison with palliative chemotherapy in patients with advanced head-neck tumors in Germany]." Laryngorhinootologie 84(10): 725-732. Kufe, D. W. and P. P. Major (1981). "5-Fluorouracil incorporation into human breast carcinoma RNA correlates with cytotoxicity." J Biol Chem 256(19): 9802-9805. Kumar, A., R. Kumar, A. Sandilium, J. Shukla and S. Pradhan (1999). "5-Fluorouracil induces defects in platelet function." Platelets 10(2-3): 137-140. Kumar, A. and J. Pillai (2018). Chapter 13 - Implantable drug delivery systems: An overview. Nanostructures for the Engineering of Cells, Tissues and Organs. A. M. Grumezescu, William Andrew Publishing: 473-511. Kumar Naraharisetti, P., B. Yung Sheng Ong, J. Wei Xie, T. Kam Yiu Lee, C.-H. Wang and N. V. Sahinidis (2007). "In vivo performance of implantable biodegradable preparations delivering Paclitaxel and Etanidazole for the treatment of glioma." Biomaterials 28(5): 886-894. Kuo, M. T. (2009). "Redox Regulation of Multidrug Resistance in Cancer Chemotherapy: Molecular Mechanisms and Therapeutic Opportunities." Antioxidants & Redox Signaling 11(1): 99-133. 70

Kwon, G. S. and D. Y. Furgeson (2007). 4 - Biodegradable polymers for drug delivery systems. Biomedical Polymers. M. Jenkins, Woodhead Publishing: 83-110. Langevin, P. B. and J. L. Atlee (2007). Chapter 30 - Chemotherapeutic Agents. Complications in Anesthesia (Second Edition). Philadelphia, W.B. Saunders: 110-118. Leach, M. K., Z.-Q. Feng, S. J. Tuck and J. M. Corey (2011). "Electrospinning Fundamentals: Optimizing Solution and Apparatus Parameters." Journal of Visualized Experiments : JoVE(47): 2494. Lee, S. Y., J. J. Lee, J. H. Park, J. Y. Lee, S. H. Ko, J. S. Shim, J. Lee, M. Y. Heo, D. D. Kim and H. J. Cho (2016). "Electrosprayed nanocomposites based on hyaluronic acid derivative and Soluplus for tumor-targeted drug delivery." Colloids Surf B Biointerfaces 145: 267-274. Li, Z. and C. Wang (2013). Effects of Working Parameters on Electrospinning. One- Dimensional nanostructures: Electrospinning Technique and Unique Nanofibers. Z. Li and C. Wang. Berlin, Heidelberg, Springer Berlin Heidelberg: 15-28. Liechty, W. B., D. R. Kryscio, B. V. Slaughter and N. A. Peppas (2010). "Polymers for Drug Delivery Systems." Annual review of chemical and biomolecular engineering 1: 149-173. Lin, R., L. Shi Ng and C.-H. Wang (2005). "In vitro study of anticancer drug doxorubicin in PLGA-based microparticles." Biomaterials 26(21): 4476-4485. Liu, S., G. Zhou, D. Liu, Z. Xie, Y. Huang, X. Wang, W. Wu and X. Jing (2013). "Inhibition of orthotopic secondary hepatic carcinoma in mice by doxorubicin-loaded electrospun polylactide nanofibers." Journal of Materials Chemistry B 1(1): 101-109. Longley, D. B., D. P. Harkin and P. G. Johnston (2003). "5-Fluorouracil: mechanisms of action and clinical strategies." Nature Reviews Cancer 3: 330. Luana, P., C. Andrea, T. Cagri and P. Dario (2013). "Industrial Upscaling of Electrospinning and Applications of Polymer Nanofibers: A Review." Macromolecular Materials and Engineering 298(5): 504-520. Lukáš, D., A. Sarkar, L. Martinová, K. Vodsed'álková, D. Lubasová, J. Chaloupek, P. Pokorný, P. Mikeš, J. Chvojka and M. Komárek (2009). "Physical principles of electrospinning (Electrospinning as a nano-scale technology of the twenty-first century)." Textile Progress 41(2): 59-140. Luo, X., C. Xie, H. Wang, C. Liu, S. Yan and X. Li (2012). "Antitumor activities of emulsion electrospun fibers with core loading of hydroxycamptothecin via intratumoral implantation." Int J Pharm 425(1-2): 19-28. Luu, Y. K., K. Kim, B. S. Hsiao, B. Chu and M. Hadjiargyrou (2003). "Development of a nanostructured DNA delivery scaffold via electrospinning of PLGA and PLA–PEG block copolymers." Journal of Controlled Release 89(2): 341-353. Lv, G., F. He, X. Wang, F. Gao, G. Zhang, T. Wang, H. Jiang, C. Wu, D. Guo, X. Li, B. Chen and Z. Gu (2008). "Novel Nanocomposite of Nano Fe3O4 and Polylactide Nanofibers for Application in Drug Uptake and Induction of Cell Death of Leukemia Cancer Cells." Langmuir 24(5): 2151-2156. Macmillan-Cancer-Support. (2018). "Fluorouracil (5FU)." Retrieved 21/06/2018, from https://www.macmillan.org.uk/information-and-support/treating/chemotherapy/drugs- and-combination-regimens/individual-drugs/fluorouracil.html. Macossay, J., A. Marruffo, R. Rincon, T. Eubanks and A. Kuang (2007). "Effect of needle diameter on nanofiber diameter and thermal properties of electrospun poly(methyl methacrylate)." Polymers for Advanced Technologies 18(3): 180-183.

Maeda, H. (2001). "The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumor-selective macromolecular drug targeting." Advances in Enzyme Regulation 41(1): 189-207. Maeda, H., J. Wu, T. Sawa, Y. Matsumura and K. Hori (2000). "Tumor vascular permeability and the EPR effect in macromolecular therapeutics: a review." J Control Release 65(1-2): 271-284. Malet-Martino, M., P. Jolimaitre and R. Martino (2002). "The prodrugs of 5-fluorouracil." Curr Med Chem Anticancer Agents 2(2): 267-310. Manasa, V., K. Vanitha, M. Venkataswamy, P. Prabhakar and A. Ramesh (2016). "Formulation and Evaluation of Aqueous Enteric Coated Delayed Release Omeprazole Pellets." Research Journal of Pharmaceutical Dosage Forms and Technology 8(1): 1. Mao, C. and L. M. Obeid (2008). "Ceramidases: regulators of cellular responses mediated by ceramide, sphingosine, and sphingosine-1-phosphate." Biochim Biophys Acta 1781(9): 424-434. Mathur, S. and J. Sutton (2017). "Personalized medicine could transform healthcare." Biomedical Reports 7(1): 3-5. Matthews, J. A., G. E. Wnek, D. G. Simpson and G. L. Bowlin (2002). "Electrospinning of Collagen Nanofibers." Biomacromolecules 3(2): 232-238. Maury, M., K. Murphy, S. Kumar, A. Mauerer and G. Lee (2005). "Spray-drying of proteins: effects of sorbitol and trehalose on aggregation and FT-IR amide I spectrum of an immunoglobulin G." European Journal of Pharmaceutics and Biopharmaceutics 59(2): 251-261. Meenach, S. A., K. W. Anderson, J. Zach Hilt, R. C. McGarry and H. M. Mansour (2013). "Characterization and aerosol dispersion performance of advanced spray-dried chemotherapeutic PEGylated phospholipid particles for dry powder inhalation delivery in lung cancer." European Journal of Pharmaceutical Sciences 49(4): 699-711. Megelski, S., J. S. Stephens, D. B. Chase and J. F. Rabolt (2002). "Micro- and Nanostructured Surface Morphology on Electrospun Polymer Fibers." Macromolecules 35(22): 8456-8466. Mills, J. K. and D. Needham (1999). "Targeted drug delivery." Expert Opinion on Therapeutic Patents 9(11): 1499-1513. Mitchell, G. and A. Tojeira (2013). Role of Anisotropy in Tissue Engineering. Moghimi, S. M. and Z. S. Farhangrazi (2014). Chapter 6 - Nanoparticles in Medicine: Nanoparticle Engineering for Macrophage Targeting and Nanoparticles that Avoid Macrophage Recognition. Nanoparticles and the Immune System. D. Boraschi and A. Duschl. San Diego, Academic Press: 77-89. Mönckedieck, M., J. Kamplade, E. M. Littringer, A. Mescher, S. Gopireddy, M. Hertel, E. Gutheil, P. Walzel, N. A. Urbanetz and M. Köster (2016). Spray Drying Tailored Mannitol Carrier Particles for Dry Powder Inhalation with Differently Shaped Active Pharmaceutical Ingredients. Process-Spray, Springer: 517-566. Moses, M. A., H. Brem and R. Langer (2003). "Advancing the field of drug delivery: Taking aim at cancer." Cancer Cell 4(5): 337-341. Muller, R. H. and C. M. Keck (2004). "Challenges and solutions for the delivery of biotech drugs – a review of drug nanocrystal technology and lipid nanoparticles." Journal of Biotechnology 113(1): 151-170.

Muller, R. H., K. Mader and S. Gohla (2000). "Solid lipid nanoparticles (SLN) for controlled drug delivery - a review of the state of the art." Eur J Pharm Biopharm 50(1): 161-177. Mumenthaler, M., C. C. Hsu and R. Pearlman (1994). "Feasibility study on spray-drying protein pharmaceuticals: Recombinant human growth hormone and tissue-type plasminogen activator." Pharmaceutical research 11(1): 12-20. Munir, M. M., A. B. Suryamas, F. Iskandar and K. Okuyama (2009). "Scaling law on particle-to-fiber formation during electrospinning." Polymer 50(20): 4935-4943. Nakamura, T., M. Ohno, Y. Tabuchi, T. Kamigaki, H. Fujii, H. Yamagishi and Y. Kuroda (2001). "Optimal duration of oral adjuvant chemotherapy with Carmofur in the colorectal cancer patients: the Kansai Carmofur Study Group trial III." Int J Oncol 19(2): 291-298. National Cancer Institute. (2017, April 6, 2017). "Types of Cancer Treatment." Retrieved 26/06/2018, from https://www.cancer.gov/about-cancer/treatment/types. National Cancer Institute. (2018). "Cancer Survival Statistics." Retrieved 26/06/2018, from https://seer.cancer.gov/statistics/types/survival.html. Naveed, A., C. Mora‐Huertas, H. Jaafar‐Maalej and E. Fessi Abdelhamid Polymeric Drug Delivery Systems for Encapsulating Hydrophobic Drugs. Drug Delivery Strategies for Poorly Water‐Soluble Drugs. Newman, A. (2015). Pharmaceutical Amorphous Solid Dispersions, Wiley. NHSChoices. (2011). "Cancer survival rates 'threatened by rising cost'." from https://www.nhs.uk/news/cancer/cancer-survival-rates-threatened-by-rising-cost/. Nie, H., Y. Fu and C. H. Wang (2010). "Paclitaxel and suramin-loaded core/shell microspheres in the treatment of brain tumors." Biomaterials 31(33): 8732-8740. Noguchi, Y., J. Wu, R. Duncan, J. Strohalm, K. Ulbrich, T. Akaike and H. Maeda (1998). "Early phase tumor accumulation of macromolecules: a great difference in clearance rate between tumor and normal tissues." Jpn J Cancer Res 89(3): 307-314. Nurwaha, D., W. Han and X. Wang (2013). "Effects of processing parameters on electrospun fiber morphology." The Journal of The Textile Institute 104(4): 419-425. O’Rourke, C., C. Hopper, A. J. MacRobert, J. B. Phillips and J. H. Woodhams (2017). "Could clinical photochemical internalisation be optimised to avoid neuronal toxicity?" International Journal of Pharmaceutics 528(1-2): 133-143. Pal, S. K. and A. Hurria (2012). Chapter 9 - Chemotherapy A2 - Naeim, Arash. Management of Cancer in the Older Patient. D. B. Reuben and P. A. Ganz. Philadelphia, W.B. Saunders: 89-94. Palumbo, M. O., P. Kavan, W. H. Miller, L. Panasci, S. Assouline, N. Johnson, V. Cohen, F. Patenaude, M. Pollak, R. T. Jagoe and G. Batist (2013). "Systemic cancer therapy: achievements and challenges that lie ahead." Frontiers in Pharmacology 4: 57. Patrizia, A., B. Kristian, C. K. A., F. T. H., G. A. W., G. S. O., H. S. M., H. M. R., J. Asta, K. David, K. Mladen, M. Johan, M. Pawel, N. Dominika, P. Jacques, W. B. C. and G. Jakub (2011). "Photodynamic therapy of cancer: An update." CA: A Cancer Journal for Clinicians 61(4): 250-281. Pillai, O. and R. Panchagnula (2001). "Polymers in drug delivery." Current Opinion in Chemical Biology 5(4): 447-451. Pizzirani, D., C. Pagliuca, N. Realini, D. Branduardi, G. Bottegoni, M. Mor, F. Bertozzi, R. Scarpelli, D. Piomelli and T. Bandiera (2013). "Discovery of a New Class of Highly Potent Inhibitors of Acid Ceramidase: Synthesis and Structure–Activity Relationship (SAR)." Journal of Medicinal Chemistry 56(9): 3518-3530. 73

Poste, G. and R. Kirsh (1983). "Site–specific (targeted) drug delivery in cancer therapy." Nature Biotechnology 1(10): 869-878. Pray, L. (2008). "DNA replication and causes of mutation." Nature education 1(1): 214. Pridgen, E. M., F. Alexis and O. C. Farokhzad (2015). "Polymeric Nanoparticle Drug Delivery Technologies for Oral Delivery Applications." Expert opinion on drug delivery 12(9): 1459-1473. Ramachandran, R., V. R. Junnuthula, G. S. Gowd, A. Ashokan, J. Thomas, R. Peethambaran, A. Thomas, A. K. K. Unni, D. Panikar, S. V. Nair and M. Koyakutty (2017). "Theranostic 3-Dimensional nano brain-implant for prolonged and localized treatment of recurrent glioma." Scientific Reports 7: 43271. Ramakrishna, S., K. Fujihara, W.-E. Teo, T. Yong, Z. Ma and R. Ramaseshan (2006). "Electrospun nanofibers: solving global issues." Materials Today 9(3): 40-50. Ré, M.-I. (2006). "Formulating Drug Delivery Systems by Spray Drying." Drying Technology 24(4): 433-446. Realini, N., C. Solorzano, C. Pagliuca, D. Pizzirani, A. Armirotti, R. Luciani, M. P. Costi, T. Bandiera and D. Piomelli (2013). "Discovery of highly potent acid ceramidase inhibitors with in vitro tumor chemosensitizing activity." Sci. Rep. 3. Ribeiro, C., V. Sencadas, C. M. Costa, J. L. Gómez Ribelles and S. Lanceros-Méndez (2011). "Tailoring the morphology and crystallinity of poly(L-lactide acid) electrospun membranes." Science and Technology of Advanced Materials 12(1): 015001. Roh, K.-H., D. C. Martin and J. Lahann (2005). "Biphasic Janus particles with nanoscale anisotropy." Nat Mater 4(10): 759-763. Saini, R., N. V. Lee, K. Y. P. Liu and C. F. Poh (2016). "Prospects in the Application of Photodynamic Therapy in Oral Cancer and Premalignant Lesions." Cancers 8(9): 83. Sampath, M., R. Lakra, P. Korrapati and S. Balasubramanian (2014). Curcumin loaded poly (lactic-co-glycolic) acid nanofiber for the treatment of carcinoma. Sanoff, H. K., W. R. Carpenter, J. Freburger, L. Li, K. Chen, L. L. Zullig, R. M. Goldberg, M. J. Schymura and D. Schrag (2012). "Comparison of Adverse Events During 5- fluorouracil versus 5-fluorouracil/Oxaliplatin Adjuvant Chemotherapy for Stage III Colon Cancer: A Population Based Analysis." Cancer 118(17): 4309-4320. Santos, C., C. F. Rovath, R. P. Franke, M. M. Almeida and M. E. V. Costa (2009). "Spray- dried hydroxyapatite-5-Fluorouracil granules as a chemotherapeutic delivery system." Ceramics International 35(1): 509-513. Savage, P. and S. Mahmoud (2015). "Development and economic trends in cancer therapeutic drugs: a 5-year update 2010–2014." British Journal of Cancer 112(6): 1037- 1041. Schulmeister, L. (2006). "Preventing chemotherapy errors." The oncologist 11(5): 463- 468. Semwal, R., R. Semwal and D. Semwal (2015). Drug Delivery Systems: Use and Selection Criteria. Shafirstein, G., D. Bellnier, E. Oakley, S. Hamilton, M. Potasek, K. Beeson and E. Parilov (2017). "Interstitial Photodynamic Therapy—A Focused Review." Cancers 9(2): 12. Sharquie, K. E. and A. A. Noaimi (2012). "Basal cell carcinoma: Topical therapy versus surgical treatment." Journal of the Saudi Society of Dermatology & Dermatologic Surgery 16(2): 41-51.

Shirshahi, V., F. Shamsipour, A. H. Zarnani, J. Verdi and R. Saber (2013). "Active targeting of HER2-positive breast cancer cells by Herceptin-functionalized organically modified silica nanoparticles." Cancer Nanotechnology 4(1-3): 27-37. Shrestha, H., R. Bala and S. Arora (2014). "Lipid-Based Drug Delivery Systems." Journal of Pharmaceutics 2014: 10. Sill, T. J. and H. A. von Recum (2008). "Electrospinning: Applications in drug delivery and tissue engineering." Biomaterials 29(13): 1989-2006. Silverstein, A., V. B. Silverstein and L. S. Nunn (2006). Cancer: Conquering a Deadly Disease, Twenty-First Century Books. Simone, C. B., J. S. Friedberg, E. Glatstein, J. P. Stevenson, D. H. Sterman, S. M. Hahn and K. A. Cengel (2012). "Photodynamic therapy for the treatment of non-small cell lung cancer." Journal of Thoracic Disease 4(1): 63-75. Singh, R. and J. W. Lillard (2009). "Nanoparticle-based targeted drug delivery." Experimental and molecular pathology 86(3): 215-223. Society, A. C. (2018). "Treatment Types." Retrieved 25/06/2018, 2018, from https://www.cancer.org/treatment/treatments-and-side-effects/treatment-types.html. Socinski, M. A., D. E. Morris, G. A. Masters and R. Lilenbaum (2003). "Chemotherapeutic Management of Stage IV Non-small Cell Lung Cancer^*." CHEST 123(1): 226S-243S. Sollohub, K. and K. Cal (2010). "Spray Drying Technique: II. Current Applications in Pharmaceutical Technology." Journal of Pharmaceutical Sciences 99(2): 587-597. Song, M., D. Guo, C. Pan, H. Jiang, C. Chen, R. Zhang, Z. Gu and X. D. Wang (2008). The application of poly(N-isopropylacrylamide)-co-polystyrene nanofibers as an additive agent to facilitate the cellular uptake of an anticancer drug. Songsurang, K., N. Praphairaksit, K. Siraleartmukul and N. Muangsin (2011). "Electrospray fabrication of doxorubicin-chitosan-tripolyphosphate nanoparticles for delivery of doxorubicin." Archives of Pharmacal Research 34(4): 583. Sosnik, A. and K. P. Seremeta (2015). "Advantages and challenges of the spray-drying technology for the production of pure drug particles and drug-loaded polymeric carriers." Advances in Colloid and Interface Science 223: 40-54. Spring, B. Q., I. Rizvi, N. Xu and T. Hasan (2015). "The role of photodynamic therapy in overcoming cancer drug resistance." Photochemical & photobiological sciences : Official journal of the European Photochemistry Association and the European Society for Photobiology 14(8): 1476-1491. Sridhar, R. and S. Ramakrishna (2013). "Electrosprayed nanoparticles for drug delivery and pharmaceutical applications." Biomatter 3(3): e24281. Strebhardt, K. and A. Ullrich (2008). "Paul Ehrlich's magic bullet concept: 100 years of progress." Nature Reviews Cancer 8: 473. Strimbu, K. and J. A. Tavel (2010). "What are Biomarkers?" Current opinion in HIV and AIDS 5(6): 463-466. Sun, Z., E. Zussman, A. L. Yarin, J. H. Wendorff and A. Greiner (2003). "Compound Core–Shell Polymer Nanofibers by Co‐Electrospinning." Advanced Materials 15(22): 1929-1932. Tahover, E., Y. P. Patil and A. A. Gabizon (2015). "Emerging delivery systems to reduce doxorubicin cardiotoxicity and improve therapeutic index: focus on liposomes." Anticancer Drugs 26(3): 241-258.

Tanaka, H., H. Matsushima, N. Mizumoto and A. Takashima (2009). "Classification of Chemotherapeutic Agents Based on Their Differential In vitro Effects on Dendritic Cells." Cancer Research 69(17): 6978-6986. Tang, P., D.-M. Zhang, N.-H. Xuan, S. Tsui, M. Waye, S.-K. Kong, W.-P. Fong and K.- P. Fung (2009). "Photodynamic therapy inhibits p-glycoprotein mediated multidrug resistance via JNK activation in human hepatocellular carcinoma using the photosensitizer pheophorbide a." Molecular Cancer 8(1): 56. Taylor, G. (1964). Disintegration of Water Drops in an Electric Field. Tiancheng, L., J. Xiabing, S. Xiangfu and W. Xiuran (2012). "Doxorubicin‐loaded ultrafine PEG‐PLA fiber mats against hepatocarcinoma." Journal of Applied Polymer Science 123(1): 209-217. Tiwari, G., R. Tiwari, B. Sriwastawa, L. Bhati, S. Pandey, P. Pandey and S. K. Bannerjee (2012). "Drug delivery systems: An updated review." International Journal of Pharmaceutical Investigation 2(1): 2-11. Torchilin, V. P. (2010). "Passive and active drug targeting: drug delivery to tumors as an example." Handb Exp Pharmacol(197): 3-53. Triesscheijn, M., P. Baas, J. H. M. Schellens and F. A. Stewart (2006). "Photodynamic Therapy in Oncology." The Oncologist 11(9): 1034-1044. Trotta, M., R. Cavalli, C. Trotta, R. Bussano and L. Costa (2010). "Electrospray technique for solid lipid-based particle production." Drug Dev Ind Pharm 36(4): 431-438. Tu, H., C. P. Wen, S. P. Tsai, W.-H. Chow, C. Wen, Y. Ye, H. Zhao, M. K. Tsai, M. Huang, C. P. Dinney, C. K. Tsao and X. Wu (2018). "Cancer risk associated with chronic diseases and disease markers: prospective cohort study." BMJ 360. UK, C. R. (2017). "Cancer diagnosis and treatment statistics." Retrieved 25/06/2018. Vasir, J. K. and V. Labhasetwar (2005). "Targeted Drug Delivery in Cancer Therapy." Technology in Cancer Research & Treatment 4(4): 363-374. Ventola, C. L. (2012). "The Nanomedicine Revolution: Part 2: Current and Future Clinical Applications." Pharmacy and Therapeutics 37(10): 582-591. Wachsberger, P., R. Burd and A. P. Dicker (2003). "Tumor Response to Ionizing Radiation Combined with Antiangiogenesis or Vascular Targeting Agents: Exploring Mechanisms of Interaction." Clinical Cancer Research 9(6): 1957-1971. Webb, C. E. and J. D. C. Jones (2004). Handbook of Laser Technology and Applications: Laser Design and laser systems, CRC Press. Wei, J., J. Hu, M. Li, Y. Chen and Y. Chen (2014). Multiple drug-loaded electrospun PLGA/gelatin composite nanofibers encapsulated with mesoporous ZnO nanospheres for potential postsurgical cancer treatment. Weinberg, R. (2013). The Biology of Cancer, Second Edition, Taylor & Francis Group. Wilczewska, A. Z., K. Niemirowicz, K. H. Markiewicz and H. Car (2012). "Nanoparticles as drug delivery systems." Pharmacol Rep 64(5): 1020-1037. Williams, G. R., N. P. Chatterton, T. Nazir, D. G. Yu, L. M. Zhu and C. J. Branford-White (2012). "Electrospun nanofibers in drug delivery: recent developments and perspectives." Ther Deliv 3(4): 515-533. Wilson, B. C. (2002). "Photodynamic therapy for cancer: principles." Can J Gastroenterol 16(6): 393-396. Wolinsky, J. B., Y. L. Colson and M. W. Grinstaff (2012). "Local Drug Delivery Strategies for Cancer Treatment: Gels, Nanoparticles, Polymeric Films, Rods, and Wafers." Journal 76

of controlled release : official journal of the Controlled Release Society 159(1): 10.1016/j.jconrel.2011.1011.1031. Woo, M. W. and B. Bhandari (2013). 2 - Spray drying for food powder production. Handbook of Food Powders. B. Bhandari, N. Bansal, M. Zhang and P. Schuck, Woodhead Publishing: 29-56. Xie J., et al (2008) "Putting Electrospun Nanofibers to Work for Biomedical Research" Macromolecular Rapid Communications 29(22): 1775-1792 Xie, C., X. Li, X. Luo, Y. Yang, W. Cui, J. Zou and S. Zhou (2010). "Release modulation and cytotoxicity of hydroxycamptothecin-loaded electrospun fibers with 2-hydroxypropyl- beta-cyclodextrin inoculations." Int J Pharm 391(1-2): 55-64. Xie, J., L. K. Lim, Y. Phua, J. Hua and C.-H. Wang (2006). "Electrohydrodynamic atomization for biodegradable polymeric particle production." Journal of Colloid and Interface Science 302(1): 103-112. Xie, J., J. C. Marijnissen and C. H. Wang (2006). "Microparticles developed by electrohydrodynamic atomization for the local delivery of anticancer drug to treat C6 glioma in vitro." Biomaterials 27(17): 3321-3332. Xie, J. and C. H. Wang (2006). "Electrospun micro- and nanofibers for sustained delivery of paclitaxel to treat C6 glioma in vitro." Pharm Res 23(8): 1817-1826. Xu, X., X. Chen, Z. Wang and X. Jing (2009). "Ultrafine PEG-PLA fibers loaded with both paclitaxel and doxorubicin hydrochloride and their in vitro cytotoxicity." Eur J Pharm Biopharm 72(1): 18-25. Xu, X., X. Chen, X. Xu, T. Lu, X. Wang, L. Yang and X. Jing (2006). "BCNU-loaded PEG- PLLA ultrafine fibers and their in vitro antitumor activity against Glioma C6 cells." J Control Release 114(3): 307-316. Yoo, J. J., C. Kim, C. W. Chung, Y. I. Jeong and D. H. Kang (2012). "5-aminolevulinic acid-incorporated poly(vinyl alcohol) nanofiber-coated metal stent for application in photodynamic therapy." Int J Nanomedicine 7: 1997-2005. Zahedi, P., Z. Karami, I. Rezaeian, S.-H. Jafari, P. Mahdaviani, A. H. Abdolghaffari and M. Abdollahi (2012). "Preparation and performance evaluation of tetracycline hydrochloride loaded wound dressing mats based on electrospun nanofibrous poly(lactic acid)/poly(ϵ-caprolactone) blends." Journal of Applied Polymer Science 124(5): 4174- 4183. Zamani, M., M. P. Prabhakaran and S. Ramakrishna (2013). "Advances in drug delivery via electrospun and electrosprayed nanomaterials." International J of Nanomedicine 8: 2997-3017. Zhang, C.-L., F.-H. Cao, J.-L. Wang, Z.-L. Yu, J. Ge, Y. Lu, Z.-H. Wang and S.-H. Yu (2017). "Highly Stimuli-Responsive Au Nanorods/Poly(N-isopropylacrylamide) (PNIPAM) Composite Hydrogel for Smart Switch." ACS Applied Materials & Interfaces 9(29): 24857-24863. Zhang, J., C. Jiang, J. P. Figueiró Longo, R. B. Azevedo, H. Zhang and L. A. Muehlmann (2018). "An updated overview on the development of new photosensitizers for anticancer photodynamic therapy." Acta Pharmaceutica Sinica B 8(2): 137-146. Zhang, J., X. Wang, T. Liu, S. Liu and X. Jing (2016). "Antitumor activity of electrospun polylactide nanofibers loaded with 5-fluorouracil and oxaliplatin against colorectal cancer." Drug Deliv 23(3): 794-800. Zhou, L., L. Zhou, S. Wei, X. Ge, J. Zhou, H. Jiang, F. Li and J. Shen (2014). "Combination of chemotherapy and photodynamic therapy using graphene oxide as drug delivery system." Journal of Photochemistry and Photobiology B: Biology 135: 7-16. 77

2. Janus structures generated by electrohydrodynamic atomization

2.1 Introduction

Janus structures

Janus structures have a compartmentalized architecture. They are “two-faced” materials with different surface structure on opposing sides (Pang, Wan et al. 2014). Their synthesis and fabrication are challenging, which hindered their development for a long time (Walther and Muller 2008). However, today there are various ways to produce these particles; for instance, self-assembly of block copolymers, the masking/unmasking technique, phase separation, controlled surface nucleation and electrodynamic jetting techniques (Perro, Reculusa et al. 2005). The structures of some Janus architectures are depicted in Figure 2-1.

The range of potential applications of these structures is wide, as they exhibit distinct chemistry and morphology: they can be used as imaging sensors, electronic paper, photonic crystals, catalyst support stabilizers, imaging agent nanosensors and for drug delivery (Lee, Yoon et al. 2011).

Figure 2-1. Schematic representations of Janus type particles. Adapted from Perro et.al. (Perro, Reculusa et al. 2005)

Multi drug co-delivery

Janus structures can be useful for drug delivery because different active ingredients can be incorporated into a single particle to allow simultaneous release (Jiang, 2012).

Therapies in which multiple therapeutic agents act on different molecular pathways could lead to a synergistic effect using such systems (Mokhtari, Homayouni et al. 2017). For example, Khan et al. developed drug loaded poly(acrylamide)/poly(methyl acrylate) particles using a microfluidic device. These gave a controlled release of both ketoprofen and sodium fluorescein, which proves that the incorporation and co-delivery of two different active ingredients can be successful even with molecules of very different hydrophilicity (Khan, Serra et al. 2014).

The entrapment of hydrophobic and hydrophilic compounds in a single particle can be a challenge, and Janus structures can be used to overcome this issue. In 2014, Janus particles were synthesized using biodegradable and biocompatible materials to entrap curcumin, a hydrophobic drug, and doxorubicin, a hydrophilic chemotherapeutic, to 79

achieve local delivery to the lungs (Garbuzenko, Winkler et al. 2014). These particles appeared to have therapeutic efficiency when tested in human lung cancer cells, and accumulated in the lungs when inhaled by mice. The Janus particles demonstrated efficacy in suppressing tumour growth, while free doxorubicin only slightly limited growth

(Garbuzenko, Winkler et al. 2014).

In other work, biocompatible Janus particles made of poly(lactic-co-glycolic acid) (PLGA) have been shown to be able to entrap hydrophilic and hydrophobic chemotherapeutic drugs (doxorubicin chloride and paclitaxel, respectively) via fluidic based precipitation

(Xie, She et al. 2012). The release of the drugs from the Janus particles presented biphasic drug release, one fast release and one sustained release. When compared to the drug release from monophasic particles, paclitaxel was released in a higher quantity from the Janus particles in the burst phase, while doxorubicin release was at the same rate as the monophasic particles. This avoids the interference of drugs and facilitates the absorption of both drugs in the same carrier (Xie, She et al. 2012).

Targeted drug delivery

Compartmentalized particles comprising poly(ethylene oxide) (PEO) and poly(acrylamide-co-acrylic acid) in one hemisphere and a crosslinked copolymer of dextran and poly(acrylamide-co-acrylic acid) segments in the second compartment were made with electrohydrodynamic atomization (see Section 1.4) and shown to display selective degradation at different pHs, as well as dual-phase release kinetics (Hwang and Lahann 2012). A fluorescein isothiocyanate (FITC)‐conjugated PEO was used to track the degradation of the PEO compartment. It was demonstrated that both compartments of the particles were stable at pH 3.0, but the PEO-containing compartment underwent selective degradation at pH 7.4 over a period of 5 days (Hwang and Lahann 2012). Such particles with different degradable polymer compartments may be used for a range of oral drug delivery applications because of their capacity to provide a decoupled release profile. 80

Theranostics

Multifunctional nanoparticles can be useful for delivering drugs and imaging tags at the same time, which enables the monitoring of the circulation and biodistribution of a drug carrier (Sahu and Mohapatra 2013). For example, effective targeting for breast cancer has been achieved by making Janus particles composed of gold and polystyrene selectively functionalized with ligands (Yang, Guo et al. 2012). Gold permits in vitro monitoring via surface-enhanced Raman spectroscopy while the polystyrene functions as a targeting agent (Wu, Ross et al. 2010). Simultaneous delivery of siRNA and imaging has been proved possible using materials prepared by electrohydrodynamic atomization

(Misra, Bhaskar et al. 2012). These Janus particles were based on PLGA and polyethyleneimine (PEI), each of them in a different compartment. The PLGA compartment was employed as an imaging compartment and was labelled with a blue fluorescent dye that remained throughout the study. On the other hand, the PEI was utilized as an endosome sensing and escape compartment loaded with siRNA, which could consequently silence the green fluorescent protein (GFP) gene (Misra, Bhaskar et al. 2012).

Preparation of Janus structures

Micron-sized particles with anisotropic bicompartmental characteristics can be prepared effectively by electrohydrodynamic co-jetting. Essentially, this involves a side-by-side capillary needle system, which permits two different solutions to move under a continuous flow. Two polymer solutions are placed in different syringes. The syringes are mounted on two pumps, one for each polymer solution, and these are programmed according to the parameters desired. The polymer solutions are electrospun/electrosprayed at the same flow rate and at the same voltage (Figure 2-2).

As mentioned in Chapter 1, this technique offers reliable control and facile incorporation of active ingredients (i.e. drugs, metallic particles, ligands) into the two sides of the particles (Walther and Muller 2008).

Figure 2-2. A schematic illustration of side by side electrospinning (left) and electrospraying (right).

Two-faced Janus particles can be generated with these techniques, and in addition multiple compartments can also be produced by changing the needle arrangement of the equipment. For example, tri-compartmental particles (Roh, Martin et al. 2006), multi- compartmentalized fibres (Mandal, Bhaskar et al. 2009) and multifunctional cylinders

(Sokolovskaya, Yoon et al. 2013) have been successfully generated by electrohydrodynamic techniques (EHD) techniques.

Janus fibres can have various advantages over the monolithic fibres arising from single fluid electrospinning. For example, Yu and his group developed highly tuneable side-by- side drug loaded fibres (Yu, Yang et al. 2016). Each compartment was made of a different polymer (one of PVP and one of ethylcellulose (EC)), and both contained ketoprofen. PVP provided a fast release of the drug, while release from the EC side was much slower and provided sustained release. The doping of small amounts of PVP to the EC side offers the possibility to make the release faster, which allows this delivery system to be precisely tuned. Both fluids were also electrospun separately to create monolithic fibres, which released the entire drug load in one min; burst release is one of

the main disadvantages of single fluid electrospinning, where Janus fibres prove to overcome this challenge (Yu, Yang et al. 2016).

Wang et al investigated a new approach to create Janus structures by using an acentric spinneret (Wang, Liu et al. 2018). Janus fibres were prepared encapsulating a hydrophobic herbal medicine (helicid), used to cure insomnia and headaches. This medicine needs rapid drug delivery, as a fast onset of therapeutic action is required.

Wang wanted to prove that a non-spinnable solution (a solution that cannot be converted into solid fibers) can be electrospun using the acentric spinneret. Two fluids were employed, one for each side of the fibres: PVP K10 and sodium dodecyl sulfate (SDS)

(non-spinnable), and PVP K90 and helicid (spinnable). Drug-loaded Janus fibres were successfully prepared, proving the concept that an acentric spinneret can create Janus nanostructures from a spinnable and an unspinnable solution. In addition, the Janus fibres demonstrated improved hydrophilicity and transmembrane permeation of helicid when compared with monolithic fibres (Wang, Liu et al. 2018).

2.2 Aims and objectives

EHD techniques exhibit remarkable versatility and robustness in terms of the variety of polymers/drugs that can be used to construct complex nanomedicinal formulations. This chapter explores the formation of Janus formulations co-loaded with anticancer drugs

(e.g. carmofur,5-FU) and a photosensitizer (rose bengal (RB) and hematoporphyrin).

The key aims are to:

• Determine if Janus structures containing a photosensitizer and a

chemotherapeutic drug can be produced by EHD fabrication techniques,

exploring a range of polymers.

• Characterise the physicochemical and in vitro drug release properties of the

fabricated materials.

• Evaluate the formulations’ cytotoxicity in model non-cancerous and a cancerous

cell lines. 83

2.3 Materials and Methods

2.3.1 Materials

Rose bengal (RB), hematoporphyrin (HP), polyvinylpyrrolidone (PVP, molecular weight

360 kDa) and polycaprolactone (PCL, molecular weight 45 kDa), were supplied by

Sigma-Aldrich. Ethanol was purchased from Fisher Scientific Ltd. PVP (molecular weight

56 kDa), was sourced from Alfa Aesar. Carmofur was supplied by Cambridge Scientific.

2.3.2 Preparation of drug-loaded electrospun fibres and electrosprayed particles

PVP solutions at 10% w/v or 5% w/v were prepared in ethanol and stirred in sealed vials until complete dissolution of the polymer had occurred. Hematoporphyrin (HP) and RB were added at a concentration of 10% w/w (with respect to the PVP content), with stirring continuously at room temperature for 24h to allow complete drug dissolution. These solutions were then processed by EHD.

PCL solutions were prepared at a concentration of 10% w/v, using chloroform as the solvent. RB was added at a concentration of 1% or 0.1% w/v. A mixed solution of carmofur (0.8% w/v) and PCL (10% w/v) was also prepared in chloroform. The solutions were stirred for 1 h at room temperature.

Solutions were pumped at a controlled flow rate between 0.5-1 mL/h. The polymer solution was ejected through a 5 mL syringe (Terumo) connected to a syringe pump

(KDS100, Cole Parmer). A high voltage DC power supply (0-35kV, 0-1 mA, HCP 35-

35000, FuG Elektronik) was employed to provide a high voltage (in the range of 10-15 kV). Two different spinnerets were used. For side-by-side spinning, the spinneret consisted of two adjacent 20G (outer diameter: 0.9 mm, inner diameter: 0.6 mm) stainless steel needles (Figure 2-2). Monolithic fibres were processed using a single 21G

(outer diameter: 0.81 mm, inner diameter: 0.51 mm) needle. A grounded collector (20x20

cm) was wrapped in aluminium foil, and placed at a tip to collector distance of 10-15 cm.

Electrospinning was performed within a temperature range from 17-26 oC and a relative humidity 21-55% After processing, the aluminium foil was stored in a desiccator for at least 12 h to ensure there was no remaining solvent in the fibres/particles. Silica gel beads were used as the desiccant.

2.3.3 Scanning electron microscopy

The morphology of the materials produced was studied using scanning electron microscopy (SEM). Samples were adhered to an SEM stub with carbon-coated double- sided tape. Afterwards, the samples were coated with a 20 nm gold-sputter (Quorum

Q150T). A field emission scanning electron microscope (FEI Quanta 200F) connected to a secondary electron detector (Everheart-Thornley Detector-ETD) was used to analyze the materials. The size distribution was determined from the SEM micrographs with the ImageJ software (National Institutes of Health, US) by measuring the diameter of at least 100 fibres/particles.

2.3.4 Fluoresence microscopy

A fluorescence microscope (EVOS® FL Cell Imaging System, ThermoFisher Scientific) with GFP and DAPI filters was used to analyze the samples, and a digital camera attached to the microscope employed to acquire images.

2.3.5 Fourier-transform infrared spectroscopy (FTIR)

FTIR spectra of the samples (approx. 0.2 x 0.2 cm2 of the fibre mat) were obtained using a PerkinElmer Spectrum 100 spectrometer fitted with an ATR attachment. The scanning range was 650-4000 cm-1, with a resolution of 1 cm-1 and four scans obtained.

2.3.6 Differential scanning calorimetry (DSC)

Analysis was conducted using a Q2000 DSC (TA Instruments). A small amount of sample was placed inside a non-hermetically sealed aluminium pan (T130425, TA

Instruments). DSC analysis was carried out from 0 ºC - 140 ºC or 0 ºC – 200 ºC 85

depending on the sample, at a temperature ramp of 10 ºC/min. Oxygen-free nitrogen gas at a purge rate of 50 mL/min was supplied to the furnace throughout the experiment.

Data analysis was carried out using the TA Universal Analysis software (Waters LLC).

2.3.7 Thermogravimetric analysis (TGA)

TGA was conducted on a Discovery TGA (TA Instruments). Approximately 7 mg of sample was added to a tared aluminum pan (TA Instruments). The sample was then heated from 40 oC to 300 oC or 1100 oC at a temperature ramp of 10 oC/min. Experiments were conducted with a nitrogen plunge of 25 mL/min. Data were analysed using the Trios software (TA Instruments).

2.3.8 X-ray diffraction

X-ray diffraction (XRD) patterns of the samples and reference materials were obtained using a Miniflex 600 (Rigaku) diffractometer supplied with Cu Kα radiation. An aluminium holder was used. The patterns were recorded in the 2Ө range 3-40º at a speed of 5º/min.

The generator voltage was set at 40 kV and the current at 15 mA. The data were analysed using OriginPro 9.1.

2.3.9 Drug release studies

Drug release was carried out in 750 mL of phosphate buffered saline (PBS; pH 7.4) at a temperature of 37 ºC, with these conditions employed to mimic the average pH and temperature of the blood (Rhoades and Bell 2012).

10 mg of the fibre mats or particles was accurately weighed and placed in a sinker or a capsule (fast-dissolving disintegrating size1 HPMC capsule, Capsugel), respectively.

Samples were placed in the dissolution containers under continuous stirring at 100 rpm in order to ensure good mixing. Experiments were performed for 2-4 hours, depending on the formulation tested. Drug release was quantified using a Cary 100 UV-vis spectrometer (Agilent). The wavelengths utilized for the experiments were: RB 547 nm,

hematoporphyrin 397 nm and carmofur 258 nm, as found in the literature (Yang, Chen et al. 2010, Cayman Chemical 2018, Sigma-Aldrich 2018).

2.3.10 Cell culture

The human dermal fibroblast (HDF) cell line was purchased from Life Technologies (lot

771555). The cells were maintained at 37 °C, under a 5% CO2 atmosphere in Dulbecco’s modified Eagle’s medium-high glucose (DMEM-HG) supplemented with 10 % (v/v) heat- inactivated foetal bovine serum (FBS), 2 mM L-glutamine (Life Technologies), 1 % MEM non-essential amino acids, gentamicin solution (100 μg mL-1) and amphotericin B solution (0.25 μg mL-1), Sigma Aldrich. Cells were passaged when a confluence of 70 –

80 % was reached. This process involved a treatment with 0.05 % trypsin-EDTA solution and reseeding at a concentration of 1.5 x 105 cells mL-1.

The lung cancer cell line A549 (ATCC CCL-185) was a kind gift from Dr Satyanarayana

Somavarapu (UCL School of Pharmacy). The cells were maintained at 37 °C in a 5%

-1 CO2 atmosphere in RPMI medium (Gibco) supplemented with penicillin (100 µg mL ), streptomycin (100 µg mL-1), L-glutamine (2 mM), and 10 % (v/v) heat-inactivated FBS

(all Gibco). The cells were passaged every 3 days and reseeded prior to use at a concentration of 9 x 105 cells mL-1.

For assessment of the formulations, the materials to be tested were first dissolved in complete DMEM-HG or RPMI medium as appropriate to form solutions at 1 mg mL-1.

These were filtered through a 0.22 µm filter, and cells were directly resuspended in 180

μL of each solution in a 96-well plate (Greiner Bio-One Cellstar). Cell densities were 7.5 x 104 cells mL-1 for HDF and 5.5 x 104 cells mL-1 for A549. Doses of carmofur and RB in the controls were matched to their concentrations in the single-fluid particle solutions.

The cells were incubated with the dissolved formulations for 24 h, and then irradiated at

521 nm using a microscope illuminator (DiCon LED) for 20 min (1050 mW, 0.32 cm2).

The lid of the cell plates was removed before irradiating the well plates. Control

experiments were also performed in which the cells were not exposed to light. Cell viability was determined using the CellTiter-Glo™ luminescent cell viability assay

(Promega). The luminescent reagent was prepared following the manufacturer’s instructions and added to the culture plates with a reagent volume equal to the volume of cell culture medium present in each well.

The CellTitre-Glo assay determines the viability of cells by measuring ATP levels, which are a signal of metabolically active cells. The amount of ATP is directly proportional to the number of viable cells present in the culture (Figure 2-3).

Figure 2-3. Cell Titer Glo luciferase reaction

The substrate was prepared following the manufacturer’s instructions and added to the culture plates. After addition of the reagent, the plate was equilibrated for 10 minutes.

Luminescence was then read using a SpectraMax M2e spectrophotometer (Molecular

Devices).The cell viability was calculated using Equation 2-1.

Equation 2-1. Calculations of cell viability using Cell Titer Glo

(퐹푙푢표푟푒푠푒푛푐푒 표푓 푠푎푚푝푙푒−퐹푙푢표푟푒푠푒푛푐푒 표푓 푏푎푐푘푔푟표푢푛푑) % 푣𝑖푎𝑖푙𝑖푡푦 푐표푚푝푎푟푒푑 푡표 푐표푛푡푟표푙 = × 100 (퐹푙푢표푟푒푠푒푛푐푒 표푓 푐표푛푡푟표푙−퐹푙푢표푟푒푠푒푛푐푒 표푓 푏푎푐푘푔푟표푢푛푑)

The selectivity index of the formulations was calculated using Equation 2-2:

Equation 2-2. Selectivity index formula

푣𝑖푎푏𝑖푙𝑖푡푦 표푓 퐻퐷퐹 푐푒푙푙푠 푤𝑖푡ℎ 푙𝑖푔ℎ푡 푒푥푝표푠푢푟푒 푆푒푙푒푐푡𝑖푣𝑖푡푦 𝑖푛푑푒푥 = 푣𝑖푎푏𝑖푙𝑖푡푦 표푓 퐴549 푐푒푙푠 푤𝑖푡ℎ 푙𝑖푔ℎ푡 푒푥푝표푠푢푟푒

The Cell Titre Glo assay is based on the amount of ATP produced by cells. However,

ATP does not always correlate with cell viability because non-lethal perturbations leading to reduced proliferation or inhibited mitochondrial respiration can cause metabolic interference and could thus produce false positive results. (Kepp, et al, 2011).

Alternatives assays such as the MTT assay can also be used.

2.3.11 Statistical analysis

Statistical analysis was made with the SPSS Statistics Software. Statistical significance of differences was evaluated by one-way ANOVA using Games-Howell or Bonferroni post hoc tests. The level of significance was set at probabilities of p < 0.001 (***) and p

< 0.05 (*)

2.4 Results and discussion

2.4.1 Janus fibres

Fibres fabricated with electrospinning two fluids using a side-by-side spinneret are difficult to manufacture, due to the parallel fluids repelling each other and thus separating. As PVP has been reported to be an easily spinnable polymer (Chuangchote and Sagawa 2013), it was chosen to perform the proof-of-concept of electrospinning a photo-chemotherapy formulation

After an extensive exploration of different polymers and processing parameters (such as the concentration of the polymer solution, voltage and collector distance), PVP Janus fibres were successfully produced.

2.4.2 Proof-of-concept

Single fluid electrospinning was first used to prepare blank fibres (PVP blank fibres), fibres containing hematoporphyrin (1% w/w, PVP-HP-F) and fibres loaded with RB (1% w/w PVP--RB-F). SEM images of these are given in Figure 2-4. The average diameters of the PVP blank fibres and PVP-RB-F fibres are similar at 0.587 ± 0.126 µm and 0.474

± 0.139 µm respectively. However, the PVP-HP-F fibres had an average size of 1.075

±0.255 µm, almost twice the diameter of the PVP blank fibres (PVP-F) (Figure 2-4).

Using the optimal single-fluid parameters (15 kV, 1 ml h1, 20 cm tip-to-collector distance),

Janus fibres (PVP-RB-HP-F) were created from the two separate PVP/photosensitiser solutions. These active ingredients impart a strong coloration to the solution, allowing facile visualisation of the structure. Taking advantage of the strong coloration, a visual examination of the Taylor cone was performed, and verified that a two-compartment cone was produced (Figure 2-5).

a b

PVP-HP-F c PVP-RB-F

PVP-F

Figure 2-4. SEM images of the monolithic fibres a) PVP-HP-F, b) PVP-RB-F, and c) PVP-F, prepared at 1 ml h-1, 20 cm tip-to-collector distance and a voltage of 15 kV

Figure 2-5. A photograph of the Janus Taylor cone obtained in side-by-side spinning of PVP

SEM images of the fibres from side-by-side spinning are shown in Figure 2-6. The fibres have smooth surfaces with a clear indent in the centre, and a width of 1.18 ± 0.153 µm.

The homogeneity achieved is high, and no evidence of “bead on string” morphology is seen. Fluorescence microscopy was also utilized to visualise the fibres (Figure 2-7): the sides of the fibres can be seen to have two different colours. Physicochemical characterization (XRD, DSC) demonstrate the polymer and polymer/drug formulations to all be amorphous, RB is a crystalline material while HP is amorphous (data can be found in Appendix I). The successful fabrication of the Janus fibres showed that it was possible to produce drug-loaded Janus fibres using electrospinning (this had not been reported at the time of performing these experiments). However, it is clear that doing so is non- facile and can only be achieved in favourable situations.

a b

Figure 2-6.a) A representative SEM image of PVP Janus fibres (PVP-HP-RB-F) and b) a graph showing their size distribution.

Figure 2-7. A fluorescence microscopy image of Janus fibre loaded with HP (red) and RB (green)

2.5 Photo-chemotherapeutic fibres

Having established that Janus fibres could be produced, the formulations were changed to achieve a Janus structure containing a photosensitizer and a chemotherapeutic agent.

The compound RB was selected because it is a very effective photosensitizer, which produces singlet oxygen with high yield (Fischer, Krieger-Liszkay et al. 2004) and it has been widely studied as a photodynamic agent (Panzarini, Inguscio et al. 2011). In addition, the handling of the substance is easier than hematoporphyrin as it does not require low temperature storage. The chemotherapeutic agent chosen for this formulation was carmofur.

The electrospinning conditions were maintained the same as for the PVP-HP-RB-F formulation. An SEM image of the resultant Janus system PVP-RB-C) is shown in Figure 92

2-8. The fibre size is similar to the previous Janus system, with a mean diameter of 1.282

± 0.479 µm. The fibres present a smooth morphology, without evidence any bead-on- string morphology. However, the morphology is not as defined as for the PVP-HP-RB formulation.

Figure 2-8. An SEM image of Janus fibres loaded with hematoporphyrin and carmofur (PVP-RBC-F), and the fibre size distribution

2.5.1 FTIR

To study the interactions of RB and carmofur with the polymer, FTIR spectra were recorded. The raw materials (PVP, RB and carmofur) were investigated first (Figure 2-9).

The PVP spectrum shows a distinctive C=O band at 1660 cm-1 and C-N stretch at 1290 cm-1 (Dhumale, Gangwar et al. 2012). The chemical structures of carmofur and rose

Bengal can be seen in Chapter 1 (Figure 1-4 and Figure 1-7), Carmofur has peaks between 1660-1770 cm-1 and a broad peak at 1501 cm-1 which could all be attributed to

C=O groups. The RB spectrum presents peaks attributed to vibrations of aromatic rings at around 1547, 1492, and 1448 cm-1. C–O–C groups can be observed at 1235 cm-1.

Bands at 1613 and 1339 cm-1 can be attributed to asymmetric and symmetric stretching vibrations of COO-, respectively. The C-Cl stretching vibration can be observed at 759

cm-1, and the band at 1020 cm-1 can be assigned to Na–O vibrations (Jiao, Wu et al.

2016).

Figure 2-9. FTIR spectra of the raw materials used to fabricate Janus structures

Figure 2-10 displays the spectra of the electrospun fibres. The spectra of raw PVP and the PVP fibres are identical, as would be expected. The spectra of the formulations appear to be a combination of the raw materials, with the most significant changes observed in the carboxylate region. In the PVP-RB-F, a single carbonyl band can be seen at 1652 cm-1, which means that the PVP and RB bands have merged. This suggests the presence of interactions between the two components (e.g. H- bonding).

The spectrum of PVP-RBC-F looks similar to the PVP-F presumably because of the small amounts of RB and carmofur compared with the PVP content in the system.

Figure 2-10. FTIR spectra of single fluid and Janus fibres containing carmofur and RB

2.5.2 XRD

XRD patterns were recorded for the raw materials utilized throughout this chapter (Figure

2-11). It can be observed that PVP displays only a broad halo, with no sharp peaks, suggesting it is amorphous (Dupeyrón, Kawakami et al. 2013). Meanwhile, carmofur presents a number of Bragg reflections in its pattern confirming the crystallinity of the material, as has been reported in the literature (Ren, Liu et al. 2008). RB also shows distinct reflections, and therefore is also a crystalline material.

Figure 2-11. XRD patterns of the raw materials utilised to fabricate electrospun and electrosprayed materials

The patterns from the formulations can be seen in Figure 2-12. All show broad haloes, consistent with amorphous structures. The Bragg reflections of both carmofur and RB have disappeared in the electrospun materials (PVP-C-F and PVP-RBC-F). This can be explained as solvent evaporation during electrospinning occurs very rapidly, such that it does not allow the rearrangement, nucleation and crystallization of the drug molecules, leading to the fabrication of an amorphous material (Wang, Zhao et al. 2013). This effect is widely seen in the literature, where electrospinning leads to the creation of amorphous solid dispersions (Illangakoon, Yu et al. 2015, Nagy, Balogh et al. 2015, Yu, Li et al.

2018).

Figure 2-12. X-ray diffraction patterns of the electrospun Janus fibres.

2.5.3 DSC

TGA was used to determine the appropriate temperature range for the DSC experiments

(to avoid degradation occurring in the furnace). Subsequently, DSC data were obtained, and are shown in Figure 2-13. PVP is a naturally hygroscopic polymer which tends to absorb water. When heated the water absorbed by the polymer leads to a broad water evaporation endotherm in the thermograms, making it difficult to identify whether there are any other changes in the formulation over the temperature range from 70-110 ºC. In addition, PVP carbonises at 200 ºC. Thus, for the samples containing PVP, it was decided to perform two heating cycles: 0-100 ºC and then 0-160 ºC.

As shown in the thermographs in Figure 2-13 a sharp melting point around 110-115 ºC is observed for free carmofur; it also undergoes degradation at around 150 ºC, which is in agreement with the literature (Masahiko, Fumitoshi et al. 1987). RB has a melting point

>300° C, and thus its melting cannot be seen.

Figure 2-13. DSC thermograms of the raw materials (PVP, RB, Carmofur). The second heating cycle is shown for PVP.

No melting endotherms are depicted for the drug-loaded fibres, as seen in Figure 2-14, implying carmofur is amorphous in the polymer matrix. The absence of melting endotherms and of Bragg reflections for both carmofur and RB demonstrate that the formulations comprise amorphous solid dispersions, as widely reported in the literature when using electrospinning (Haider, Haider et al. 2015, Agrahari, Agrahari et al. 2017).

Figure 2-14. DSC thermograms showing the second heating cycle of PVP-RBC-F, PVP-RB-F, PVP-C-F and PVP-F.

2.5.4 Drug release studies

The release of the drugs was explored in PBS, and the cumulative release data are plotted in Figure 2-15. PVP is a fast dissolving polymer, and hence the release of the entrapped active pharmaceutical ingredients (APIs) is expected to be rapid. In the first

60 min all the electrospun fibres release > 50% of their drug loading. PVP-RCB-F behaves similarly to the single fluid fibres: for instance, at 60 min PVP-RB-F released

72.8 ± 1.2% of its RB loading while PVP-RBC-F released 81.2 ± 12.7%. The dissolution of carmofur from the fibres is faster than the dissolution of the pure drug. This can be explained as the API exists in an amorphous state in the fibres, and amorphous materials are more soluble and have faster dissolution rates than their crystalline counterparts

(Hancock and Parks 2000). The fast release of drugs from PVP fibres has been widely reported in the literature (Papageorgiou, Bikiaris et al. 2008, Illangakoon, Gill et al. 2014).

Figure 2-15. Dissolution data for the electrospun fibres and raw APIs. Data are presented as mean ± S.D. from three independent experiments.

2.6 Janus particles

Fibres are the most common product generated by EHD techniques. However, preparing micron-sized or nanoparticles is more desirable to take advantage of the EPR effect, and it has been demonstrated that electrosprayed particles achieve a good cell uptake efficiency (Ding, Lee et al. 2005). Therefore, since PVP fibres were successfully prepared, the next step was to create Janus particles using the electrospraying method.

2.6.1 PVP particles

A 5% w/v concentration of PVP was used for electrospraying: this has previously been suggested as the optimal concentration for 40 kDa PVP (Fischer, Krieger-Liszkay et al.

2004). SEM images of blank PVP particles (Figure 2-16) showed a high degree of monodispersity, which this is very important as the size of the particles can affect cellular uptake and lead to different biological responses (Xie and Castracane 2009). The particles have smooth and uniform surfaces with no evidence of fibres and a mean average size of 1.170 ± 0.162µm.

Mean= 1.170 ± 0.162 µm 25

10 No. of particles of No.

0 0.8 1.0 1.2 1.4 Size (µm)

Figure 2-16. An SEM image of electrosprayed PVP particles and their size distribution

PVP-RB particles (PVP-RB-P) were prepared using solutions containing RB at a concentration of 0.5% w/v. The electrospraying lead to the fabrication of particles with a size of 0.37 ± 0.20 µm, as shown in Figure 2-17. Here however the particles show high polydispersity. 100

Figure 2-17.SEM images and distribution of PVP-RB-P

Particles containing carmofur (PVP-C-P) were generated using the same electrospraying conditions as for PVP-RB-P particles. The results were smooth and round particles with a size of 0.67 ± 0.15 µm, as observed in the SEM image in Figure 2-18

Figure 2-18. PVP-C-P electrosprayed particles and the size distribution

Bicompartmental particles (PVP-RBC-P) were created using carmofur at a concentration of 0.80% w/v in one compartment, and RB at 0.1% w/v in the other. SEM data are presented in Figure 2-19. The particles have a mean size of 0.607 ± 0.191 µm. In the

SEM images, the presence of two compartments is visible in most of the particles (Figure

2-19). There is no sign of any fibres, and the particles seem to have a smooth but 101

irregular shape. To verify the presence of two compartments, fluorescence microscopy was undertaken (Figure 2-20). The two different sides of the particles are clearly visible.

a b

Figure 2-19. a) An SEM image showing PVP-RBC-P and b) their size distribution

Figure 2-20. A fluorescence microscopy image showing the two compartments of PVP-RBC-P loaded with carmofur (red) and RB (green).

2.6.2 Solid state properties of Janus particles

The solid state properties for the Janus and single fluid particles, as expected, were very similar to the data obtained for the analogous fibres. The data are presented in Appendix

1. The XRD patterns and DSC data show that that PVP and the PVP-based formulations are amorphous, presenting only broad bands in their XRD patterns and no melting 102

endotherms in DSC. This is expected given the very rapid solidification from solution which occurs in electrospraying, and many studies in the literature report similar findings

(Jahangiri and Adibkia 2016). Additionally, FTIR confirms the interactions between the polymer and the drugs.

2.6.3 Drug release studies

Drug release profiles for the formulations are given in Figure 2-21. As expected for PVP- based systems, release occurs rapidly and plateaus after 250 min (ca. 4 h). This is rather slower than usually seen for PVP materials, which was due to the use of a capsule as a container for the particles in these experiments. There is a significant burst release with

>25% of the loaded drug being released within 25 min. While the release profiles of RB are essentially identical in both the monolithic and Janus systems, the formation of Janus particles appears to reduce the release extent of carmofur: The profile of the release plots is very similar, but the PVP-C-P release 95.9 ± 6.6% of the drug loading after 350 min, while the PVP-RBC-P Janus particles release only 66.8 ± 9.9% in the same time interval. RB, used here as its disodium salt, is highly soluble in water (1 mg mL−1), but carmofur is a poorly water-soluble drug (0.0273 mg mL−1).

The different release of drugs of different solubility when encapsulated in the same carrier has been previously reported. For example, Li showed that doxorubicin hydrochloride, a hydrophilic drug, reduces the release rate of curcumin a hydrophobic drug (Li, Zhu et al. 2017). Xie demonstrated that the release of hydrophobic drugs from

Janus particles is differently than the release from monophasic particles, while more hydrophilic drugs tend to behave similarly in Janus and monophasic particles (Xie, She et al. 2012). Xie suggested that the distinct release profiles can also be taken as evidence that the particles have Janus features (Xie, She et al. 2012). These findings might apply for PVP-RB-C. Attempts were made to analyse the data using the Korsmeyer-Peppas model, but it was found that this model does not provide a good fit to the experimental data. 103

Figure 2-21. Drug release of PVP-RB-P, PVP-C-P and PVP-RB-P. The particles were placed in a capsule and release undertaken using a PBS buffer (pH 7.4) medium at 37ºC and 100 rpm. Percentage of drug release was calculated using UV measurements. Data are from three independent experiments and are given as mean ± S.D.

The difference in drug release between the Janus fibres and Janus particles can be attributed to the surface area to volume ratio of the materials (Kamble, Gaikwad et al.

2016). The drug release is faster when using fibres with most formulations achieving

100% release by 120 min, which might be explained as fibres can provide a fast- dissolving drug delivery system because of the high surface area of these nanostructures and their high porosity facilitating fast-wetting surface properties (Yu, Shen et al. 2009,

Li, Kanjwal et al. 2013). Note that, for the fibres a sinker was used, meanwhile a capsule was used for the dissolution tests.

2.6.4 PCL particles

After having fabricated Janus particles made with PVP, the polymer was changed to polycaprolactone (PCL) in the hope of developing extended release formulation. Single fluid electrospraying of the blank polymer and polymer solutions containing RB and 104

carmofur were first explored. Following the protocol published by Bock (Bock, Woodruff et al. 2011) electrospraying of PCL-P was achieved successfully (PCL-P), producing microparticles with a smooth surface and an average diameter of 10.68 µm ± 1.16 µm

(Figure 2-22). No evidence of fibres is seen.

Mean= 10.679 ±1.592µm a b 30

No. of particles of No. 10

0 1 2 3 4 5 6 7 8 9 10 11 12 13 (µm) Size

Figure 2-22.a) SEM image of electrosprayed PCL-P in chloroform and b) the particle size distribution

Particles loaded with RB (PCL-RB-P) were fabricated with the same technique (Figure

2-23). The resultant materials comprise particles with a few visible fibres. The surfaces of the particles appear smooth, but the presence of some elongated particles is noted, the average size was of 7.0 ±1.13 µm.

a b

Figure 2-23.a) An SEM image and b) the size distribution of electrosprayed PCL-RB-P

105

After corroborating that PCL microparticles could be fabricated, a single fluid experiment containing carmofur was undertaken (PCL-C-P) (see Figure 2-24). The appearance and morphology of the particles were very similar to the electrosprayed PCL-P. Drug crystals are not visible on the surface of the particles, but there is some evidence of small fibres.

a b Mean = 8.801 ± 1.641 µm 30

10 No. of microspheres of No.

0 4 5 6 7 8 9 10 11 12 13 Size of microspheres (micrometers)(µm)

Figure 2-24.a) SEM images of PCL-C-P microspheres, with the b)size distribution

To determine whether side-by-side electrospraying could be achieved with PCL, hematoporphyrin and RB were utilized, again to allow visualisation. SEM images (Figure

2-25) showed that bi-compartmental particles (PCL-RB-HP-P) were successfully achieved, with an average size of 7.621 ±1.124 μm. The particles seem to have smooth surfaces, but a few very fine fibres can be noticed at the bottom of the image.

a a Mean= 7.621±1.124µm

10 No. of particles of No.

0 4 5 6 7 8 9 Size (µm)

Figure 2-25. An SEM image and the size distribution of PCL-RB-HP-P

106

After achieving the electrospraying of both photosensitizers, side by side electrospraying of PCL solutions containing a photosensitizer and a chemotherapeutic agent was performed. Despite multiple attempts, this was not successful. SEM images (Figure 2-26) show that the particles produced lack a defined morphology. It seems that the solvent is not fully evaporating, creating clumps in the product. This effect could arise from several factors. As explained in Section 1.5.1, EHD techniques are dependent on multiple factors including viscosity, surface tension, humidity and conductivity, among others (Haider,

Haider 2015). These factors can complicate the experiments, meaning that the solutions cannot be effectively electrosprayed simultaneously using the same parameters in a side-by-side spinneret.

Figure 2-26. SEM images from electrospraying PCL-RB-C-P

2.6.5 In vitro cytotoxicity assays

In vitro assays were performed with the PVP particles (fibres were excluded due to unexpected circumstances) as the creation of PCL Janus particles was unsuccessful.

The formulations were tested using human dermal fibroblasts (HDF) as a model non- cancerous cell line and the A549 lung cancer cell line. The results are shown in Figure

2-27. PVP does not present any cytotoxicity to the chosen cell lines. This is expected given that the polymer is generally regarded as safe for use in pharmaceutical products

(OECD 2001). 107

There is a decrease in cell viability (of 25-40%) in both cell lines when they are treated with RB and PVP-RB-P in the absence of light. This can be caused by the intrinsic cytotoxicity of RB(Nascimento, Baesler et al. 2013). However, much more significant cell death can be seen with A549 cells after light irradiation. This is because green light with a wavelength 520-550 nm is required to generate the cytotoxic species required to induce photodamage from the RB. The complete death of both types of cells (A549 and

HDF) is observed after they are exposed to raw carmofur, and very low viability was observed with PVP-C-P. These results were expected as carmofur has been reported to have a non-selective nature, killing both cancerous and non-cancerous cells (Realini,

Solorzano et al. 2013, Domracheva, Muhamadejev et al. 2015).

108

The Janus particles caused only small amounts of HDF cell death, with or without exposure to light. In contrast, PVP-RBC-P has low cytotoxicity when no light exposure is applied, but in the presence of light kills almost all the A549 cells. The selectivity index

(viability of HDF cells with light exposure /viability of A549 cells with light exposure) for the formulations was calculated, and the results can be seen in Table 2-1. Clearly, the

PVP-RBC formulation shows a very high level of selectivity for cancerous cells, inducing

1500-fold more cell death in A549 cells than with the non-cancerous HDF cells. This effect could be a result of the combination of photodynamic therapy and chemotherapy, which is proven to result in a significant inhibition of tumour proliferation, increased induction of apoptosis (Canti, Nicolin et al. 1998, Zhou, Zhou et al. 2014). Khdair et al. have also suggested that enhanced cytotoxicity could be correlated with improved intracellular and nuclear delivery of the two drugs (Khdair, Di et al. 2010). One of the key benefits of photodynamic therapy is its high selectivity for tumour cells as a result of the ability of the photosensitizers to accumulate in tumour tissue rather than in normal cells.

Overall, it is clear from these experiments that the PVP-RBC particles are effective in the selective killing of cancer cells.

Table 2-1. Selectivity index for the formulations containing RB, carmofur and both active ingredients together

HDF viability A549 viability Selectivity ID (%) (%) index

PVP-P 91.1 ± 12.3 91.6 ± 14.8 0.99

PVP-RB-P 92.3 ± 13.7 72.8 ± 17.6 1.27

PVP-C-P 1.2 ± 3.5 0.85 ± 5.6 1.41

PVP-RBC-P 72.4 ± 11.8 0.05 ± 7.5 1448.00

RB 1.3 ± 4.2 29.4 ± 45.5 0.04

Carmofur 1.6 ± 3.0 2.16 ± 7.8 0.74

109

2.7 Conclusions

The aim of this study was to develop a compartmentalized structure combining two cancer treatments. Initial steps towards the development of Janus structures containing a photosensitizer and a chemotherapeutic agent were made by EHD fabrication techniques. Model drugs (hematoporphyrin, RB, carmofur) and polymers (PVP, PCL) were used and the results were encouraging when using PVP. PCL did not allow the fabrication of and reproducible Janus structures, however. Characterization of the formulations by XRD and DSC indicated that the APIs were encapsulated in the PVP- based fibres or particles in an amorphous state. Drug release studies showed a two- phase release from the Janus particles, a burst release and a continuous release. In vitro cell viability tests conducted with HDF and A549 cells confirm the effect of the chemotherapeutic agent in the formulations and the activation of the photosensitizer upon exposure to light. The PVP-RBC-P formulation shows very high selectivity towards cancerous cells.

We thus demonstrate here that Janus particles prepared by electrospraying have great potential in combined photo-chemotherapy, showing high selectivity for cancerous cells.

The materials which could be successfully prepared in this work used PVP as the carrier.

This polymer dissolves very rapidly upon addition to water, and while this allowed us to obtain a rapid assessment of functional performance onward formulation will be required to develop practicable drug delivery systems, for instance by embedding the particles prepared here in a secondary polymer matrix, or by exploring an alternative technique to create Janus structures such as the self-assembly of electrospun/electrosprayed materials. This will be addressed in subsequent chapters.

110

2.8 References

Agrahari, V., V. Agrahari, J. Meng and A. K. Mitra (2017). Chapter 9 - Electrospun Nanofibers in Drug Delivery: Fabrication, Advances, and Biomedical Applications. Emerging Nanotechnologies for Diagnostics, Drug Delivery and Medical Devices. Boston, Elsevier: 189-215. Bock, N., M. A. Woodruff, D. W. Hutmacher and T. R. Dargaville (2011). "Electrospraying, a Reproducible Method for Production of Polymeric Microspheres for Biomedical Applications." Polymers 3(1): 131. Canti, G., A. Nicolin, R. Cubeddu, P. Taroni, G. Bandieramonte and G. Valentini (1998). "Antitumor efficacy of the combination of photodynamic therapy and chemotherapy in murine tumors." Cancer Letters 125(1–2): 39-44. Cayman Chemical. (2018). "Carmofur. Product description." Retrieved on 10/8/2018 from https://www.caymanchem.com/product/14243. Chuangchote, S. and T. Sagawa (2013). "Electron-Acceptor Nanomaterials Fabricated by Electrospinning for Polymer Solar Cells." Energy Procedia 34: 848-853. Dhumale, V. A., R. K. Gangwar, S. S. Datar and R. B. Sharma (2012). "Reversible Aggregation Control of Polyvinylpyrrolidone Capped Gold Nanoparticles as a Function of pH." Materials Express 2(4): 311-318. Ding, L., T. Lee and C.-H. Wang (2005). "Fabrication of monodispersed Taxol-loaded particles using electrohydrodynamic atomization." Journal of Controlled Release 102(2): 395-413. Domracheva, I., R. Muhamadejev, M. Petrova, E. Liepinsh, A. Gulbe, I. Shestakova, G. Duburs and P. Arsenyan (2015). "1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC) increases Carmofur stability and in vitro antiproliferative effect." Toxicology Reports 2: 377-383. Dupeyrón, D., M. Kawakami, A. M. Ferreira, P. R. Cáceres-Vélez, J. Rieumont, R. B. Azevedo and J. C. T. Carvalho (2013). "Design of indomethacin-loaded nanoparticles: effect of polymer matrix and surfactant." International Journal of Nanomedicine 8: 3467- 3477. Fischer, B. B., A. Krieger-Liszkay and R. I. Eggen (2004). "Photosensitizers neutral red (type I) and rose bengal (type II) cause light-dependent toxicity in Chlamydomonas reinhardtii and induce the Gpxh gene via increased singlet oxygen formation." Environmental science & technology 38(23): 6307-6313. Garbuzenko, O. B., J. Winkler, M. S. Tomassone and T. Minko (2014). "Biodegradable Janus nanoparticles for local pulmonary delivery of hydrophilic and hydrophobic molecules to the lungs." Langmuir 30(43): 12941-12949. Haider, A., S. Haider and I.-K. Kang (2015). "A comprehensive review summarizing the effect of electrospinning parameters and potential applications of nanofibers in biomedical and biotechnology." Arabian Journal of Chemistry. Hancock, B. C. and M. Parks (2000). "What is the True Solubility Advantage for Amorphous Pharmaceuticals?" Pharmaceutical Research 17(4): 397-404. Hwang, S. and J. Lahann (2012). "Differentially degradable janus particles for controlled release applications." Macromol Rapid Commun 33(14): 1178-1183. Illangakoon, U. E., H. Gill, G. C. Shearman, M. Parhizkar, S. Mahalingam, N. P. Chatterton and G. R. Williams (2014). "Fast dissolving paracetamol/caffeine nanofibers prepared by electrospinning." Int J Pharm 477(1-2): 369-379.

111

Illangakoon, U. E., D. G. Yu, B. S. Ahmad, N. P. Chatterton and G. R. Williams (2015). "5-Fluorouracil loaded Eudragit fibers prepared by electrospinning." Int J Pharm 495(2): 895-902. Jahangiri, A. and K. Adibkia (2016). "Applications of electrospinning/electrospraying in drug delivery." BioImpacts : BI 6(1): 1-2. Jiao, W., Y. Wu, G. Lu and H. Jing (2016). "Inhibition of the excited-state Rose Bengal (RB) nonradiative process by introducing DMSO for highly efficient photocatalytic hydrogen evolution." RSC Advances 6(35): 29538-29544. Kamble, R. N., S. Gaikwad, A. Maske and S. S. Patil (2016). "Fabrication of electrospun nanofibres of BCS II drug for enhanced dissolution and permeation across skin." Journal of Advanced Research 7(3): 483-489.Kepp, O., Galluzzi L., Lipinski, M., Yuan J and Kroemer G. (2011) "Cell death assays for drug discovery" Nature Reviews Drug Discovery 10: 221-237 Khan, I. U., C. A. Serra, N. Anton, X. Li, R. Akasov, N. Messaddeq, I. Kraus and T. F. Vandamme (2014). "Microfluidic conceived drug loaded Janus particles in side-by-side capillaries device." Int J Pharm 473(1-2): 239-249. Khdair, A., C. Di, Y. Patil, L. Ma, Q. P. Dou, M. P. V. Shekhar and J. Panyam (2010). "Nanoparticle-mediated combination chemotherapy and photodynamic therapy overcomes tumor drug resistance." Journal of Controlled Release 141(2): 137-144. Lee, K. J., J. Yoon and J. Lahann (2011). "Recent advances with anisotropic particles." Current Opinion in Colloid & Interface Science 16(3): 195-202. Li, H., J. Zhu, S. Chen, L. Jia and Y. Ma (2017). "Fabrication of aqueous-based dual drug loaded silk fibroin electrospun nanofibers embedded with curcumin-loaded RSF nanospheres for drugs controlled release." RSC Advances 7(89): 56550-56558. Li, X., M. A. Kanjwal, L. Lin and I. S. Chronakis (2013). "Electrospun polyvinyl-alcohol nanofibers as oral fast-dissolving delivery system of caffeine and riboflavin." Colloids and Surfaces B: Biointerfaces 103: 182-188. Mandal, S., S. Bhaskar and J. Lahann (2009). "Micropatterned fiber scaffolds for spatially controlled cell adhesion." Macromolecular rapid communications 30(19): 1638-1644. Masahiko, K., H. Fumitoshi and U. Kaneto (1987). "Improvement of oral and rectal bioavailabilities of carmofur by methylated β-cyclodextrin complexations." International Journal of Pharmaceutics 38(1–3): 191-198. Misra, A. C., S. Bhaskar, N. Clay and J. Lahann (2012). "Multicompartmental particles for combined imaging and siRNA delivery." Adv Mater 24(28): 3850-3856. Mokhtari, R. B., T. S. Homayouni, N. Baluch, E. Morgatskaya, S. Kumar, B. Das and H. Yeger (2017). "Combination therapy in combating cancer." Oncotarget 8(23): 38022- 38043. Nagy, Z. K., A. Balogh, B. Demuth, H. Pataki, T. Vigh, B. Szabo, K. Molnar, B. T. Schmidt, P. Horak, G. Marosi, G. Verreck, I. Van Assche and M. E. Brewster (2015). "High speed electrospinning for scaled-up production of amorphous solid dispersion of itraconazole." Int J Pharm 480(1-2): 137-142. Nascimento, P., K. Baesler, J. Knuever, G. Douglas, A. Anfosso, W. Weninger and N. Haass (2013). Rose Bengal-Phototoxicity versus intrinsic cytotoxicity. Journal der Deutschen Dermatologischen Gesellschaft, Wiley-Blackwell Verlag GmbH. OECD (2001). OECD Guidelines for the Testing of Chemicals, Oecd/Ocde.

112

Pang, X., C. Wan, M. Wang and Z. Lin (2014). "Strictly biphasic soft and hard Janus structures: synthesis, properties, and applications." Angew Chem Int Ed Engl 53(22): 5524-5538. Panzarini, E., V. Inguscio and L. Dini (2011). "Overview of Cell Death Mechanisms Induced by Rose Bengal Acetate-Photodynamic Therapy." International Journal of Photoenergy 2011. Papageorgiou, G. Z., D. Bikiaris, F. I. Kanaze, E. Karavas, A. Stergiou and E. Georgarakis (2008). "Tailoring the release rates of fluconazole using solid dispersions in polymer blends." Drug Dev Ind Pharm 34(3): 336-346. Perro, A., S. Reculusa, S. Ravaine, E. Bourgeat-Lami and E. Duguet (2005). "Design and synthesis of Janus micro- and nanoparticles." Journal of Materials Chemistry 15(35- 36): 3745-3760. Realini, N., C. Solorzano, C. Pagliuca, D. Pizzirani, A. Armirotti, R. Luciani, M. P. Costi, T. Bandiera and D. Piomelli (2013). "Discovery of highly potent acid ceramidase inhibitors with in vitro tumor chemosensitizing activity." Sci. Rep. 3. Ren, J., W. Liu, J. Zhu and S. Gu (2008). "Preparation and characterization of electrospun, biodegradable membranes." Journal of Applied Polymer Science 109(5): 3390-3397. Rhoades, R. A. and D. R. Bell (2012). Medical Phisiology: Principles for Clinical Medicine, Wolters Kluwer Health. Roh, K. H., D. C. Martin and J. Lahann (2006). "Triphasic nanocolloids." J Am Chem Soc 128(21): 6796-6797. Sahu, S. and S. Mohapatra (2013). "Multifunctional magnetic fluorescent hybrid nanoparticles as carriers for the hydrophobic anticancer drug 5-fluorouracil." Dalton Transactions 42(6): 2224-2231. Sigma-Aldrich. (2018). "Rose Bengal. Product description." Retrieved 23/07/2018, from https://www.sigmaaldrich.com/catalog/product/aldrich/330000?lang=en®ion=US. Sokolovskaya, E., J. Yoon, A. C. Misra, S. Brase and J. Lahann (2013). "Controlled microstructuring of janus particles based on a multifunctional poly(ethylene glycol)." Macromol Rapid Commun 34(19): 1554-1559. Walther, A. and A. H. E. Muller (2008). "Janus particles." Soft Matter 4(4): 663-668. Wang, K., X.-K. Liu, X.-H. Chen, D.-G. Yu, Y.-Y. Yang and P. Liu (2018). "Electrospun Hydrophilic Janus Nanocomposites for the Rapid Onset of Therapeutic Action of Helicid." ACS Applied Materials & Interfaces 10(3): 2859-2867. Wang, X., H. Zhao, L.-S. Turng and Q. Li (2013). "Crystalline Morphology of Electrospun Poly(ε-caprolactone) (PCL) Nanofibers." Industrial & Engineering Chemistry Research 52(13): 4939-4949. Wu, L. Y., B. M. Ross, S. Hong and L. P. Lee (2010). "Bioinspired nanocorals with decoupled cellular targeting and sensing functionality." Small 6(4): 503-507. Xie, H., Z. G. She, S. Wang, G. Sharma and J. W. Smith (2012). "One-step fabrication of polymeric Janus nanoparticles for drug delivery." Langmuir 28(9): 4459-4463. Xie, Y. and J. Castracane (2009). "High-voltage, electric field-driven micro/nanofabrication for polymeric drug delivery systems." IEEE Eng Med Biol Mag 28(1): 23-30. Yang, S., F. Guo, B. Kiraly, X. Mao, M. Lu, K. W. Leong and T. J. Huang (2012). "Microfluidic synthesis of multifunctional Janus particles for biomedical applications." Lab Chip 12(12): 2097-2102. 113

Yang, Y.-T., C.-T. Chen, J.-C. Yang and T. Tsai (2010). "Spray-Dried Microparticles Containing Polymeric Micelles Encapsulating Hematoporphyrin." The AAPS Journal 12(2): 138-146. Yu, D.-G., X.-X. Shen, C. Branford-White, K. White, L.-M. Zhu and S. A. Bligh (2009). "Oral fast-dissolving drug delivery membranes prepared from electrospun polyvinylpyrrolidone ultrafine fibers." Nanotechnology 20(5): 055104. Yu, D. G., J. J. Li, G. R. Williams and M. Zhao (2018). "Electrospun amorphous solid dispersions of poorly water-soluble drugs: A review." J Control Release. Yu, D. G., C. Yang, M. Jin, G. R. Williams, H. Zou, X. Wang and S. W. Bligh (2016). "Medicated Janus fibers fabricated using a Teflon-coated side-by-side spinneret." Colloids Surf B Biointerfaces 138: 110-116. Zhou, L., L. Zhou, S. Wei, X. Ge, J. Zhou, H. Jiang, F. Li and J. Shen (2014). "Combination of chemotherapy and photodynamic therapy using graphene oxide as drug delivery system." Journal of Photochemistry and Photobiology B: Biology 135: 7-16.

114

3. Electrohydrodynamic production of templates for the self-assembly of solid lipid nanoparticles

3.1 Introduction

Drug carrier development is of paramount importance to achieve controlled, localized and improved delivery of therapeutic agents. Solid lipid nanoparticles (SLNs) were introduced 26 years ago as an alternative carrier system to traditional colloid carriers such as liposomes, emulsions or polymeric nano/microparticles (Muller, Mehnert et al.

1995). SLNs are colloidal drug carriers with size ranging between 50-400 nm (see Figure

3-1). They differ from liposomes by dint of the substitution of a liquid lipid for a solid lipid.

SLN compositions often include fatty acids, steroids, glycerides and waxes (Almeida and

Souto 2007).

SLNs have a small size, large surface area, and they are fabricated with biocompatible ingredients. This makes them potentially advantageous over other carriers because of their small size, large surface area, good biocompatibility (due to the low toxicity of their components), high drug loading, possibility of controlled drug release and drug targeting, and potential to be applied through a wide range of administration routes. Drug-loaded

SLNs have the potential to achieve site-specific and controlled drug delivery for both lipophilic and hydrophilic drugs (Naseri, Valizadeh et al. 2015)

115

Figure 3-1. An illustration of SLNs, liposomes and polymeric particles.

One of the major disadvantages SLNs encounter is their colloidal instability. This can arise as a result of several factors, one being gelation after storage. This is accelerated with increased exposure to temperature and light (Naseri, Valizadeh et al. 2015). It has also been reported that recrystallization of the lipid fraction leads to an increase in particle size and an unpredictable gelation tendency (Naseri, Valizadeh et al. 2015).

Another problematic factor is the recognition of SLN by the reticuloendothelial system, as this means they get cleared quickly from the blood circulation (Zhang, Liu et al. 2007).

SLNs have been investigated for a wide range of applications such as ocular drug delivery (as SLN increase the bioavailability of the drug into the aqueous humour)

(Hippalgaonkar, Adelli et al. 2013), protein stability enhancement (Gaspar, Serra et al.

2017), improved delivery to the colon (as SLN can be absorbed intact through cells in the ileum and colon) (Li, Zhao et al. 2009), and the encapsulation of anti-cancer drugs

(taking advantage of the EPR effect) (Wong, Bendayan et al. 2007).

There are several methods reported in the literature for producing SLNs. These include high-pressure or high-shear homogenization, oil-in-water emulsions, solvent emulsification-evaporation, solvent injection, or the use of double emulsions (Mukherjee,

116

Ray et al. 2009). These processes can be problematic because some of them involve high temperatures, limiting the ability to process heat-labile drugs (Vasconcelos,

Sarmento et al. 2007). SLNs usually consist of a lipid core and an emulsifier that stabilizes the particle in an aqueous environment, (Bruschi, Villa Nova et al. 2016) however EHD offers a one-step and heat free process that offers advantages from other techniques whilst reducing the ingredients needed (i.e. surfactants).

In this chapter, a novel technique first described by Yu et al. (Yu, Williams et al. 2011) will be explored to generate SLNs. The technique consists of using electrohydrodynamic techniques to create a matrix which will self-assemble into SLN when added to water.

The matrix used is based on polyviynlpyrrolidone (PVP) and glyceryl tristearate (GTS), and loaded with a drug. Self-assembly of SLN is expected to occur because, upon addition to water, the strong interactions between the C=O bond of the PVP matrix and water will cause water to be taken up into the formulations. These interactions result in the swelling of the fibres and microparticles, allowing mobility of the building blocks inside the structures. Subsequently, the polymer will disentangle and dissolve and the GTS and drug can aggregate into hybrid nanoparticles in order to minimise the interactions between these hydrophobic species and the aqueous medium. The high hygroscopicity and hydrophilicity of the PVP polymer are thought to be essential for the self-assembly of the SLNs (Yu, Williams et al. 2011). A schematic representation of the process can be seen in Figure 3-2.

117

Figure 3-2. Schematic representation of the self-assembly of SLNS from electrosprayed particles

In this chapter, the encapsulated drugs will comprise indomethacin and 5-fluorouracil (5-

FU). Indomethacin is commonly used as a model for BCS II drugs, because of its low solubility in water and high permeability (ElShaer et al., 2011; Yazdanian et al., 2004). It is a non-steroidal anti-inflammatory drug (NSAID), and acts through the inhibition of cyclooxygenase 1 and 2. Indomethacin finds use in the reduction of inflammation, and the resultant pain, fever and swelling (Ricciotti and FitzGerald 2011). Indomethacin, as well as 5-FU, have been reported to have anti-colorectal cancer activity (Hull, Gardner et al. 2003, Foley, Pippin et al. 2008). The anti-neoplastic activity of IMC is not fully understood; however, there are several mechanisms of action suggested such as directly acting on cell division or modulating the destruction of cancer cells by macrophages

118

(Hull, Gardner et al. 2003). If used at the correct concentration, IMC could offer a viable and safer alternative for more aggressive chemotherapeutic agents.

5-FU is a cytotoxic drug approved by the Food and Drug Administration (FDA) for colorectal cancer; it stops DNA replication by blocking the action of an enzyme called thymidylate synthetase. If delivered to its intended site of action, it will reduce the viability of cancerous cells (Wang and DuBois 2006). Section 1.2 contains more information about these agents.

3.2 Aims and objectives:

The overarching aim of the research in this chapter was to explore a bottom-up technique to create solid lipid nanoparticles loaded with model drug carmofur. The objectives are as follows:

• Prepare and characterise composite PVP-GTS fibres and particles loaded with

different drugs and assess whether SLN self-assembly occurs upon addition to

water

• Fully characterise the resultant products, and determine whether the nature of

the template particles affects the functional performance of the self-assembled

SLNs

• Explore the development of SLNs loaded with a chemotherapy agent

3.3 Materials and methods

3.3.1 Synthetic procedures

Solutions containing indomethacin (IMC, Alfa Aesar) or 5-fluorouracil (5-FU, Sigma-

Aldrich) were prepared prior to the electrohydrodynamic (EHD) processes. These solutions were composed of the drug, polyvinylpyrrolidone (PVP, 360 kDa or 40kDa,

Sigma-Aldrich) and glyceryl tristearate (GTS, Sigma-Aldrich). Chloroform was used as

119

the solvent for IMC. In the case of 5-FU, the drug was fully dissolved in dimethylformamide (DMF) and then mixed with a PVP/GTS solution in chloroform for 10 min before electrospinning/spraying. All solvents were from Fisher Scientific. The solutions contained a final concentration of 10 % w/v PVP, 5% w/v GTS and 2.5% w/v drug.

For the fabrication process, Terumo plastic syringes (10 mL) were loaded with the polymer solutions and pumped through a stainless steel needle (gauge 20, inner diameter 0.61 mm and outer diameter of 0.91 mm). The flow rates for electrospraying and electrospinning were set to 0.5 ml/h and 1.5 ml/h, respectively. The collector consisted of a metal plate wrapped in foil of 20x30 cm. Different solutions were made for single fluid electrospinning/spraying (Table 3-1):

Solvent composition Self- Intended PVP: GTS : API Drug assembled product w/w/w ratio Chloroform DMF (v/v) (v/v) structures Fibres. 10 : 5 : 2.5 100% - SLN-IMC-F PVP-GTS-IMC-F IMC Particles 20 : 5 : 2.5 100% - SLN-IMC-P PVP-GTS-IMC-P Fibres 10 : 5 :2.5 65% 35% SLN-5FU-F PVP-GTS-5FU-F 5-FU Particles 20 : 5 : 2.5 65% 35% SLN-5FU-P PVP-GTS-5FU-P

For the self-assembly process, 10 mg of the formulations was accurately weighed, added to 10 mL of distilled water, and sonicated for 10 min. The resultant suspension was then filtered using a 0.22 µm syringe filter.

120

3.3.2 Scanning electron microscopy

The morphologies of the materials were analysed by scanning electron microscopy

(SEM), which was performed using a field emission microscope (FEI Quanta 200F) connected to a secondary electron detector. Samples were first adhered to a SEM stub with carbon-coated double-sided tape, and sputter coated with gold to render them conductive prior to measurement. Size measurements were calculated using the ImageJ software to assess at least 50 fibres or particles.

3.3.3 Transmission electron microscopy

One drop of the self-assembled SLN suspension was mixed with a drop of 1 % w/v aqueous uranyl acetate, and the resultant mixture dropped on a carbon-coated copper grid. Transmission electron microscopy (TEM) was undertaken on a JEOL KEM-2100F microscope.

3.3.4 Fourier-transform infrared spectroscopy (FTIR)

FTIR was conducted using a PerkinElmer Spectrum 100 spectrometer fitted with an ATR attachment with approximately 2 mg of the electrosprayed particles or a fibre sample of approx. 0.2 x 0.2 cm2. The scanning range was 650-4000 cm-1, with a resolution of 1 cm-1.and four scans obtained.

3.3.5 Differential scanning calorimetry

Thermograms were obtained on a Q2000 differential scanning calorimeter (DSC, TA

Instruments). Around 2-5 mg of sample was placed in a non-hermetically sealed aluminium pan (T130425, TA). The samples were heated from 0 – 110 °C at 10°C min-1

(to remove any adsorbed water), followed by cooling to 0 °C at a rate of 10°C min-1 and reheating to 200 °C at a ramp of 10 °C min-1. All experiments were performed under a

50 mL min-1 flow of oxygen-free nitrogen gas. The TA Universal Analysis software was used to analyse the data.

121

3.3.6 X-ray diffraction

X-ray diffraction (XRD) patterns were collected on a Rigaku Miniflex 600 diffractometer supplied with Cu Kα radiation (1.5418 Å) at 40 kV and 15 mA. Patterns were recorded over the 2θ range 3 – 40° at a speed of 5° min-1.

3.3.7 Dynamic light scattering (DLS)

The size of the self-assembled SLNs was quantified using dynamic light scattering (DLS) on a Zetasizer Nano ZS instrument (Malvern Instruments). Each formulation was dissolved in distilled water at a concentration of 1 mg mL-1, and a disposable polystyrene cuvette employed for sizing. The experiment was performed in triplicate, with each suspension being prepared three times. To investigate the stability of the SLNs, the suspensions were stored and further DLS measurements collected after 6 months.

3.3.8 Entrapment efficiency and drug loading

The entrapment efficiency (EE) was calculated as follows. 3 mL of the self-assembled

SLNs were loaded into a filter centrifuge tube (Amicon Ultra-15, 3000 MWCO, Merck

Millipore) and centrifuged at 9500 rpm for 10 min at 25 °C, with acceleration and brake set to 9. After centrifugation, the filtrate was recovered and analysed by UV spectroscopy

(Cary-100 UV-Vis, Agilent) at 240 nm (IMC) and 265 nm (5FU). %EE was calculated using Equation 3-3:

Equation 3-3

푤푇표푡푎푙−푊 퐸퐸% = 푓푟푒푒 푥 100 푤푇표푡푎푙

Wtotal drug represents the overall mass of drug in the EHD product, and Wfree is the mass of drug present in the supernatant.

122

The drug loading (DL) was also quantified, using Equation 3 4:

Equation 3-4.

푊 푡표푡푎푙 푑푟푢푔 %퐷퐿 = 푥 100 푊 푝푎푟푡푖푐푙푒푠

Wtotal drug represents the overall mass of drug in the formulation, and Wparticles the overall mass of the polymer and lipid.

3.3.9 Drug release studies

5 mL of the aqueous SLN suspension was transferred into a cellulose dialysis bag

(Fisher Scientific, 3500 MWCO, volume/cm=1.5). The latter was then placed in a 50 mL autoclave bottle containing phosphate buffered saline (PBS; pH 7.4 or 6.8) or fasted simulated small intestinal fluid (FaSSIF, Biorelevant) at 37 °C, and stirred at 80 rpm. 2 mL aliquots were withdrawn at certain time points and replaced with the same volume of fresh preheated media. The aliquots were filtered (0.22 µm filter) and the drug concentration quantified by UV-vis spectroscopy (Cary 100 spectrophotometer, Agilent

Technologies). Experiments were performed in triplicate and data are presented as mean ± S.D.

3.3.10 Cell culture

The colorectal adenocarcinoma cell line Caco-2 (ATCC HTB-37) was employed for in vitro studies. Cells were maintained at 37 °C, under 5 % CO2, in Dulbecco’s modified

Eagle medium (DMEM-HG; Gibco) supplemented with penicillin-streptomycin (1 % v/v) and L-glutamine (1 % v/v) solutions (Life Technologies), non-essential amino acid solution (1 % v/v, Life Technologies) and 10 % v/v heat-inactivated fetal bovine serum

(Gibco) (termed “complete DMEM”). Cells were passaged until required for further studies. The passage numbers for the viability experiments were between 26 and 30.

123

For viability tests, Biolite 24 well multidish clear plates (ThermoScientific) were used. The seeding density was 5x104 cells mL-1, and each well contained 1 mL of media. The formulations to be tested were dissolved in complete DMEM (1 mg mL-1), filtered through a 0.22 µm filter, and 180 µL of the resultant suspension was added to the wells of the plate. The cells were incubated with the dissolved formulations for 48 h. Cell viability was determined using the Alamar Blue™ cell viability assay (ThermoFisher). The reagent was prepared following the manufacturer’s instructions and added to the culture plates with a reagent volume equal to the volume of cell culture medium present in each well.

After addition, the plate was incubated for 60 minutes at 37 °C and 5 % CO2 before absorbance at 570 nm and 600 nm was read using a SpectraMax M2e spectrophotometer (Molecular Devices). The viability of the cells was calculated using the following Equation 3-5:

Equation 3-5.Calculation for % viability

(퐴 − 퐴 ) 푋 100 (푇푟푒푎푡푒푑 푐푒푙푙푠) % 푣𝑖푎푏𝑖푙𝑖푡푦 = 570 600 (퐴570 − 퐴600) (퐶표푛푡푟표푙 푐푒푙푙푠)

3.3.11 Statistical analysis

Size data obtained from DLS were analysed using a one-tail Student’s T-test. The level of significance was set at probabilities of p <0.05 (denoted *). Data from cell culture experiments are presented as mean ± S.D. and were statistically analysed using the

MiniTab17 Software. The statistical significance of differences was evaluated by one- way ANOVA using Dunnett simultaneous 95% CI tests.

124

3.4 Results and discussion

3.4.1 Optimisation

3.4.1.1 Electrospinning optimisation

The electrospinning parameters were first optimised to obtain smooth and uniform fibres.

Initial optimisation consisted of adapting the solution parameters, and various polymer: lipid ratios were explored. Firstly, a solution using a ratio of 20% PVP, 5% GTS and 2.5%

IMC was prepared, but it was observed that the viscosity of this solution was too high and no Taylor cone was achieved during electrospinning. Subsequent solutions were prepared to maintain the same ratios of GTS and IMC while changing the polymer concentration to 10%. In addition, a solution of 10% PVP, 2.5% GTS and 2.5% IMC was also electrospun. The solutions with 2.5% and 5% of GTS were both able to produce fibres (Figure 3-3).

The next step to follow during the optimisation of the electrospinning process is the flow rate: a lower flow rate helps produce uniform fibres, but it also increases the possibility of clogging the spinneret (Haider, Haider et al. 2015). A flow rate of 1 ml h-1 was found to be too low, as the solvent evaporated too quickly, and the resultant solid material blocked the spinneret. The flow rate was adjusted gradually until it resulted in a continuous jet and a stable Taylor cone, which was achieved at 1.5 ml h-1.

A critical voltage is known to exist for an electrospinning process: at this point high-quality fibres are formed, while exceeding this value results in excessively rapid evaporation of solvent and falling below it leads to insufficient solvent evaporation (Şener, Altay et al.

2011). Smooth fibres were observed for 5 % w/v GTS at 15kV. Solutions containing 2.5% w/v GTS produced beaded fibres at both 10 and 15 kV, while at 13 kV bead-free structures was observed but the fibres are still malformed. It has been noted previously that beaded nanofibres can form at both above and below the critical applied voltage, albeit due to different mechanisms (Şener, Altay et al. 2011). The 5% w/v concentration

125

of GTS, achieved more smooth and uniform fibres as seen in Figure 3-3, thus it was chosen for further investigation.

Figure 3-3. SEM images of unloaded fibres (a,b) 5%GTS, 15 Kv,, 1.5 ml/h, collector distance 20 cm (c,d) 2.5%GTS electrospun fibres, 13Kv, flow rate: 1.0 ml/h collector distance 20 cm.

The spinneret to collector distance was next optimised and ultimately set at 20 cm as distances below this resulted in a wet deposit on the collector as a result of incomplete solvent evaporation.

3.4.1.2 Electrospraying optimisation

This optimisation was based on previous work where a lower molecular weight PVP (Mw

40,000Da), was used to produce electrosprayed microparticles entrapping naproxen

(Yu, Williams et al. 2011). A solution of 20% w/v PVP, 5% GTS, 2.5% IMC or 5-FU was employed. Some experimental conditions were reproduced from Yu’s work: a flow rate 126

of 1 mL h-1 and collector distance of 25 cm. Yu utilised a voltage of 6-12 kV, however using a voltage between 6 and 10 kV resulted in neither particles nor fibres, as observed in Figure 3-4 (a). It seems that there is an agglomaration of particles or a fusing of fibres forming an undefined structure. The voltage was increased from 10 -12 kV, leading to more clearly defined particles, but there are still structures with undefined shape observed (Figure 3-4 b)

a) b)

Figure 3-4. SEM image of PVP-GTS-IMC-P using a) 7 kV at 1ml h-1 with a collector distance of 25 cm b) 11 kV at 1ml h-1 with a collector distance of 25 cm.

To try to improve the shape of the PVP-GTS-IMC particles, the voltage was increased to 15 kV and the flow rate decreased to 0.8 ml h-1, as it has been reported that higher flow rates and lower spraying voltages produce malformed particles (Wu,

MacKay et al. 2009) .Using these new parameters, spherical particles and rods were observed as observed in Figure 3-6. However, as the most important part is the self- assembly of the SLNs and the electrosprayed particles were to be used only as templates, the morphology of the particles was not felt to be of paramount importance, and no further optimisation was carried out.

127

The optimisation of the PVP-GTS-5-FU-P formulation, which included a mixture of solvents, was more challenging. The same conditions used for the IMC formulation were tried, however these parameters fabricated particles that seem to fuse, probably as a consequence of lack of solvent evaporation or solvent dripping on the existing sample (Figure 3-5 a). Thus, the flow rate was decreased to 0.5 ml h-1 and the voltage was increased to 10 kV leading to the creation of more defined particles, although the solvent evaporation problem was still present, leading to the fusing of the particles (Figure 3-5 b).

a) b)

Figure 3-5. SEM image of PVP-GTS-IMC-P using a) 6 kV at 1 ml h-1 with a collector distance of 25 cm b) 10 kV at 0.5 ml h-1 with a collector distance of 20 cm

To try to overcome the problem with solvent evaporation, the tip-to-collector distance was increased to 25 cm, allowing more time for the solvent to evaporate: this lead to the complete evaporation of the solvent, having as a result particles and rods with smooth morphology (Figure 3-6). Thus, the optimal set conditions were: 10 kV at 0.5 ml h-1 with a collector distance of 25 cm.

128

3.4.2 SEM of optimised products

SEM images of the optimised formulations for IMC fibres (PVP-GTS-IMC-F) and particles

(PVP-GTS-IMC-P) and 5-FU fibres (PVP-GTS-5FU-F) and particles (PVP-GTS-5FU-P) are shown in Figure 3-6. The PVP-GTS-IMC-F fibres are observed to be smooth and cylindrical, with an average size of 3.25 ± 0.26 μm, while for PVP-GTS-5FU-F the size is much smaller (0.15 ± 0.05 μm). Both sets of electrosprayed particles are of rods and particle morphologies. PVP-GTS-IMC-P was found to have a mean size of 1.63 ± 0.43

μm, similar to PVP-GTS-5FU-P.with 1.35 ± 0.5 μm.

Figure 3-6. SEM images of the optimal products from a) electrospinning of PVP-GTS-IMC b) electrospraying of PVP-GTS-IMC, c) electrospinning of PVP-GTS-5FU, d) electrospraying of PVP- GTS-5FU. Conditions used for electrospinning 5FU and IMC: 13Kv, flow rate: 1.0 ml h-1 collector distance 20 cm. Conditions for elecrospraying IMC: 15 kV at 0.8 ml h-1 with a collector distance of 25 cm. Conditions used for electrospraying 5FU: 10 kV at 0.5 ml h-1 with a collector distance of 25 cm

129

3.4.3 Characterisation of PVP-GTS-5FU and PVP-GTS-IMC

formulations

3.4.3.1 Fourier transform infrared spectroscopy (FTIR)

FTIR was used to infer the possible molecular interactions between the components.

The chemical structures of the raw materials are given in Figure 3-7, and the IR spectra are presented in Figure 3-8.

a) b)

c) d)

Figure 3-7. Chemical structures of a) IMC b) GTS c) 5-FU, d) PVP.

The PVP spectrum is characteristic in having two weak and broad bands between 3200

– 3600 cm-1 and 2800-3000 cm-1, indicating O-H and C-H stretching. The O-H stretching is believed to be due to the water adsorbed by the polymer, as it has a hygroscopic nature. The peak at 1660 cm-1 corresponds to the C=O amide stretching.

In the indomethacin spectrum, a triplet of medium intensity peaks between 1580-1620 cm-1 corresponds to aromatic C=C stretching. The C=O stretching vibration can be seen as a strong peak at 1680 cm-1. At 1261 cm-1 there is an asymmetric aromatic O-C stretch, and symmetric aromatic O-H stretching arises at 1086 cm-1. A weak broad band between

2800 - 3000 cm-1 corresponds to C-H stretching (Taylor and Zografi 1997). For 5-FU, the

130

characteristic bands are shown in the 1429-1600 cm-1 region, corresponding to the C=N and C=C ring stretching vibrations. The C=O band can be seen at 1720 cm-1, while vibrations from the pyridine compound occur at around 1345 cm-1. The N-H stretch was assigned to the broad band seen at 2989 cm-1.

The spectrum of GTS shows two strong peaks at 2917 and 2850 cm-1, as well as a

-1 shoulder at 2960 cm ; all of these arise from C-H stretching. C=O stretching from the ester group can be observed in a strong peak at 1740 cm-1, while C-O vibrations can be seen in the broad range of peaks between 1100 – 1300 cm-1.

Figure 3-8. FTIR spectra of the raw materials used for formulation development

In the spectra of the IMC-loaded formulations (Figure 3-9) twin peaks at 2917 cm-1 and

2850 cm-1, corresponding to the C-H stretching of GTS, can be observed. Major changes can be seen in the carboxylate region of the spectra of the indomethacin loaded

131

formulations, where the characteristic bands of indomethacin are shifted to lower wavenumber values and overlap with the carbonyl band of PVP at 1660 cm-1. This indicates the presence of intermolecular interactions between the drug and polymer.

Additionally, two peaks can be seen at 2850 and 2920 cm-1 which indicate stretching of several C-H bonds formed between GTS and PVP.

The 5-FU spectra show clear evidence for the polymer and lipid material, but the characteristic 5-FU bands at around 3000 cm-1 are not visible. This might arise due to the weak intensity of the band, or could be the result of hydrogen bonding with PVP giving a broader and less visible band. FTIR confirmed the interaction between PVP-

GTS-IMC, however from the 5-FU formulations, FTIR just provides evidence of the interaction of the PVP and GTS. Due to the high complexity of the spectra, it is difficult to obtain sufficient information from the particles and fibres components from FTIR.

Figure 3-9. FTIR spectra of the fabricated materials

132

3.4.3.2 DSC

The physical form of the formulations was first investigated using DSC; the thermograms of the raw material can be found in Figure 3-10. The thermograms for 5-FU and IMC show that they are both crystalline materials. IMC displays a sharp endotherm at 161 °C due to the melting of the drug. Similarly, 5-FU has a sharp endothermic peak at 284 °C corresponding to melting. Both values agree with the literature (Tummala, Satish Kumar et al. 2015). In the case of GTS, there are two endotherms (49 oC, 60 oC) and one exothermic (51.2 oC) peak. The first endotherm indicates the melting of α-GTS, followed by recrystallisation to β–GTS (Singh, Jalali et al. 1999), which leads to the exothermic peak. The second endotherm thus represents the melting of β–GTS. PVP is amorphous as no melting peaks can be observed in the DSC data. In all cases, the use of a pre-heat means that there are no residual solvent peaks in the thermograms.

Figure 3-10. DSC of the raw materials.

133

The thermograms for the fibres and particles can be found in Figure 3-11. PVP carbonises at temperatures above 200 °C and thus DSC data could not be obtained at higher temperatures for the formulations. Therefore, it is not possible to detect from DSC whether 5-FU is present in the crystalline form in the drug-loaded formulations. The sharp endotherm at 49 OC observed in all formulations correlates to the melting of β-GTS seen in Figure 3-10; GTS is thus incorporated into the electrospun fibres and electrosprayed particles in a semi-crystalline form. No melt for IMC is present in the thermograms for the fibres or particles, confirming the presence of amorphous iMC.

Figure 3-11. DSC thermograms for the formulations loaded with IMC and 5FU.

3.4.3.3 XRD

XRD patterns are shown in Figure 3-12, 5-FU and IMC exhibit sharp Bragg reflections due to their crystalline physical form. PVP does not present any diffraction peaks as it is an amorphous polymer, its XRD pattern can be found in Chapter 2. The presence of a broad peak at 21.4o shows semi-crystallinity of the GTS powder, matching literature XRD

134

patterns for the α-polymorph (Amasya, Badilli et al. 2016). This agrees with the findings from DSC, where α-GTS is noted at low temperatures.

Figure 3-12. XRD of the raw materials used to create polymer / drug formulations.

As expected, there is a total lack of sharp Bragg reflections from the drugs in the XRD patterns of the fibres and particles loaded with IMC and 5-FU (Figure 3-13). This confirms the DSC data, showing these systems contain the active ingredients in the amorphous form. The patterns primarily display a broad double-hump, similar to that observed in the pattern of pure PVP. They also include a large, broad, peak at 21.63o which is consistent with the presence of the semi-crystalline GTS. This is to be expected, as PVP and GTS are the main constituent of the particles and fibres. The XRD patterns of the fibres and particles are consistent with the work of Yu et al, which also showed PVP/GTS particles to have one or two broad halos, similar to the raw materials (Yu, Williams et al. 2011).

135

Figure 3-13. XRD patterns of the electrospun and electrosprayed materials

3.4.4 Solid lipid nanoparticle characterisation

3.4.4.1 TEM

The morphology of the self-assembled SLNs was visualised using TEM. The images of the obtained SLNs are shown in Figure 3-14. The images clearly show the presence of spherical vesicles.

Figure 3-14. TEM images of SLNs self-assembled from electrospun/electrosprayed materials

136

The particle size was measured using the ImageJ software, and the diameter of the SLNs seems to be <100 nm for all the samples (Table 3-2). It can also be noted that the standard deviation in the sizes of SLN-5FU-F, SLN-GTS-PVP-F and SLN-GTS-PVP-P is very high: this is a result of the number of particles in each image being rather low, and therefore caution must be taken in drawing firm conclusions from these data.

The hydrodynamic size of the SLNs from the electrospun and electrosprayed materials was characterised by DLS. The polydispersity index (PdI) was also determined. This is calculated from a simple two-parameter fit of the data and has a value between 0 and 1

(Malvern 2011). Colloidal systems with very narrow size distributions have values closer to 0 (i.e. are monodisperse) and samples with very broad size distributions have values closer to 1 (i.e. are polydisperse). The results are shown in Table 3-2. The mean diameters calculated using DLS are significantly higher than those measured by TEM.

This is reported to be for two possible reasons: the agglomeration of nanoparticles in solution at high concentrations (Domingos, 2011), or the hydrodynamic diameter effect.

It is assumed that a hydration layer surrounds the particles in aqueous formulations, producing the increased sizes noted in DLS compared to TEM (where the samples were dehydrated). (Malvern 2011)

Table 3-2. Particle sizes of the SLNs as determined from TEM and DLS.

Size (nm), Z-average (nm), Formulation PDI TEM DLS 33.1 ± 8.6 SLN-IMC-F 121.5 ± 30.7 0.5 ± 0.0 (n=41)

27.7 ± 8.0 SLN-IMC-P 143.6 ± 0.6 0.1 ± 0.0 (n=68)

93.9 ± 53.2 SLN-5FU-F 88.1± 6.1 0.5 ± 0.1 (n=24)

137

26.0 ± 7.1 SLN-5FU-P 143.9 ± 0.7 0.1 ± 0.1 (n=49)

91.5 ± 43.4 SLN-GTS-PVP-F 127.5 ± 17.2 0.7 ± 0.3 (n=4)

SLN-GTS-PVP- 55.5 ± 21.0 98.3 ± 3.4 0.3 ± 0.0 P (n=19)

The PDIs of the SLNs imply than those assembled from the fibres are more polydisperse than the ones assembled from the particles. The sizes of the SLNs are also smaller than is generally reported in the literature. For example, SLNs fabricated by the hot- homogenization method and loaded with indomethacin were reported to have a size of

226 ± 5 nm with a PDI of 0.17 (Balguri, Adelli et al. 2016), while drug-loaded SLNs prepared by ultrasonication had a size of 216 nm (Castelli, Puglia et al. 2005). In the case of 5-FU, Yassin et al. reported SLN with a size between 258 nm and 2 µm using the double emulsion-solvent evaporation technique, and found the size was dependant on the lipid composition (Yassin, Anwer et al. 2010). A size range of 137 ± 5.5 nm to 800

± 53.6 nm was noted for 5-FU-loaded SLNs fabricated with a solvent injection- lyophilization procedure by Khallaf (Khallaf, Salem et al. 2016).

3.4.4.2 Entrapment efficiency

The results of the DLS and TEM show that SLNs are self-assembling from the drug- loaded polymer fibres and particles; however, these techniques give no information regarding whether the drug is present in the SLNs. Therefore, the entrapment efficiency was calculated for both particles and fibres (see Table 3-3). For this experiment, the SLN suspension was filtered by centrifugation using a centrifugal filter unit with a molecular weight cut off of 3000 Da. The filter does not allow the SLNs to pass through, and the free drug present in the supernatant can be quantified to determine EE.

Table 3-3. Results of encapsulation efficiency experiments. Values are shown as mean ± standard deviation (n = 3).

138

Entrapment Efficiency Formulation (%)

SLN-IMC-F 90.2 ± 0.6

SLN-IMC-P 89.4 ± 0.9

SLN-5FU-F 88.2 ± 2.1

SLN-5FU-P 89.5 ± 0.9

The encapsulation of hydrophilic drugs can be challenging and in some studies, the incorporation of the drugs can be as low as 15% (Wang, Wang et al. 2010). Table 3-3 shows that the SLNs assembled from the electrospun/electrosprayed materials have a high encapsulation efficiency and are highly reproducible. 88.2 ± 2.1% is the lowest EE observed, and there are no clear differences between the IMC and 5-FU formulations.

These results show great improvement from previous work done by the group preparing self-assembled drug-loaded liposomes, where encapsulation efficiencies for 5-FU of 47

– 60 % were observed (Askin 2015). These also show slightly improved encapsulation to similar studies utilising a PVP matrix and GTS as a lipophilic carrier (88 ± 2%) (Yu,

Williams et al. 2011).

The EHD technique used to create the templates does not appear to have any effect on the entrapment efficiency of the SLN. Where the theoretical drug loading of the fibres was of 16.6 %, while of the particles it was of 10%.

3.4.4.3 In-vitro drug release

Drug release was studied for the self-assembled SLNs. The data obtained for IMC are depicted in Figure 3-15. For the IMC formulations, the SLNs assembled from the fibres seem to have a slightly lower release rate than the ones from electrosprayed particles.

A burst release is seen in the first 7 h for both formulations, but this slows with time. At

48 h there is an 11.2% difference in the release extents. SLN-IMC-F has a release extent

139

of 77.5 ± 3.3 % while SLN-IMC-F reached 66.3 ± 5.4% release at this time point. Such a biphasic drug release pattern (the release is divided in two phases, fast/immediate release and sustained release) is commonly reported by many research groups working with SLNs (Liu, Gong et al. 2007, Jain and Jain 2008).

Figure 3-15. In vitro release from the optimized IMC SLN formulations. Data are shown as mean ± S.D. from three independent experiments

In contrast, from Figure 3-16 it is clear that the vast majority of the 5-FU is released from

SLN-5FU-P during the first 6 h. This suggests that the SLNs are not having a significant effect on the release of 5-FU, as they seem to have a similar dissolution profile to the raw 5-FU. The SLN assembled from the fibres (SLN-5FU-F) extent of release appear to be lower than the particles. However, a bust release is still seen in the first 5 h: this can potentially be attributed to the rapid dissolution of drug molecules present at the surface layer of the particles. The release rate slows after the first 5 h and a constant extent of release was observed until 48 hours. 140

Figure 3-16. In vitro release from the optimized 5-FU SLN formulations. Data are shown as mean ± S.D. from three independent experiments.

Overall, it seems that the drug release from the SLN assembled from the fibres is slower than for those obtained from the electrosprayed particles. This may be because of the difference on the molecular weight used for each of the materials; a high-molecular weight of PVP was used to form fibres (360 kDa PVP), while a lower molecular weight

(40 kDa) was used to produce the particles. This means that the fibres have longer polymer chains, leading to a greater extent of entanglement which might result in slower dissolution of the polymer matrix, giving more time for the SLN to self-assemble and more densely packed SLNs. The size and PDI of the SLNs from the fibres are smaller than the particles (see Table 3-2) which supports this hypothesis. 141

3.4.4.4 Cell culture studies

All the drug-loaded formulations were tested on colon cancer cells (Caco-2) and viability quantified using the Alamar® blue assay. The viability of cells in the previous chapter was tested using the CellTiter-Glo™ luminescent cell viability assay, purchased from

Promega UK as RB interferes with the Alamar Blue signal. However as neither RB nor

HP were employed in the formulations of this chapter, viability was quantified using the simpler Alamar® blue assay. This monitors the reducing environment of living cells. It uses resazurin, a water-soluble, non-toxic and membrane-permeable compound.

Resazurin is added to the cell suspension and is able to freely diffuse into cells. If a cell is viable there will be a reducing environment within the cell which converts resazurin (a non-fluorescent compound) into a fluorescent compound, resorufin (Figure 3-17).

Resorufin has an excitation wavelength of 555 nm and an emission wavelength of 585 nm Figure 3-17 (Rampersad 2012).

Figure 3-17. The Alamar blue reaction, involving the reduction of resazurin to resorufin

The cytotoxicity of the formulations is presented in Figure 3-18 and Figure 3-19. The unloaded SLNs (vehicle) do not seem to have a major effect on cell viability, which is expected as PVP and GTS are known to have low cytotoxicity (Nair 1998, Santa Cruz

Biotechnology 2015).

For indomethacin, SLN-IMC-F give a cell viability of 87.7 ± 3.8%, higher than the viability of 63.7 ± 15.0% obtained from SLN-IMC-P (Figure 3-18). The latter is not significantly different from IMC alone. The increased cytotoxicity of the fibres can be attributed to the 142

difference observed in the drug release rate: this is slower with the fibers than the particles, which could translate to less cytotoxicity of the materials. The size of the SLN can also have an effect on the cytotoxicity with an inverse relationship between toxicity and size (Kong, Seog et al. 2011).

SLNs loaded with 5-FU appear to have similar cytotoxicity (Figure 3-19), whether they were self-assembled from fibres or particles, with viabilities of 59.3 ± 8.0% and 64.7 ±

9.4% for SLN-5FU-F and SLN-5FU-P respectively. The reduction in viability of Caco-2 cells when they are exposed to 5-FU-SLNs can be compared with previously published results, where a reduction of 33% in cell viability was noted (Patel, Lakkadwala et al.

2014). The cytotoxic effect observed here is not significantly different than the drug alone, which shows that the SLNs are as effective in causing the death of cancerous cells as raw 5FU.

143

3.5 Conclusions

In this chapter, a polymer matrix template was created via electrohydrodynamic techniques and used to create solid lipid nanoparticles with entrapped APIs. The fabricated fibres and particles were characterised, XRD, DSC, confirmed that indomethacin and 5FU were incorporated in an amorphous state.

Drug-loaded SLNs were successfully assembled from the template materials. The entrapment efficiency of all formulations was >82%. The fabrication method

(electrospraying or spinning) did not seem to have any major differences in the assembly of SLNs, but the optimisation of the fabrication process was more challenging when electrospraying. It also appears that the SLN assembled from the fibres have a smaller size and a slower release of the drug than those from the particles. The drug release from the SLN assembled from the fibres appear to be slower than the drug release from the particles. The cytotoxicity of the SLNs is comparable to the one obtained from the pure drug. Overall, we can conclude that the single self-assembled solid lipid nanoparticles have promise for the treatment of cancers. 144

3.6 References

Almeida, A. J. and E. Souto (2007). "Solid lipid nanoparticles as a drug delivery system for peptides and proteins." Adv Drug Deliv Rev 59(6): 478-490.

Amasya, G., U. Badilli, B. Aksu and N. Tarimci (2016). "Quality by design case study 1: Design of 5-fluorouracil loaded lipid nanoparticles by the W/O/W double emulsion — Solvent evaporation method." European J of Pharmaceutical Sciences 84: 92-102.

Askin, S. (2015). Liposomal Drug Delivery Systems from Electrospun Polymer Nanofibres, UCL School of Pharmacy: 1-44.

Balguri, S. P., G. R. Adelli and S. Majumdar (2016). "Topical ophthalmic lipid nanoparticle formulations (SLN, NLC) of indomethacin for delivery to the posterior segment ocular tissues." European J of Pharmaceutics and Biopharmaceutics 109: 224-235.

Bruschi, M. L., M. Villa Nova and P. M. Outuki (2016). 12 - Relevant Aspects on Peptide Delivery from Nanostructured Therapeutic Systems. Nanoarchitectonics for Smart Delivery and Drug Targeting. A. M. Holban and A. M. Grumezescu, William Andrew Publishing: 315-344.

Castelli, F., C. Puglia, M. G. Sarpietro, L. Rizza and F. Bonina (2005). "Characterization of indomethacin-loaded lipid nanoparticles by differential scanning calorimetry." Int J Pharm 304(1): 231-238.

Foley, P. J., J. A. Pippin, C. J. Smolock and J. A. Drebin (2008). "Growth inhibitory effects of indomethacin on human colon cancer cells are mediated, in-part, by effects on beta- catenin." Journal of the American College of Surgeons 207(3, Supplement): S99.

Gaspar, D. P., C. Serra, P. R. Lino, L. Goncalves, P. Taboada, C. Remunan-Lopez and A. J. Almeida (2017). "Microencapsulated SLN: An innovative strategy for pulmonary protein delivery." Int J Pharm 516(1-2): 231-246.

Haider, A., S. Haider and I.-K. Kang (2015). "A comprehensive review summarizing the effect of electrospinning parameters and potential applications of nanofibers in biomedical and biotechnology." Arabian Journal of Chemistry.

Hippalgaonkar, K., G. R. Adelli, K. Hippalgaonkar, M. A. Repka and S. Majumdar (2013). "Indomethacin-Loaded Solid Lipid Nanoparticles for Ocular Delivery: Development, Characterization, and In Vitro Evaluation." Journal of Ocular Pharmacology and Therapeutics 29(2): 216-228.

Hull, M. A., S. H. Gardner and G. Hawcroft (2003). "Activity of the non-steroidal anti- inflammatory drug indomethacin against colorectal cancer." Cancer Treatment Reviews 29(4): 309-320.

Jain, A. and S. K. Jain (2008). "In vitro and cell uptake studies for targeting of ligand anchored nanoparticles for colon tumors." Eur J Pharm Sci 35(5): 404-416.

Khallaf, R. A., H. F. Salem and A. Abdelbary (2016). "5-Fluorouracil shell-enriched solid lipid nanoparticles (SLN) for effective skin carcinoma treatment." Drug Delivery 23(9): 3452-3460.

Kong, B., J. H. Seog, L. M. Graham and S. B. Lee (2011). "Experimental considerations on the cytotoxicity of nanoparticles." Nanomedicine (London, England) 6(5): 929-941. 145

Li, H., X. Zhao, Y. Ma, G. Zhai, L. Li and H. Lou (2009). "Enhancement of gastrointestinal absorption of quercetin by solid lipid nanoparticles." J Control Release 133(3): 238-244.

Liu, J., T. Gong, C. Wang, Z. Zhong and Z. Zhang (2007). "Solid lipid nanoparticles loaded with insulin by sodium cholate-phosphatidylcholine-based mixed micelles: Preparation and characterization." International Journal of Pharmaceutics 340(1): 153- 162.

Malvern. (2011). "Dynamic Light Scattering." Retrieved 30/07/2018, from http://www.biophysics.bioc.cam.ac.uk/wp- content/uploads/2011/02/DLS_Terms_defined_Malvern.pdf.

Mukherjee, S., S. Ray and R. S. Thakur (2009). "Solid Lipid Nanoparticles: A Modern Formulation Approach in Drug Delivery System." Indian Journal of Pharmaceutical Sciences 71(4): 349-358.

Muller, R. H., W. Mehnert, J. S. Lucks, C. Schwarz, A. Zur Muhlen, H. Weyhers, C. Freitas and D. Ruhl (1995). "Solid lipid nanoparticles (SLN) - An alternative colloidal carrier system for controlled drug delivery." European Journal of Pharmaceutics and Biopharmaceutics 41(1): 62-69.

Nair, B. (1998). "Final Report On the Safety Assessment of Polyvinylpyrrolidone (PVP)." International Journal of Toxicology 17(4_suppl): 95-130.

Naseri, N., H. Valizadeh and P. Zakeri-Milani (2015). "Solid Lipid Nanoparticles and Nanostructured Lipid Carriers: Structure, Preparation and Application." Advanced Pharmaceutical Bulletin 5(3): 305-313.

Patel, M. N., S. Lakkadwala, M. S. Majrad, E. R. Injeti, S. M. Gollmer, Z. A. Shah, S. H. Boddu and J. Nesamony (2014). "Characterization and evaluation of 5-fluorouracil- loaded solid lipid nanoparticles prepared via a temperature-modulated solidification technique." AAPS PharmSciTech 15(6): 1498-1508.

Rampersad, S. N. (2012). "Multiple applications of Alamar Blue as an indicator of metabolic function and cellular health in cell viability bioassays." Sensors (Basel) 12(9): 12347-12360.

Ricciotti, E. and G. A. FitzGerald (2011). "Prostaglandins and Inflammation." Arteriosclerosis, thrombosis, and vascular biology 31(5): 986-1000.

Santa Cruz Biotechnology. (2015). "Glyceril tristearate- Safety Data Sheet." Retrieved 10/09/2018, from http://datasheets.scbt.com/sc-215094.pdf.

Şener, A. G., A. S. Altay and F. Altay (2011). Effect of voltage on morphology of electrospun nanofibers. 2011 7th International Conference on Electrical and Electronics Engineering (ELECO).

Singh, S. K., A. F. Jalali and M. Aldén (1999). "Modulated temperature differential scanning calorimetry for examination of tristearin polymorphism: I. Effect of operational parameters." Journal of the American Oil Chemists' Society 76(4): 499-505.

Taylor, L. S. and G. Zografi (1997). "Spectroscopic characterization of interactions between PVP and indomethacin in amorphous molecular dispersions." Pharm Res 14(12): 1691-1698.

146

Tummala, S., M. N. Satish Kumar and A. Prakash (2015). "Formulation and characterization of 5-Fluorouracil enteric coated nanoparticles for sustained and localized release in treating colorectal cancer." Saudi Pharmaceutical Journal : SPJ 23(3): 308-314.

Vasconcelos, T., B. Sarmento and P. Costa (2007). "Solid dispersions as strategy to improve oral bioavailability of poor water soluble drugs." Drug Discovery Today 12(23– 24): 1068-1075.

Wang, D. and R. N. DuBois (2006). "Prostaglandins and cancer." Gut 55(1): 115-122.

Wang, T., N. Wang, Y. Zhang, W. Shen, X. Gao and T. Li (2010). "Solvent injection- lyophilization of tert-butyl alcohol/water cosolvent systems for the preparation of drug- loaded solid lipid nanoparticles." Colloids Surf B Biointerfaces 79(1): 254-261.

Wong, H. L., R. Bendayan, A. M. Rauth, Y. Li and X. Y. Wu (2007). "Chemotherapy with anticancer drugs encapsulated in solid lipid nanoparticles." Adv Drug Deliv Rev 59(6): 491-504.

Wu, Y., J. A. MacKay, J. R. McDaniel, A. Chilkoti and R. L. Clark (2009). "Fabrication of elastin-like polypeptide nanoparticles for drug delivery by electrospraying." Biomacromolecules 10(1): 19-24.

Yassin, A. E. B., M. K. Anwer, H. A. Mowafy, I. M. El-Bagory, M. A. Bayomi and I. A. Alsarra (2010). "Optimization of 5-flurouracil solid-lipid nanoparticles: a preliminary study to treat colon cancer." International Journal of Medical Sciences 7(6): 398-408.

Zhang, J., J. Liu, X. Li and B. Jasti (2007). "Preparation and characterization of solid lipid nanoparticles containing silibinin." Drug delivery 14(6): 381-387.

147

4. Scaling up production of EHD-generated sacrificial templates

4.1 Introduction

The scale-up of a technique is of paramount importance for taking a laboratory-based process to an industrial environment. Electrohydrodynamic techniques have been investigated for the production of materials with applications in fields ranging from textiles to drug delivery and regenerative medicine (Chakraborty, Liao et al. 2009). The wide range of potential applications has led to the development of high-throughput procedures to meet clinical and industrial requirements (Persano, Camposeo et al. 2013). A standard benchtop electrospinning laboratory set-up produces approximately 0.1 – 1.0 g h -1, which translates to a maximum of around 10g of fibres a day (Krogstad and Woodrow

2014); for electrospraying, the highest production rate on a benchtop set-up is around

30 ml h-1. These volumes are substantially below the amounts required for industrial use.

Multi-needle instruments can raise production values to 2 tons a year of nanofibers and

> 1 kg/h of powder-based products from electrospinning and electrospraying, respectively (Bioinicia 2018). Many other scale-up techniques exist, utilising a variety of approaches to increase nanofibre and nanoparticle production rates.

One of the most elementary steps taken towards increasing nanofibre and nanoparticle production volumes is the use of multiple spinnerets to increase throughput (Rulison and

Flagan 1993, Niu, Lin et al. 2009, Lhernould and Lambert 2011). The needles are arranged in a linear or circular array (Figure 4-1). This arrangement can also allow for

148

the production of nanofibre mats comprised of different materials (Persano, Camposeo et al. 2013).

The main problem encountered with multi-needle arrays arises from the repulsion between the parallel electrical fields surrounding the polymer jets, requiring adequate distancing between the needles to overcome these issues (Teo and Ramakrishna 2006,

Niu, Lin et al. 2009). The distance between the arrays can also impact the electric field surrounding the emitters (Rulison and Flagan 1993, Lhernould and Lambert 2011). Kim has described an approach utilising an extra-cylindrical electrode surrounding the multi- needle array to stabilise and reduce the repulsive forces between the polymer jets (Kim and Kim 2006). Equipment such as Yflow® multi-nozzle devices (Figure 4-2) is commercially available for large scale production of nanoparticles and nanofibres.

Figure 4-2. Equipment for the large scale production of fibres and particles by electrohydrodynamic techniques. Images reproduced with permission from http://www.yflow.com/services/custom_devices/

149

One consideration in the scalability of electrospinning is its capacity for continuous processing, which is dependent on the geometry of the collector. The flat, stationary, collectors found in basic laboratory set-ups, and the rotating collectors used for uniaxial alignment of nanofibres (Teo et al., 2011), result in batch processes, due to the eventual loss of grounding as the thickness of the fibre mat increases. For effective scale-up, a moving collector such as a conveyer belt is required, allowing for continuous processing

(Persano, Camposeo et al. 2013, Shuai, Yubo et al. 2014).

Needleless methods for the electrospinning of polymer fibres were first reported in 1979, but it has only been in the last 12 years that their potential for scaling-up the production of nanofibres has been capitalised on with the development of the Nanospider by

Elmarco (Haitao and Tong 2012). The Nanospider design first utilised a rotating cylindrical spinneret partially submerged in polymer solution, with the collector located above; multiple Taylor cones form on the surface of the rotating cylinder as the electrical field is applied and the polymer jet is drawn up towards the collector, where the nanofibres are deposited (Jirsak, Saternik et al. 2005). The Nanospider is reported to be able to produce over 50 million m2 of non-woven nanofibres annually (Persano,

Camposeo et al. 2013). Although the Nanospider equipment is designed for the production of fibres, electrospraying can also be achieved by controlling the polymer solution parameters, such as molecular weight and concentration (Böttjer, Grothe et al.

2018).

Other groups moved to develop the spinneret geometry into rotating discs, spiral coils, balls and beaded chains, which are reported to confer further advantages over the cylindrical spinneret (Niu and Lin, 2012). Niu indicated that the rotating disc showed a greater rate of fibre production over the cylinder, due to the presence of a more intense electrical field along the disc edge; the spiral coil provided advantages in greater control over fibre morphology and further improved production rates (Niu and Lin, 2012). On the other hand, Parhizkar demonstrated that a circular plate configuration was easier to 150

control and achieved higher size uniformity than the linear array (Parhizkar, Reardon et al. 2017).

High throughput spinning has seen further advancements with the very recent advent of high speed, or corona, electrospinning, developed by Molnar and Nagy to overcome the drawbacks of previous needleless methods. In particular, they sought to minimise the exposed surface of the polymer solution to make it easier to use highly volatile solvents and minimise possible water vapour absorption (Molnar and Nagy 2016). Nagy and his group have utilised this approach to produce amorphous itraconazole loaded Kollidon® fibres as a solid dispersion at a rate of 450 g h-1 (Nagy, Balogh et al. 2015).

Alternative nanofibre production techniques include alternating current electrospinning

(ACES) and centrifugal spinning. In ACES, the alternating current applied to the spinneret produces an electric wind which draws fibres out of the Taylor cone in a fibrous plume; flow rates of 10-40 ml h-1 have been achieved (Balogh, Cselkó et al. 2015).

Centrifugal spinning is a non-electrical process whereby fibres are formed through the centrifugal force produced by a high-speed rotating spinneret; a basic lab setup is capable of producing 50 g h-1, compared to the 0.1-1.0 g h-1 of basic ES (Zhang and Yu

2014).

An alternative to using electrical energy to evaporate a solvent and yield a solid product is to use heat to this end, using approaches such as spray drying. This is a technique used widely in the pharmaceutical and food industry. The industrial application of the spray drying technique began in 1920, with the spray drying of milk and detergent (Parikh

2008). However, the first patent regarding spray drying was in 1994 (Keshani, Daud et al. 2015). There are an estimated of 25,000 spray drying facilities being used for commercial purposes, from agrochemical to pharmaceutical products, with capacities from a few kg/h to over 50 tons/h evaporation capacity (Mujumdar 2014). In the pharmaceutical industry there are several products produced by spray drying (e.g.

151

penicillin, blood products, enzymes, vaccines, etc) (Jain, Ganesh et al. 2012, Kanojia,

Have et al. 2017).

Spray drying offers advantages such as high tonnage production in continuous operation and relatively simple equipment. It can be optimised to produce uniform spherical particles. However, disadvantages include high installation costs, and a lack of flexibility in terms of the fabrication of different type of products in the same unit (e.g. a spray drying unit that produces fine atomization cannot produce a coarse product and vice versa) (Mujumdar 2014).

4.2 Aims and objectives

The aim of this chapter is to investigate whether current scale-up techniques are suitable for the production of SLN-forming nanofibres and nanoparticles. The formulations first developed in Chapter 3, using polyvinylpyrrolidine (PVP), glyceryl tristearate (GTS) and indomethacin (IMC) / 5- fluorouracil (5-FU) to form SLNs from electrosprayed microparticles and electrospun fibres, were explored in higher throughput production methods. These comprised both high-throughput EHD approaches, and also spray drying.

We sought to fulfil the following objectives:

• Production of indomethacin or 5FU/GTS loaded PVP formulations via medium

speed electrospinning, corona electrospinning, and spray drying

• Characterisation of the materials in terms of physical form and component

compatibility

• Formation of SLNs from the materials generated in all three processes

• Investigation of the SLN in terms of their in vitro drug release and entrapment

efficiency

• To compare and contrast the results from all the processes investigated 152

4.3 Materials and methods

4.3.1 Medium speed electrospinning

Scale-up of the IMC-loaded fibres was performed using a TL-Pro-BM Nanofibre

Electrospinning Unit (Tong Li Tech) (Figure 4-3). Solutions containing indomethacin

(IMC, Alfa Aesar) or 5-fluorouracil (5-FU, Sigma-Aldrich) were prepared prior to the electrohydrodynamic (EHD) processes. These solutions were composed of the drug, polyvinylpyrrolidone (PVP, 360 kDa or 40kDa, Sigma-Aldrich) and glyceryl tristearate

(GTS, Sigma-Aldrich). Chloroform was used as the solvent for IMC. In the case of 5-FU, the drug was fully dissolved in dimethylformamide (DMF) and then mixed with a

PVP/GTS solution in chloroform for 10 min before electrospinning/spraying. All solvents were from Fisher Scientific. The solutions contained a final concentration of 10 % w/v

PVP (360 kDa), 5% w/v GTS and 2.5% w/v drug for the creation of fibres, while the solutions to create particles consisted of 20 % w/v PVP (40 kDa) , 5% w/v GTS and 2.5% w/v.

Each solution was prepared and loaded into a plastic syringe (Terumo). The syringe was attached to a 21 gauge spinneret (internal diameter 0.51mm) by PTFE tubing, and the solution was pumped at a flow rate of 20 ml h-1 using the inbuilt syringe pump. The spinneret arm was set to scan over a distance of 100 mm at a rate of 10.0 mm s-1, with the spinneret tip set at 20 cm from the collector. A rotating mandrel collector (diameter

100 mm), was coated in aluminium foil to allow for easy removal of the sample and set to rotate at 50 rpm. A voltage of 15 kV was applied to the spinneret, while the mandrel was grounded. Samples were stored in a desiccator after preparation. Scale-up of 5-FU- loaded fibres was performed as above, but using a voltage of 20 kV and a spinneret-to- collector distance of 30 cm.

Electrospraying scale-up was performed using the same parameters as above, except with a flow rate to 10 ml h-1. The spinneret arm was set to scan over a distance of 80 mm

153

at a rate of 10.0 mm s-1, with the spinneret tip set at 20 cm from the collector. A voltage of 15kV was applied to the spinneret, and a grounded aluminium plate of 20 x 20 cm served as a collector instead of the rotating mandrel. Samples were stored in a desiccator after fabrication.

Figure 4-3. A photograph of the TL-Pro-BM Nanofibre Electrospinning Unit.

4.3.2 High speed electrospinning (HSES)

HSES-IMC-F was performed by Dr Zsombor Nagy and Attila Balogh at Budapest

University of Technology and Economics (BUTE), courtesy of Prof. Gyorgy Marosi. The corona setup uses a stainless-steel spinneret with a spherical cap connected to a high- speed motor (Figure 4-4). The polymer solution was fed at a rate of 1000 ml h-1 using an

SEP-10S Plus syringe pump. The grounded collector was covered in aluminium foil and placed 30 cm from the spinneret. The spinneret was rotated at 25,000 rpm and a voltage of 45 kV was applied using an NT-65 high voltage DC supply (Uintronik Ltd) (Nagy et al.,

2015). 154

Figure 4-4. Diagram of the Corona HSES system. 1- High Voltage supply, 2- sharp-edged circular electrode, 3- grounded collector, 4- nanofibre formation, 5-lid, 6-solution inflow, 7- collector conveyor. Reproduced from (Molnar and Nagy 2016)

4.3.3 Spray drying

Spray drying was undertaken using a mini spray dryer (Buchi, B-290, Laboratory-Technik

LTD) with a closed loop. The spray nozzle tip diameter was 0.7 mm. The inlet air temperature was 70 °C for all the solutions investigated, and the outlet air temperature

40–48 °C. The liquid feed to the dryer was established at a rate of 10 mL min−1, while the flow of drying gas was approximately 35 m3 h-1. The experiments were performed under constant process conditions. The spray dryer was set up for the desired inlet temperature, and neat solvent pumped until the desired outlet temperature of between

68-70 °C was achieved and stabilized, after which spraying was commenced. All the solutions were spray dried individually, and after the process was complete the instrument temperature was allowed to cool to below 60 °C before the powder product was collected from the electrostatic particle chamber. The final product was immediately transferred to a glass vial and stored in a desiccator to prevent absorption of environmental moisture. 155

4.3.4 Physicochemical characterisation

SEM, XRD, DSC and FTIR were carried out as stated in Chapter 3, Section 3.3

4.3.5 Functional performance assays

Functional performance assays were carried out essentially as stated in Chapter 3,

Section 3.3. The only difference is that drug release was additionally tested using Fasted

Simulated Small Intestinal Fluid, FaSSIF (Biorelevant, UK).

4.3.6 Permeation assays

Caco-2 cells were seeded by Dr. Jong Bong Lee at the University of Nottingham. They were then grown in complete DMEM in a Falcon® 24-multiwell insert system plate

(Corning) for 21 days, changing the medium every 2 days. The seeding density was 3.75 x 104 cells/cm2. On the day of the assay, the Caco-2 monolayer was washed twice with transport buffer (Hanks Balanced Salt Solution supplemented with 10 mM 4-(2- hydroxyethyl)-1-piperazineethanesulfonic acid and with the pH adjusted to 7.4). The cells were left to equilibrate for 30 min at 37 °C. The assay was initiated when the donor solution (containing 200 µM of the drug) was placed on the apical side of the monolayer, and samples of 250 µL were withdrawn from the basal side at different time points over

2 h. Fresh buffer was supplied each time, to maintain a constant volume. The transepithelial electrical resistance (TEER) was measured before and after the experiment to assess the integrity of the monolayer. All experiments were performed in triplicate. The apparent permeability coefficient (Papp) was calculated using Equation

4-1:

Equation 4-1

푑푄 1 푃푎푝푝 = ( ) × 푑푡 퐴퐶0

-1 Where dQ/dt is the steady state flux (µmol s ), C0 is the initial concentration in the donor chamber (µM), and A represents the effective filter area of each well (cm2).

156

Samples obtained from permeation studies were subjected to a liquid-liquid extraction using ethyl acetate as the organic solvent. After the addition of ethyl acetate (2 mL) to each sample, they were vortexed for 2 min and then left to separate and for the organic layer to evaporate. The latter was accelerated using a stream of air.

4.3.7 High performance liquid chromatography

High performance liquid chromatography (HPLC) was performed on the permeation study samples using previously published and validated methods (Nassim, Shirazi et al.

2002). The residue from liquid-liquid extraction was first reconstituted in an appropriate mobile phase. The IMC mobile phase comprised 0.5 % v/v aqueous orthophosphoric acid, methanol, and acetonitrile (all Fisher Scientific) at volume ratios of 40: 20: 40. For

5-FU the mobile phase was acetonitrile: water (10: 90 v/v). For both analyses, a Luna

C18 column (Phenomenex) was utilised, with an injection volume of 10 µL. IMC experiments were undertaken at a flow rate of 2 mL min-1, and 5-FU chromatograms recorded at a flow rate of 1 mL min-1 (Tsvetkova 2012).

4.3.8 Statistical analysis

Sizes obtained from dynamic light scattering were assessed using a one tailed Student’s t-test. Data from cell culture experiments are presented as mean ± SD, and were statistically analysed using the MiniTab17 software (Minitab, Inc.). Statistical significance of differences was evaluated by one-way ANOVA using Dunnett Simultaneous 95% CI tests

157

4.4 Results and discussion

4.4.1 Materials characterisation

Details of the formulations produced are given in Table 4-1

Table 4-1. Identification of the materials and SLNs in Chapter 4.

Drug Material Method of production IMC-SLN Benchtop Electrospinning (BTES-IMC-F) F1 Medium Speed Electrospinning (MSES- Fibres F2 IMC-F)

IMC High speed Electrospinning (HSES-IMC-F) F3 Benchtop Electrospinning (BTES-IMC-P) F4 Medium Speed Electrospraying (MSES- Particles F5 IMC-P) Spray Drying (SD-IMC-P) F6 Drug Material Method of production 5FU-SLN Benchtop Electrospinning (BTES-5FU-F) F7 Fibres Medium Speed Electrospinning (MSES- F8 5FU-F)

5-FU HSES-F F9 Benchtop Electrospraying (BTES-5FU-P) F10

Particles Medium Speed Electrospraying (MSES- F11 5FU-P) Spray Drying (SD-5FU-P) F12

4.4.2 SEM

4.4.2.1 Medium speed and corona electrospinning

The SEM images obtained using IMC to fabricate fibres from medium-speed electrospinning (MSES) and corona electrospinning (HSES) are shown in Figure 4-5.

Diameter measurements were performed using the ImageJ software (Table 4-2), and show that the fibre diameter increases with throughput. The MSES-IMC-F images

(Figure 4-3) display cylindrical microfibres with smooth surfaces and a diameter of 5.71 158

± 2.74 µm, an approx. 2 µm size difference compared to the benchtop electrospinning product (3.25 ± 0.26 µm). Meanwhile, the fibers obtained from high-speed spinning,

HSES-IMC-F, have a mean diameter of 10.85 ± 3.15 µm and display irregular morphologies, including wavy and branched fibres. This morphology has been previously reported when using HSES (Molnar and Nagy 2016).

Figure 4-5. SEM images of electrospun fibres: a) MSES-IMC-F, b) HSES-IMC-F

Table 4-2. The diameters of IMC loaded fibres prepared using different electrospinning variants

Formulation Mean diameter (µm)

BTES-IMC-F 3.25 ± 0.26

MSES-IMC-F 5.71 ± 2.74

HSES-IMC-F 10.85 ± 3.15

The scale up of the 5-FU formulation using MSES was challenging. First, the same parameters used to fabricate the IMC loaded fibres were explored. SEM images of the products are shown in Figure 4-6. Individual fibres cannot be made out when using a 20 cm tip-to-collector distance, and it appears that not all the solvent is evaporating, leading to the fusion of the fibres. To try to overcome this issue, the tip-to-collector distance and voltage were increased to help with the evaporation of the solvent (raising the distance

159

between the needle and collector allows a longer time for the solvent to evaporate, and increasing the voltage of has a similar effect (Leach, Feng et al. 2011). After increasing the voltage and distance, the resultant fibres had somewhat better definition, and only a few regions of aggregation can be seen. The final fibres show thick uniform fibres with largely cylindrical morphology.

Figure 4-6. Optimisation of MSES for the electrospinning of SLN-5FU fibres

As seen in the scale up of IMC electrospinning, the diameter of the 5FU loaded fibres

(Table 4-3) increases with throughput.

Table 4-3. Size comparison of the scale up process of 5FU loaded fibres

Formulation Mean diameter (µm)

BTES-5FU-F 0.18 ± 0.06

MSES-5FU-F 0.51 ± 0.20

The main difference between the different ES processes is the flow rate of the polymer solution through the spinneret; the flow rates utilised were 1.5 (BTES), 20 (MSES) and

1000 (HSES) ml h-1. An increase in fibre diameter and raised number of deformities when using higher flow rates has previously been reported (Nagy, Balogh et al. 2015, Démuth,

Farkas et al. 2016). The literature thus concurs with the observations made in this work. 160

Benchtop electrospinning also results in greater diameter uniformity when compared to the MSES and HSES.

4.4.2.2 Medium speed electrospraying

Uniform particles could not be obtained from scaled-up electrospraying. The particles produced have very irregular morphologies (Figure 4-7). PVP- IMC-P has a mean size of 3.53 ± 1.08 μm, similar to PVP- 5FU-P at 3.76 ± 1.10 μm. Flow rate is one of the most important parameters affecting size in microparticle production (Milad, Jalal et al. 2015).

In general, it is well established that higher flow rates result in greater particle diameters

(Almeria, Deng et al. 2010, Songsurang, Praphairaksit et al. 2011, Bohr, Kristensen et al. 2012). The particle sizes from these experiments agree with the literature, as the size of the particles fabricated from the MSES are larger than those made by BTES (BTES-

IMC-P: 1.63 ± 0.43 µm and BTES-5FU-P: 1.35 ± 0.5 µm).

a) b)

Figure 4-7. SEM Images for the scaled-up electrosprayed particles a) MSES-IMC-P and b) MSES- 5FU-P

a 4.4.2.3 Spray drying

The spray dried particles possess spherical shapes, smooth surfaces, and non-uniform b sizes (see Figure 4-8). The morphology of the particles appears to be more uniform than the systems obtained by electrospraying, which could be due to the faster evaporation

161

of the solvent in spray drying; the drying temperature has been reported to have an effect on particle shape and surface characteristics (Liu, Wu et al. 2011). Both types of drug loaded particles have high polydispersity. The SD-IMC-P have an average size of 7.15

± 4.39 µm and SD-5FU-P an average size of 5.78 ± 4.44 µm. One of the possible reasons for the size difference could be the use of a co-solvent in the case of 5-FU, as this has been reported to have an effect on the size of the particles (Boraey, Hoe et al. 2013).

However, the size of the droplets can also be affected by other factors such as the chemical and physical properties of the solution.

a) b)

Figure 4-8. SEM images of formulations prepared by spray drying: a) SD- IMC-P, b) SD-5FU-P

If compared with benchtop and medium speed electrospraying, the spray dried particles are larger. This is as expected, since the Buchi mini spray drier creates micron-sized particles which normally range from 20-300 µm (in contrast, the Buchi nanospray drier creates particles in the nano range 300 nm-5 µm (Roland and Huagang 1988,

Gharsallaoui, Roudaut et al. 2007, Vehring 2008).

4.4.3 FTIR

FTIR spectra of the fibres and particles were collected to confirm the structural composition of the materials. The results are depicted in Figure 4-9. FTIR data of

162

systems obtained from BTES are also included for comparison. It can be observed that the spectra of the formulations encapsulating IMC and 5FU appear to be very similar.

Major bands of interest include the C-H bonds of GTS and PVP which are seen at 2850 cm-1 and 2920 cm-1. The broad band from 3200 – 3600 cm -1 shows that water is present in all systems, due to the hygroscopic nature of the PVP used to fabricate the formulations. All spectra show a prominent peak at 1660 cm-1 which matches the amide

C=O stretching found in PVP, with a small shoulder at 1735 cm-1 representing the ester

C=O stretching from GTS.

Figure 4-9. FTIR spectra of the scaled up formulations.

However, it is hard to identify the peaks corresponding to 5-FU and IMC as those are obscured by other ingredients in the fibres and particles. For example, the IMC C=C stretching vibrations from 1580-1620 cm-1 cannot be observed in the formulations (Figure

4-9). Similarly, the 5-FU pyridine vibrations (C=N and C=C) are obscured by the PVP and GTS bands. In the carboxylate region changes can be seen in the of the spectra of the indomethacin loaded formulations, where the characteristic bands of indomethacin 163

are shifted to lower wavenumber values and overlap with the carbonyl band of PVP at

1660 cm-1. However, as the spectra is very complex, it is not possible to obtain much information about the nature of the components in the fibers and particles.

4.4.4 XRD

XRD patterns were collected for all the formulations and are displayed in Figure 4-10. In the XRD data a broad peak at 21.63o which is consistent with the presence of the semi- crystalline GTS. The IMC and 5FU, which are crystalline materials, do not appear to have any Bragg reflection, which confirms their amorphous state in the electrospun and electrosprayed materials. This corresponds to the literature as both the electrospraying and spray drying are techniques known to fabricate amorphous materials (Nagy, Balogh et al. 2015, Newman 2015).

Figure 4-10. XRD diffraction patterns of IMC and 5-FU-loaded scaled up formulations.

4.4.5 DSC

DSC data, showing the second heating cycle to prevent the overlap signals with the evaporation of the water absorbed naturally by the PVP, are given in Figure 4-11. GTS

164

presents a two endotherms (49 oC, 60 oC) and one exothermic (51.2 oC) peak which can be seen in Figure 3-10. All the materials present only the GTS melting and recrystallisation events. The scaled up formulations thus show the same behaviour as the BTES products. This suggests that scaling up the fabrication process using EHD techniques and spray drying does not affect the nature of the products, which all comprise amorphous solid dispersions.

Figure 4-11. DSC thermograms of the scaled-up formulations. Exo up.

4.5 Solid lipid nanoparticle characterisation

4.5.1 TEM

SLNs were assembled from the formulations by adding them to water, as detailed in

Section 3.3.1. TEM images of the samples, presented in Figure 4-12, confirm the presence of nanoparticles self-assembled from the fibres and particles. To note that F11, has a different morphology than the rest of the SLNs.

165

F3 F2

F5 F6

F8 F11

166

Figure 4-12- TEM images SLNs self-assembled SLN from the electrospun, electrosprayed, and spray dried materials. F1 and F4, F7 and F10 are reported in Chapter 3, Figure 3-14.

F12

Continuation of Figure 4-12

The sizes of the SLNs were analysed with ImageJ, and the results are shown in Table

4-4. The results are varied, but the SLN assembled form the fibres are more polydisperse when compared with those assembled from the particles. Caution must be taken in interpreting the data from the TEM images however, as the number of particles per image is limited. Therefore, DLS was also used to assess particle size (Table 4-4). The TEM sizes are smaller than the DLS size as expected, given that the particles are hydrated in

DLS measurements. The different IMC loaded SLNs (F1-F6) do not appear to have any significant variation in their size, suggesting that the changes in fibre and particle morphology observed upon scale up do not influence the self-assembly process.

Meanwhile, the 5FU loaded SLN seem to differ in size, especially the SLN assembled from the particles.

The trend described in Chapter 3 (Section 3.4.4) is also seen in the scale-up processes, and the sizes of the SLN assembled from the fibres are smaller than the ones assembled from particles. The PDI values tend to be smaller for the SLNs self-assembled from the particles, translating in a higher monodispersity.

167

Table 4-4. The sizes of the SLNs, as determined from TEM images and DLS (n=number of particles counted).

Formulation TEM size (nm) DLS Size (nm) PDI

F1 33.1 ± 8.6 (n=41) 121.5 ± 30.7 0.51 ± 0.03

F2 43.9 ± 20.7 (n=7) 118.1 ± 8.0 0.37 ± 0.09

F3 73.2 ± 24.9 (n=47) 137.2 ± 6.3 0.49 ± 0.15

F4 27.7 ± 8.0 (n=68) 143.6 ± 0.6 0.15± 0.00

F5 89.9 ± 33.6 (n=29) 180 ± 3.6 0.70 ± 0.01

F6 53.0 ± 19.9 (n=15) 134 ± 3.6 0.10 ± 0.01

F7 93.9 ± 53.2 (n=24) 88.08± 1.6 0.50 ± 0.02

F8 107 ± 44.7 (n=12) 104.5 ± 0.3 0.46 ± 0.14

F10 26.0 ± 7.1 (n=49) 143.9 ± 0.7 0.23 ± 0.01

F11 23.1 ± 4.6 (n=31) 66.43 ± 1.9 0.11 ± 0.02

F12 82.5 ± 23.2 (n=3) 141.8 ± 2.7 0.14 ± 0.01

4.5.2 Entrapment efficiency and drug loading

A summary of the EE% of the formulations is presented in Table 4-5. The theoretical drug loading of the fibers is16.6 %, while of the particles it was of 10 %.

Table 4-5. Entrapment efficiency of the formulations (n=3).

SLN-IMC %EE SLN-5FU %EE

F1 90.2 ± 0.6 F7 88.2 ± 2.1

F2 84.0 ± 0.9 F8 37.1 ± 2.2

F3 85.9 ±1.8 F10 49.2 ± 4.5

F4 89.5 ± 0.9 F11 54.0 ± 2.7

F5 88.2 ± 3.1 F12 64.9 ± 16.7

168

F6 86.1 ± 4.7

The SLN-IMC (F1-F6) have an EE% between 84-90%, with the highest being obtained from benchtop electrospinning. The entrapment efficiency seems to be largely consistent throughout the scale-up process and is similar to the results of Potta, who reported an

EE% of 93-94% when preparing ibuprofen-loaded loaded SLNs (Potta, Minemi et al.

2011).

Meanwhile, the SLN-5FU (F7-F12) have an EE% ranging from 37-88%. Here, there do seem to be differences between different scale processes. The highest EE% is obtained with benchtop electrospraying, and the lowest can be found when using MSES electrospinning, with a difference of 51.19%. The electrospraying EE% seems to be more consistent than that from electrospinning. The polymer entanglements of PVP may thus have an influence on the self-assembly of the particles; the molecular weight used for electrospraying is lower and would be expected to dissolve in water more quickly, facilitating the assembly and entrapment of 5FU.

The EE% is higher and more consistent with the SLN -IMC. There are many factors that can influence the entrapment efficiency of a component into SLNs, but the main factor is the hydrophobicity of the drug. Since indomethacin is more hydrophobic than 5-FU, it will have greater intermolecular interactions with the GTS. Also, during the self-assembly process the fabricated materials must be dissolved in the hydration medium to be encapsulated into the SLN. In the case of 5-FU a large amount remains un-encapsulated in the external aqueous medium owing to the relatively high solubility of the drug (Eloy,

Claro de Souza et al. 2014).

This same pattern can be seen in the literature, with hydrophilic drugs having on average a lower EE% than hydrophobic drugs; for example, an EE% of 41.65% was reported for the hydrophilic drug paromomycin (Ghadiri, Fatemi et al. 2012), while SLN encapsulating

169

hydrophobic doxorubicin have been found to have an EE of 67.5 ± 2.4 (Kang, Chun et al. 2010), and paclitaxel SLN can reach 92% EE (Chirio, Gallarate et al. 2014).

4.5.3 Drug release

The drug release of the SLNs was first examined in PBS at pH 6.8, and the results are presented in Figure 4-13 and Figure 4-14. The same pattern of IMC release is observed as with the BTES formulations presented in Chapter 3 (Section 3.4.) All the SLN show very similar release patterns. For the IMC loaded SLN, the initial rate for the first 8h is identical in all cases, where after this period F2 and F3 drug release is slower than F5 and F6. On the other hand, the release of 5-FU seems to differ from one formulation to the other: F8 and F12 which are self-assembled from the MSES and SD materials, seem to behave similarly and reach approximately 80% release after 24h. However, F11

(which had a different morphology than the other SLN in the TEM images) behaves differently, releasing 100% of the drug in 24h. The entrapment efficiency and size of F11 is also lower than the other SLNs, which could also contribute to the release differences.

In both the IMC and 5-FU-loaded formulations, an initial burst release is seen and the rate then becomes slower after the first few hours. Biphasic drug release has been reported to be desirable to provide an immediate and long-lasting effect (Thukral,

Dumoga et al. 2014). This has previously been identified as a problem encountered with

SLN (Müller, Mäder et al. 2000).

SLN are usually reported to have a drug release that lasts several weeks, however the

SLN fabricated with this method present a faster drug release. This concurs with studies by Yu et al. who fabricated naproxen-loaded SLN using an EHD sacrificial template and obtained a drug release of 87.6% over a period of 24h (Yu, Williams et al. 2011). This might imply that this fabrication method mainly encapsulates the drug on the surface of the SLNs, leading to the rapid dissolution of the drug molecules from the surface layer of the particles. This idea has previously been posited by Yassin, who prepared SLN by a double emulsion-solvent evaporation technique (w/o/w) (Yassin, Anwer et al. 2010), and 170

Shazly, who fabricated SLN by the melt-emulsification method, and observed the same effect (Shazly 2017).

Figure 4-13. Cumulative drug release from indomethacin loaded SLNs (F2-F5) in PBS. Data are given as mean ± S.D. from three independent experiments.

171

Figure 4-14. Cumulative drug release from 5-FU loaded SLNs (F8-F12) in PBS. Data are given as mean ± S.D. from three independent experiments.

Three formulations containing each drug were chosen to observe the drug release in fasted state simulated intestinal fluid (FaSSIF). Spray dried formulations (F6 and F12) were chosen to determine if the fabrication process affected the performance of the

SLNs. One formulation from each of electrospinning and electrospraying was also chosen, based on those that presented the lowest PDI. FaSSIF medium is meant to represent the composition of the gastrointestinal fluids as closely as possible (Jantratid,

Janssen et al. 2008). The results of these experiments are seen in Figure 4-15 (IMC) and Figure 4-16 (5-FU). The data do not show any major differences changes to release in PBS in the case of IMC.

172

Figure 4-15 Cumulative IMC release from F2, F5 and F6 in FaSSIF. Data are given as mean ± S.D. from three independent experiments.

The release from the formulations containing 5-FU do seem to be affected by the use of

FaSSIF, and the cumulative drug release decreases when compared the PBS data. F12 and F8 reach less than 50% drug release after 48h, while F11 releases more quickly and achieves 69.4 ± 15.6% drug release in 48h. The release profiles of the SLN are similar to that reported by Yassin for 5FU-SLN using a media which contains colonic enzymes and bacteria of the GI tract (Yassin, Anwer et al. 2010), this provides an in-vitro system that mimics the microbial environment in the human colon (Zhang and Neau 2002). IMC drug release from the SLN appears to be similar in FASSIF and PBS, on the other hand the drug release of 5-FU appears to be slower in FASSIF when compared with the release in PBS.

173

Figure 4-16. Cumulative drug release of 5FU from SLN formulations in FASSIF Data are given as mean ± S.D. from three independent experiments.

Considering that the transit times of the small intestine and colon are around 2.5-3 h and

30-40 h respectively (Metcalf, Phillips et al. 1987, Camilleri, Colemont et al. 1989, Degen and Phillips 1996), maximum drug release would be reached while the formulations are still located in the intestinal tract after oral administration, giving them sufficient time to free their drug cargo (Yassin, Anwer et al. 2010). Other approaches to increase the selective delivery of the SLNs to cancer tissues such as targeting ligands or pH sensitive systems could also be applied (Kim, Kim et al. 2015, Tran, Ramasamy et al. 2015). In order to ensure that the SLNs reach the colon intact, an enteric coated capsule can be used.

174

4.5.4 Permeation studies

The determination of the SLN permeability coefficient was performed using monolayers of a human colon carcinoma cell line, which is commonly used to predict the absorption of orally administered drugs (Stenberg, Norinder et al. 2001). When grown in specific conditions, the Caco-2 cell line has the ability to differentiate into monolayers which are similar to the single-layer of epithelial cells that covers the inner intestinal wall (Hubatsch,

Ragnarsson et al. 2007).

The experiment consists of two parts: a) the cultivation of the Caco-2 cells on filter supports, and b) drug transport experiments. The experiments performed in this study were apical-to-basolateral transport experiments, where the SLN were placed on the apical side of the Caco-2 monolayer and samples were taken at certain time point from the acceptor compartment. The transepithelial electrical resistance (TEER) values are strong indicators of the integrity of cellular barriers (Srinivasan, Kolli et al. 2015), so these values were recorded before and after the experiment to ensure that the integrity of the monolayer was not affected by the drugs used. The TEER values remained between

395.2-382.04 Ω.cm2; according to Hubatsch, values >165 Ω.cm2 are acceptable

(Hubatsch, Ragnarsson et al. 2007). From the TEER values it can be confirmed that the integrity of the membrane was intact during the duration of the experiments. On completion of the experiment, the cell monolayer was disrupted and collected. After sample collection, a liquid--liquid extraction was undertaken and drug quantification performed using HPLC analysis.

The formulations chosen for these experiments were F2, F5, F6 for IMC and F8, F11,

F12 for 5-FU (the same formulations as previously selected for the FaSSIF drug release experiments); the pure drugs were also analysed. Assessment of the permeation of pure

6 -1 IMC resulted in a Papp value (Equation 4-1) of 11.6 x 10 cm s , typical for biopharmaceutical classification system class II (BCS II) drugs and confirming the high permeability of IMC (Lee, Zgair et al. 2017). In contrast, the 5-FU Papp value was 0.10 175

x 106 cm s-1, a consistent with the literature (which shows 5-FU is a BSC class III drug which has a poor permeability in the Caco-2 monolayer model) (Buur, Trier et al. 1996).

When analysing the SLN-IMC containing samples, no drug was found in the receiver compartment with any of the formulations. Most of the drug was found in the cell monolayer lysate, as observed in Figure 4-17. Organic solvents were used to disrupt the cell monolayer, which will also result in destruction of the SLN: it is thus not possible to determine whether the drug was present in the cells incorporated in SLNs or as free drug.

Figure 4-17. Percentage of drug found in the cell lysate. Data are given as mean ± SD (n=3)

The results show that the SLN do not appear to permeate thorough the cell monolayer but instead accumulate in the cells. The SLNs have significant advantages over either drug alone. If the IMC were to permeate through tumour cells and reach the systemic circulation, unwanted side effects could arise; the SLNs offer an alternative to preclude this. 5-FU alone does not effectively permeate, meaning that it may translocate through cells/tissues but without exerting a pharmacological effect. The SLNs enable both IMC and 5-FU to be effectively localised in cancer cells, ideal for the treatment of tumours.

176

4.5.5 Cell culture studies

As discussed in Chapter 3 (Section 3.4.3.5), indomethacin and 5-FU have anti-cancer properties against colon cancer cells. The cytotoxicity of the scaled-up formulations was thus tested using the Caco-2 cell line. The cytotoxicity seems to be comparable to the pure drug for most of IMC formulations, as shown in Figure 4-18. However, two formulations stand out: F4 seems to have a greater cytotoxicity compared to the rest of the formulations and to pure IMC, while F1 seems to be less effective than the other formulations in this regard. F4 and F1 did not seem to behave differently from the others when testing the performance assays or the physicochemical characterisation, making these results surprising. The only difference that can be noted is in the PDI. PDI is an indicator of uniformity of the particles, and reflects the size distribution of the nanoparticles (Masarudin, Cutts et al. 2015). PDI has been reported to have an impact on the clinical applications of a drug delivery system (Danaei, Dehghankhold et al. 2018), as the tendency of a nanocarrier to accumulate in target tissue depends on factors such as particle size and distribution (Bulbake, Doppalapudi et al. 2017, Dong, Tchung et al.

2017). F1 has a PDI of 0.5, and thus the SLNs have high polydispersity. This high value of the PDI could also indicate agglomeration of the SLNs. On the other hand, F4 has a

PDI of 0.1 which is one of the lowest among the SLN-IMC formulations. The influence of

PDI on viability is plotted in Figure 4-19: an increase of PDI seems in general to lead to an increase in cell viability.

177

Formulation

100 90 F1 80 F6 70 F5 60 F3 50 F2 40

Viability (%) Viability 30 20 10 F4 0 0 0.1 0.2 0.3 0.4 0.5 0.6

PDI

Figure 4-19. Viability vs. PDI for IMC-loaded SLNs

The 5FU formulations seem to behave similarly (Figure 4-20), except for F11 which results in a lower viability (it also differs in size, morphology and EE% from all the other formulations). The size of F11 is smaller than the others, and it is also highly

178

monodisperse. The size of nanomaterials has been reported to have a significant influence on the intracellular responses, their cytotoxicity and potential mechanisms of toxicity (Sohaebuddin, Thevenot et al. 2010), and smaller nanoparticles tend to possess higher toxicity than larger ones (Shang, Nienhaus et al. 2014).

4.5.6 Stability studies

The physical stability of the SLNs during prolonged storage can be determined by monitoring changes in particle size. Thus, in order to assess the stability of the formulations, the hydrodynamic size of the SLNs after 6 months storage at room temperature was studied.

As shown in Figure 4-21, it is clear that the size of the IMC-SLNs self-assembled on- demand from stored materials (electrospun fibres, electrosprayed and spray dried particles) are similar to those assembled at t0 (from freshly prepared fibres/particles).

However, the SLN suspensions stored for 6 months seem to have poor colloidal stability,

179

with significant changes in particle size observed. The poor stability of the SLNs suspensions can be attributed to gelation and recrystallization of the lipid phase after during storage, which can be accelerated with temperature and light exposure

(Siekmann and Westesen 1994).

From the 5-FU formulations there are no clear trends after storage: for example, SLNs from F11 and F12 have similar sizes when prepared from fresh formulations and after the fibres or particles have been in storage for 6 months, but other formulations such as

F8 and F10 result in very different sized SLNs depending on the self-assembly is done immediately after production or after the products have been stored. This could arise as

5FU is a hydrophilic drug, which here is encapsulated in a hydrophobic carrier. Phase separation of the components in the PVP/GTS/5-FU formulations could thus occur upon storage. It is reported that when lipids solidify into crystalline matrices, drug molecules have a reduced opportunity to be incorporated within such ordered structures, which could result in these differences (Patel, Lakkadwala et al. 2014).

Figure 4-21. The results of stability studies performed for 6 months at room temperature. Data are shown from 3 independent experiments as mean ± S.D. * denotes p < 0.05 with respect to t0.

180

4.6 Conclusions

This chapter demonstrates the viability of scaling-up production of formulations able to self-assemble into SLNs. It proved possible to scale up fibre production to 1 kg h-1, and particle production to 10 g h-1 using electrohydrodynamic techniques, or 600 g h-1 with spray drying. Overall, the particle or fibre size increased and became less uniform as the scale increased. However, all the formulations could be assembled successfully into

SLNs, and the cytotoxicity of the SLNs were mostly comparable with the pure drug.

The results from scale up of indomethacin-based formulations are generally consistent.

The indomethacin encapsulation efficiency in the assembled SLNs was largely the same, regardless of the scale of production. The SLNs give sustained release of indomethacin over a period greater than eight hours. More problems were encountered during the scale-up of 5FU-based materials. The 5-FU-loaded SLN also seem to be less stable upon storage than the IMC-SLN, thus it seems that hydrophobic drugs provide a better stability and entrapment efficiency than hydrophilic drugs.

181

4.7 References

Almeria, B., W. Deng, T. M. Fahmy and A. Gomez (2010). "Controlling the morphology of electrospray-generated PLGA microparticles for drug delivery." J Colloid Interface Sci 343(1): 125-133. Balogh, A., R. Cselkó, B. Démuth, G. Verreck, J. Mensch, G. Marosi and Z. K. Nagy (2015). "Alternating current electrospinning for preparation of fibrous drug delivery systems." Int J Pharm 495(1): 75-80. Bioinicia. (2018). "Increasing the throughput of the electrospinning process" Retrivied 10/08/2018 from "http://bioinicia.com/increasing-the-throughput-of-the-electrospinning- process/. Bohr, A., J. Kristensen, M. Dyas, M. Edirisinghe and E. Stride (2012). "Release profile and characteristics of electrosprayed particles for oral delivery of a practically insoluble drug." J R Soc Interface 9(75): 2437-2449. Boraey, M. A., S. Hoe, H. Sharif, D. P. Miller, D. Lechuga-Ballesteros and R. Vehring (2013). "Improvement of the dispersibility of spray-dried budesonide powders using leucine in an ethanol–water cosolvent system." Powder Technology 236: 171-178. Böttjer, R., T. Grothe, D. Wehlage and A. Ehrmann (2018). "Electrospraying poloxamer/(bio-)polymer blends using a needleless electrospinning machine." J of Textiles and Fibrous Materials 1: 2515221117743079. Bulbake, U., S. Doppalapudi, N. Kommineni and W. Khan (2017). "Liposomal Formulations in Clinical Use: An Updated Review." Pharmaceutics 9(2). Buur, A., L. Trier, C. Magnusson and P. Artursson (1996). "Permeability of 5-fluorouracil and prodrugs in Caco-2 cell monolayers." Int J Pharm 129(1): 223-231. Camilleri, M., L. J. Colemont, S. F. Phillips, M. L. Brown, G. M. Thomforde, N. Chapman and A. R. Zinsmeister (1989). "Human gastric emptying and colonic filling of solids characterized by a new method." Am J Physiol 257(2 Pt 1): G284-290. Chakraborty, S., I. C. Liao, A. Adler and K. W. Leong (2009). "Electrohydrodynamics: A facile technique to fabricate drug delivery systems." Advanced drug delivery reviews 61(12): 1043-1054. Chirio, D., M. Gallarate, E. Peira, L. Battaglia, E. Muntoni, C. Riganti, E. Biasibetti, M. T. Capucchio, A. Valazza, P. Panciani, M. Lanotte, L. Annovazzi, V. Caldera, M. Mellai, G. Filice, S. Corona and D. Schiffer (2014). "Positive-charged solid lipid nanoparticles as paclitaxel drug delivery system in glioblastoma treatment." Eur J Pharm Biopharm 88(3): 746-758. Danaei, M., M. Dehghankhold, S. Ataei, F. Hasanzadeh Davarani, R. Javanmard, A. Dokhani, S. Khorasani and M. R. Mozafari (2018). "Impact of Particle Size and Polydispersity Index on the Clinical Applications of Lipidic Nanocarrier Systems." Pharmaceutics 10(2). Degen, L. P. and S. F. Phillips (1996). "Variability of gastrointestinal transit in healthy women and men." Gut 39(2): 299-305. Dong, Y. D., E. Tchung, C. Nowell, S. Kaga, N. Leong, D. Mehta, L. M. Kaminskas and B. J. Boyd (2017). "Microfluidic preparation of drug-loaded PEGylated liposomes, and the impact of liposome size on tumour retention and penetration." J Liposome Res: 1-9. Eloy, J. O., M. Claro de Souza, R. Petrilli, J. P. A. Barcellos, R. J. Lee and J. M. Marchetti (2014). "Liposomes as carriers of hydrophilic small molecule drugs: Strategies to enhance encapsulation and delivery." Colloids Surf B Biointerfaces 123: 345-363.

182

Ghadiri, M., S. Fatemi, A. Vatanara, D. Doroud, A. R. Najafabadi, M. Darabi and A. A. Rahimi (2012). "Loading hydrophilic drug in solid lipid media as nanoparticles: Statistical modeling of entrapment efficiency and particle size." Int J Pharm 424(1): 128-137. Gharsallaoui, A., G. Roudaut, O. Chambin, A. Voilley and R. Saurel (2007). "Applications of spray-drying in microencapsulation of food ingredients: An overview." Food Research International 40(9): 1107-1121. Haitao, N. and L. Tong (2012). "Fiber Generators in Needleless Electrospinning." J Nanomater 2012: 13. Hubatsch, I., E. G. E. Ragnarsson and P. Artursson (2007). "Determination of drug permeability and prediction of drug absorption in Caco-2 monolayers." Nature Protocols 2: 2111. Jain, M., L. Ganesh, B. Manoj, C. Randhir, B. Shashikant and S. Chirag (2012). Spray Drying in Pharmaceutical Industry: A Review, European Pharmaceutical Review. Jantratid, E., N. Janssen, C. Reppas and J. B. Dressman (2008). "Dissolution media simulating conditions in the proximal human gastrointestinal tract: an update." Pharm Res 25(7): 1663-1676. Jirsak, O., F. Saternik, D. Lukas, V. Kotek, L. Martinova and C. Jiri (2005). A method of nanofibres production from a polymer solution using electrostatic spinning and a devide for carrying out the method. C. Patent. EP 1 673 493 B1. Kang, K. W., M.-K. Chun, O. Kim, R. K. Subedi, S.-G. Ahn, J.-H. Yoon and H.-K. Choi (2010). "Doxorubicin-loaded solid lipid nanoparticles to overcome multidrug resistance in cancer therapy." Nanomedicine: Nanotechnology, Biology and Medicine 6(2): 210- 213. Kanojia, G., R. t. Have, P. C. Soema, H. Frijlink, J.-P. Amorij and G. Kersten (2017). "Developments in the formulation and delivery of spray dried vaccines." Human Vaccines & Immunotherapeutics 13(10): 2364-2378. Keshani, S., W. R. W. Daud, M. M. Nourouzi, F. Namvar and M. Ghasemi (2015). "Spray drying: An overview on wall deposition, process and modeling." Journal of Food Engineering 146: 152-162. Kim, G. and W. Kim (2006). "Formation of oriented nanofibers using electrospinning." Applied Physics Letters 88(23): 233101. Kim, J.-H., Y. Kim, K. H. Bae, T. G. Park, J. H. Lee and K. Park (2015). "Tumor-Targeted Delivery of Paclitaxel Using Low Density Lipoprotein-Mimetic Solid Lipid Nanoparticles." Molecular Pharmaceutics 12(4): 1230-1241. Krogstad, E. A. and K. A. Woodrow (2014). "Manufacturing scale-up of electrospun poly(vinyl alcohol) fibers containing tenofovir for vaginal drug delivery." Int J Pharm 475(1): 282-291. Leach, M. K., Z.-Q. Feng, S. J. Tuck and J. M. Corey (2011). "Electrospinning Fundamentals: Optimizing Solution and Apparatus Parameters." JoVE(47): 2494. Lee, J. B., A. Zgair, D. A. Taha, X. Zang, L. Kagan, T. H. Kim, M. G. Kim, H. Y. Yun, P. M. Fischer and P. Gershkovich (2017). "Quantitative analysis of lab-to-lab variability in Caco-2 permeability assays." Eur J Pharm Biopharm 114: 38-42. Lhernould, M. S. and P. Lambert (2011). "Compact polymer multi-nozzles electrospray device with integrated microfluidic feeding system." Journal of Electrostatics 69(4): 313- 319.

183

Liu, W., W. D. Wu, C. Selomulya and X. D. Chen (2011). "Uniform Chitosan Microparticles Prepared by a Novel Spray-Drying Technique." International Journal of Chemical Engineering 2011: 7. Masarudin, M. J., S. M. Cutts, B. J. Evison, D. R. Phillips and P. J. Pigram (2015). "Factors determining the stability, size distribution, and cellular accumulation of small, monodisperse chitosan nanoparticles as candidate vectors for anticancer drug delivery: application to the passive encapsulation of [(14)C]-doxorubicin." Nanotechnology, Science and Applications 8: 67-80. Metcalf, A. M., S. F. Phillips, A. R. Zinsmeister, R. L. MacCarty, R. W. Beart and B. G. Wolff (1987). "Simplified assessment of segmental colonic transit." Gastroenterology 92(1): 40-47. Milad, J. N., B. Jalal and M. Hamid (2015). "Size and morphology controlling of PLGA microparticles produced by electro hydrodynamic atomization." Polymers for Advanced Technologies 26(5): 502-513. Molnar, K. and Z. K. Nagy (2016). "Corona-electrospinning: Needleless method for high- throughput continuous nanofiber production." European Polymer Journal 74: 279-286. Mujumdar, A. S. (2014). Handbook of Industrial Drying, Fourth Edition, Taylor & Francis. Müller, R. H., K. Mäder and S. Gohla (2000). "Solid lipid nanoparticles (SLN) for controlled drug delivery – a review of the state of the art." Eur J Pharm Biopharm 50(1): 161-177. Nagy, Z. K., A. Balogh, B. Demuth, H. Pataki, T. Vigh, B. Szabo, K. Molnar, B. T. Schmidt, P. Horak, G. Marosi, G. Verreck, I. Van Assche and M. E. Brewster (2015). "High speed electrospinning for scaled-up production of amorphous solid dispersion of itraconazole." Int J Pharm 480(1-2): 137-142. Nassim, M. A., F. H. Shirazi, C. M. Cripps, S. Veerasinghan, M. J. Molepo, M. Obrocea, D. Redmond, S. Bates, D. Fry, D. J. Stewart and R. Goel (2002). "An HPLC method for the measurement of 5-fluorouracil in human plasma with a low detection limit and a high extraction yield." Int J Mol Med 10(4): 513-516. Newman, A. (2015). Pharmaceutical Amorphous Solid Dispersions, Wiley. Niu, H., T. Lin and X. Wang (2009). "Needleless electrospinning. I. A comparison of cylinder and disk nozzles." J Appl Polym Sci 114(6): 3524-3530. Parhizkar, M., P. J. T. Reardon, J. C. Knowles, R. J. Browning, E. Stride, R. B. Pedley, T. Grego and M. Edirisinghe (2017). "Performance of novel high throughput multi electrospray systems for forming of polymeric micro/nanoparticles." Materials & Design 126: 73-84. Parikh, D. (2008). Advances in spray drying technology: New applications for a proven process. Patel, M. N., S. Lakkadwala, M. S. Majrad, E. R. Injeti, S. M. Gollmer, Z. A. Shah, S. H. S. Boddu and J. Nesamony (2014). "Characterization and Evaluation of 5-Fluorouracil- Loaded Solid Lipid Nanoparticles Prepared via a Temperature-Modulated Solidification Technique." AAPS PharmSciTech 15(6): 1498-1508. Persano, L., A. Camposeo, C. Tekmen and D. Pisignano (2013). "Industrial Upscaling of Electrospinning and Applications of Polymer Nanofibers: A Review." Macromolecular Materials and Engineering 298(5): 504-520. Potta, S. G., S. Minemi, R. K. Nukala, C. Peinado, D. A. Lamprou, A. Urquhart and D. Douroumis (2011). "Preparation and characterization of ibuprofen solid lipid nanoparticles with enhanced solubility." J Microencapsul 28(1): 74-81.

184

Roland, B. and C. Huagang (1988). "Preparation of Biodegradable Poly(±)lactide Microparticles Using a Spray‐Drying Technique." Journal of Pharmacy and Pharmacology 40(11): 754-757. Rulison, A. J. and R. C. Flagan (1993). "Scale‐up of electrospray atomization using linear arrays of Taylor cones." Review of Scientific Instruments 64(3): 683-686. Shang, L., K. Nienhaus and G. U. Nienhaus (2014). "Engineered nanoparticles interacting with cells: size matters." J Nanobiotechnology 12: 5-5. Shazly, G. A. (2017). "Ciprofloxacin Controlled-Solid Lipid Nanoparticles: Characterization, In Vitro Release, and Antibacterial Activity Assessment." BioMed Research International 2017: 9. Shuai, W., Y. Yubo, Z. Yang, F. Xiaoliang, Z. Chen, Z. Yue, L. Yaoxian, Y. Qingbiao and S. Yan (2014). "Fabrication of large‐scale superhydrophobic composite films with enhanced tensile properties by multinozzle conveyor belt electrospinning." J. Appl. Polym. Sci 131(1). Siekmann, B. and K. Westesen (1994). "Thermoanalysis of the recrystallization process of melt-homogenized glyceride nanoparticles." Colloids Surf B Biointerfaces 3(3): 159- 175. Sohaebuddin, S. K., P. T. Thevenot, D. Baker, J. W. Eaton and L. Tang (2010). "Nanomaterial cytotoxicity is composition, size, and cell type dependent." Particle and Fibre Toxicology 7(1): 22. Songsurang, K., N. Praphairaksit, K. Siraleartmukul and N. Muangsin (2011). "Electrospray fabrication of doxorubicin-chitosan-tripolyphosphate nanoparticles for delivery of doxorubicin." Archives of Pharmacal Research 34(4): 583. Srinivasan, B., A. R. Kolli, M. B. Esch, H. E. Abaci, M. L. Shuler and J. J. Hickman (2015). "TEER measurement techniques for in vitro barrier model systems." J Lab Autom 20(2): 107-126. Stenberg, P., U. Norinder, K. Luthman and P. Artursson (2001). "Experimental and computational screening models for the prediction of intestinal drug absorption." J Med Chem 44(12): 1927-1937. Teo, W. E. and S. Ramakrishna (2006). "A review on electrospinning design and nanofibre assemblies." Nanotechnology 17(14): R89-r106. Thukral, D. K., S. Dumoga, S. Arora, K. Chuttani and A. K. Mishra (2014). "Potential carriers of chemotherapeutic drugs: matrix based nanoparticulate polymeric systems." Cancer Nanotechnology 5(1): 3. Tran, T. H., T. Ramasamy, J. Y. Choi, H. T. Nguyen, T. T. Pham, J.-H. Jeong, S. K. Ku, H.-G. Choi, C. S. Yong and J. O. Kim (2015). "Tumor-targeting, pH-sensitive nanoparticles for docetaxel delivery to drug-resistant cancer cells." International Journal of Nanomedicine 10: 5249-5262. Tsvetkova, B. P. I. Z., A.; Peikov, P (2012). "High Performance Liquid Chromatographic Assay of Indomethacin and its Related Substances in Tablet Dosage Form." Int. J. Pharm. Pharm. Sci. 4 (Supplement 3): 549-552. Vehring, R. (2008). "Pharmaceutical Particle Engineering via Spray Drying." Pharmaceutical Research 25(5): 999-1022. Yassin, A. E. B., M. K. Anwer, H. A. Mowafy, I. M. El-Bagory, M. A. Bayomi and I. A. Alsarra (2010). "Optimization of 5-flurouracil solid-lipid nanoparticles: a preliminary study to treat colon cancer." Int. J. Med. Sci 7(6): 398-408.

185

Yu, D.-G., G. R. Williams, J.-H. Yang, X. Wang, J.-M. Yang and X.-Y. Li (2011). "Solid lipid nanoparticles self-assembled from electrosprayed polymer-based microparticles." J Mater Chem 21(40): 15957-15961. Zhang, C.-L. and S.-H. Yu (2014). "Nanoparticles meet electrospinning: recent advances and future prospects." Chemical Society Reviews 43(13): 4423-4448. Zhang, H. and S. H. Neau (2002). "In vitro degradation of chitosan by bacterial enzymes from rat cecal and colonic contents." Biomaterials 23(13): 2761-2766.

186

5. Triaxial electrospinning for the creation of self- assembled nanostructures

5.1 Introduction

Liposomal drug delivery systems

Liposomal drug delivery systems have been of paramount importance in developing the first generation of nanomedicines: 15 of the 43 commercial nanomedicines approved so far are liposomal systems (Weissig, Pettinger et al. 2014, Bulbake, Doppalapudi et al.

2017). Liposomes were described for the first time in 1965 (Bangham, Standish et al.

1965, Sessa and Weissmann 1968). They are defined as spherical vesicles composed of one or more lipid bilayers and containing an aqueous core, with a size which varies between 30 nm and several micrometres. Liposomes can be classified according to their size and number of bilayers (Akbarzadeh, Rezaei-Sadabady et al. 2013). The encapsulation of drugs in a liposomal formulation changes their pharmacokinetic properties, which often leads to an improvement of the therapeutic index and reduction in adverse effects (Bozzuto and Molinari 2015). The nanoscale size of liposomes is also advantageous as it allows them to exploit the EPR effect and to accumulate inside solid tumours, which maximises the effective concentration at the tumour site and minimises toxic effects (Gabizon, Catane et al. 1994). Probably the most famous liposomal formulation is Doxil, the first nanomedicine approved by the FDA in 1995 (Barenholz

2012).

187

Micelles as drug delivery systems

Micelles are formed by the self-assembly of amphiphilic molecules (Noguchi and Takasu

2001). Micelles are considered closed lipid monolayers, usually with a fatty acid core and a polar surface. They are capable of delivering hydrophobic and hydrophilic small drug molecules or macromolecules (peptides, proteins), improving the stability, release and pharmacokinetics of the entrapped drugs (Trinh, Joseph et al. 2017). One of the main advantages of micelles is their small size, which ranges from 2 to 20 nm: such small sizes are very difficult to achieve when using other drug carriers. However, micelles are only stable above a concentration termed the CMC (critical micelle concentration). There are numerous methods to prepare micelles, including emulsion, solvent evaporation and dialysis methods (Cholkar, Acharya et al. 2017). Micelles have been widely investigated for their use in cancer therapeutics. (Oerlemans, Bult et al. 2010) (Kedar, Phutane et al.

2010)

Often, block copolymers are employed to create micelles. Such polymers are composed of separate hydrophobic and hydrophilic parts. One such family of materials which has been widely explored are the poloxamers, well-known non-ionic triblock copolymers composed of a central hydrophobic chain of polypropylene oxide (PPO) flanked by two hydrophilic chains of polyethyleneoxide (PEO). The poloxomer chemical structure is shown in Figure 5-1 (Parisi, Puoci et al. 2014). The length of the polymer blocks can easily be customized, creating different poloxamers with slightly different properties

(Adhikari, Goliaei et al. 2016). In this project, the poloxamer P188 was used. It has a molecular weight of 8,400 Da and 80% PEO content (Kabanov, Batrakova et al. 2002).

PEO Unit PPO Unit PEO Unit

(Hydrophilic) (Hydrophobic) (Hydrophilic) Figure 5-1. The generic chemical structure of poloxamers, in which a=20,b=27

188

Formulation background

In this chapter, we will aim to develop bi-compartment Janus particles for photo- chemotherapy. First, self-assembled liposomes comprising dioleoyl phosphatidylcholine

(DOPC) (Figure 5-2) will be explored, building on the work reported by Yu et al. (Yu,

Branford-White et al. 2011). The key difference between this work and that of Yu lies in the phospholipid used: the previous work used phosphatidylcholine, but here we use

DOPC, a high purity lipid that has been utilised in commercial liposomal formulations such as DepoDur and DepoCyt (Li, Wang et al. 2015).

Figure 5-2. The chemical structure of DOPC

First, we will electrospin poly(vinylpyrrolidone) (PVP), DOPC and the anti-cancer drug carmofur (C) to create fibres in which the molecules of PVP, DOPC and C interact with each other (the electrospinning process is explained in detailed in the materials and method section of this chapter). Upon the addition of water, the PVP matrix will expand as the water will strongly interact with the carbonyl groups of the polymer, allowing the

PC and C molecules to become mobile as they become hydrated. Due to the hydrophobic nature of the latter however, they will become concentrated in the PVP matrix. The PC molecules will then co-aggregate to form liposomes, simultaneously entrapping carmofur. The liposomes will finally be released into the aqueous medium when the PVP matrix terminates its dissolution.

Second, the fabrication of drug-loaded micelles containing (haematoporphyrin) HP will be undertaken. These micelles will be electrospun into PVP-based fibres, and subsequently released from the polymeric matrix upon addition to an aqueous media. 189

HP is a photosensitiser which, when activated by light, generates reactive oxygen species (ROS). HP has been previously reported to successfully supress the growth of human bladder cancer cells (Kim, Lee et al. 2018) and to induce apoptosis in glioma cells and head and neck cancer cells (Kim, Lim et al. 2014).

Janus particles assembled from electrospun fibres

The creation of Janus particles from the liposomes and the micelles will be performed building on the work of Yu and co-workers, who were able to create inorganic/organic

Janus nanoparticles in situ from fibers prepared using triaxial electrospinning with a

Janus core.(Yu, Li et al. 2018). To do this, two core fluids were employed, one comprising silver nanoparticles and PVP and the second PVP and phosphatidylcholine. This resulted in snowman-like Janus particles where one half comprised a liposome and the other an Ag nanoparticle.

This chapter will attempt to employ the same concept to fabricate a photo-chemotherapy formulation, using a chemotherapeutic agent (carmofur) entrapped in a liposome, while a photosensitiser (hematoporphyrin, previously used in Chapter 2), will be encapsulated in micelles (Figure 5-3).

HP-loaded micelle

Carmofur-loaded liposome

Figure 5-3. Schematic representation of the proposed self-assembled Janus structure

190

Multiaxial electrospinning

The use of multi-axial electrospinning for the creation of complex electrospun fibres is widely reported in the literature, as it provides the opportunity to fabricate multicomponent composites (Khalf and Madihally 2017). Multiaxial electrospinning uses multiple fluids to create multifunctional systems via a series of nested needles.

Triaxial electrospinning

Similarly, to coaxial electrospinning, triaxial electrospinning also uses a spinneret with multiple needles, but with three nozzles instead of two. The two inner nozzles can either be concentrically arranged or side-by-side: a schematic representation can be found in

Figure 5-4. Compared to co-axial electrospinning, triaxial electrospinning is not widely reported in the literature with only 14 papers published by 2018, probably because of the complexity of the process. Each of the three solutions being processed will have different properties, which can be challenging when processing them together under the same conditions.

Figure 5-4. Configuration of triaxial spinneret. Left: two inner nozzles concentrically arranged; right: two inner nozzles side-by-side

The use of triaxial electrospinning has been explored for drug delivery applications in a few studies. For example, Han reported nanofibres composed of PCL and PVP entrapping two colour dyes, one in the core and the other in the outer layer (keyacid blue and keyacid uranine) to simulate drug release profiles. The triaxial set up of the fibres provided a rapid release from the outer layer and a sustained release from the fibre core, while the intermediate layer served as a barrier to prevent drug leaching from the core.

191

These characteristics offer a formulation approach to achieve the release of drugs for both short and long-term treatment. The fibres provided a dual release modality in which release from the core was 24 times slower than from analogous coaxial fibres (Han and

Steckl 2013).

In other work, Yu et al. used triaxial fibres to achieve zero-order drug release over 20h

(Yu, Li et al. 2015). Their method consisted in using ethylcelulose as the polymer matrix for the three layers, but a different concentration of ketoprofen was used in each fluid.

Yu demonstrated that, contrary to monolithic fibres, no burst release was seen from the triaxial nanofibres, thus overcoming one of the major disadvantages of monolithic and co-axial fibres. Additionally, the drug loading was released at a constant rate over 20h

(Yu, Li et al. 2015).

5.2 Aims and objectives

The combination of photodynamic therapy and chemotherapy which can be obtained using Janus particles led to promising results, as described in Chapter 2. However, high- quality Janus materials could only be obtained using the fast-dissolving polymer PVP.

The creation and scaled-up fabrication of self-assembled nanomaterials using PVP- based dispersions as sacrificial templates was shown to be feasible in Chapters 3 and

4. This chapter brings together both concepts to investigate the use of the self-assembly process to create Janus structures as potential photochemotherapy formulations. The aims are as follows:

• Generate multi-compartment PVP based fibres by electrospinning

• Characterise the physicochemical properties of the fabricated fibres

• Create Janus particles via self-assembly of the electrospun templates

• Characterise the self-assembled particles

• Assess the in-vitro drug release and cytotoxicity of the formulations created

192

5.3 Materials and methods

Polyvinylpyrrolidone (average molecular weight 360,000 kDa), hematoporphyrin and poloxamer 188 (Kolliphor® P188) were purchased from Sigma Aldrich. 1,2-dioleoyl-sn- glycerol-3-phosphocholine (DOPC) was obtained from Lipoid. Ethanol was purchased from Fisher Scientific. Carmofur was provided by Cambridge Scientific. Dicloromethane

(DCM) was purchased from Alfa Aesar. All chemicals were used as supplied, and all water used distilled before use.

5.3.1 Formation of polymer micelles

Hematoporphyrin loaded micelles were prepared as described by Barichello (Barichello et al., 1999). Briefly, poloxamer P188 (10% w/v) and HP (25 µg/mL) were added to a mixture of dichloromethane (DCM) and ethanol (ratio by volume 2:1). To ensure complete dissolution, the solution was left stirring for 4 h. The DCM and ethanol were then completely evaporated under vacuum using a rotary evaporator (Buchi R-200,

Flawil) using a vacuum of 300 milibar, a rotation speed of 120 rpm, and a water bath at

70 oC. The final composition was a red-brown film, which was dissolved in 20 mL of deionized water. The micellar suspension was then sonicated for 10 min and filtered through a 0.22 µm filter to eliminate the non-entrapped drug. The HP content in the micelles was evaluated spectrophotometrically using an UV-Vis method (Cary Vis 1000 instrument) at a wavelength of 395 nm. The concentration of HP was calculated using a calibration curve which was previously prepared in water. Empty micelles were prepared using the same protocol, but without adding HP. The HP encapsulated in micelles was quantified by measuring its absorption at 395 nm after extraction in absolute ethanol.

193

5.3.2 Electrospinning solutions and parameters

A series of solutions was prepared as follows:

Unloaded materials:

Shell solution: A 10% w/v PVP solution was prepared by dissolving the appropriate

amount of PVP in ethanol and stirring overnight.

Core solution 1: 1 g of PVP was weighed and added to 10 ml of the unloaded micellar

solution previously prepared.

Core solution 2: A 10% w/v PVP and 2% w/v DOPC solution was prepared by adding

the appropriate amounts of each to chloroform

Drug-loaded materials:

Shell solution: A 10% w/v PVP solution was prepared by dissolving the appropriate

amount of PVP in ethanol and stirring overnight.

Core solution 3: 1 g of PVP was weighed and added to 10 ml of the drug-loaded

micellar solution previously prepared.

Core solution 4: A 10% w/v PVP and 2% w/v s DOPC solution was prepared by adding

the appropriate amount of both polymers to chloroform with 0.5% w/v of carmofur was

added.

Using the solutions described above, fibres were fabricated as detailed in Table 5-1.

194

Table 5-1. Electrospinning fibre composition

Shell Self-assembled Fibre ID Core solution used solution structure

PVP-DOPC-C No Core solution 4 Liposomes-C

V_PVP-DOPC No Core solution 2 V-liposomes

M_HP-PVP No Core solution 1 Micelles-HP

V_M_HP-PVP No Core solution 3 V-micelles

Core solution 3 + 3A-F Yes Janus particles Core solution 4

Core solution 1 + V_3A-F Yes V-Janus particles Core solution 2

Monoaxial electrospinning of each electrospinning solution was carried out at 16 kV with help of a high voltage DC power supply (0-35kV, 0-1 mA, HCP 35-35000, FuG

Elektronik). The polymer solution was ejected using a plastic syringe (Terumo) connected to a syringe pump with a flow rate of 1 ml h-1 (KDS100, Cole Parmer). The relative humidity during the experiments was between 30-45%, a temperature of 24-26

ºC and a tip to collector distance of 10 cm. The spinneret comprised a stainless needle

(20 gauge, inner diameter 0.61mm, outer diameter 0.91mm).

Triaxial electrospinning was performed with a bespoke triaxial spinneret which was kindly provided by Professor Deng-Guang Yu from the University of Shanghai for Science and

Technology. The spinneret is composed of three stainless steel needles, with an arrangement of needles seen as Figure 5-5.

195

Figure 5-5. Schematic representation of the triaxial spinneret

Three syringe pumps (KDS100, Cole Parmer) and a power supply (HCP 35-35000, FuG

Elektronik) were used for the process. The collector was composed of a metal plate of

20 x 20 cm wrapped with aluminium foil with a tip-to-collector distance of 10 cm.

Optimised flow rate consisted in using 0.8 ml h-1 for the shell, and 0.1 ml h-1 for both cores. The electrospinning process was conducted at ambient temperature (18-25 oC) and all products were stored in a glass desiccator.

The fibres were characterised using SEM, XRD, DSC, and FTIR measurements, as described in Section 2.3. In vitro drug release and cytotoxicity experiments were carried out as described in the experimental section of Chapter 2.

The release of micelles from the PVP matrix was performed by adding 10 mL of water to

10 mg of the fibre mat. The self-assembly of liposomes and Janus particles was done similarly, by weighing 10 mg of the fibres and dissolving them in 10 mL of distilled water.

The resultant nanoparticles were characterised by TEM and DLS.

196

5.4 Results and discussion

5.4.1 Micelle preparation

HP-loaded micelles were created using the thin-film evaporation method. After solvent evaporation, a red-brown film was observed to form, and following the addition of water the film was completely dispersed to give a brown-coloured suspension (Figure 5-6) HP is not soluble in water and brown in colour; hence, the dissolution of the film confirms the entrapment of the drug into the micelles.

Figure 5-6. Formation of micelles upon water addition. Left: the dried film; right: the micelles after the addition of water.

5.4.1.1 TEM of self-assembled micelles

To assess if micelles were formed during the thin-film evaporation method, TEM images were taken of the empty and drug-loaded micelles (Figure 5-7). The TEM data show the micelles to have roughly spherical morphologies, which corresponds to that reported in the literature (Chen, Yan et al. 2014). The micellar sizes were calculated using the

ImageJ software , and found to be 22.7 ± 4.9 nm and 21.4 ± 5.2 nm, for the drug-loaded and drug-free micelles, respectively (Chen, Yan et al. 2014).

197

a) b) Figure 5-7. TEM images of a) unloaded and b) HP- loaded micelles

5.4.2 Electrospinning

After the micelles were prepared, they were processed by electrospinning to determine if this has any effects on them. Electrospinning was first attempted with the micellar solution alone, but this process was not successful (Figure 5-8). It is known that for electrospinning to be achieved, the liquid being processed must have a minimum viscosity (Zhenyu and Ce 2013), which clearly the micellar suspension did not possess.

To increase the viscosity of the solution, PVP was added. A 5% w/v PVP was first assessed but was also unsuccessful: a film was formed because of there being insufficient polymer entanglement to yield fibres (Figure 5-8).

Figure 5-8. The results of failed electrospinning a micellar suspension with 5% w/v PVP

198

In contrast, a 10% w/v PVP solution containing micelles led to smooth and uniform fibres being obtained (M_HP-PVP), as can be seen in Figure 5-9. The diameter of the fibres is

0.52 ± 0.10 µm.

Figure 5-9. Fibres prepared from a micellar suspension with 10% w/v PVP (M_HP-PVP)

The resulting fibres were dissolved to study if the process affected the morphology of the micelles. The TEM images shown in Figure 5-10 does not seem to differ in morphology from those obtained of the micelles before electrospinning. The sizes of the particles after electrospinning were calculated to be 20.1 ± 4.8 nm and 22.4 ± 3.5 nm for micelles-

HP and V-micelles, respectively. These are identical to those noted prior to spinning.

a) b)

Figure 5-10. TEM images of a) micelles-HP and b) V-micelles

199

After achieving the fabrication of micelles and successfully electrospinning them into fibres using PVP as a carrier polymer, the next step consisted of creating the

DOPC/carmofur loaded fibres (V_PVP-DOPC and PVP-DOPC-C). In the SEM images of these, shown in Figure 5-11, there appear to be two populations of fibres (large and small) in both systems, resulting in a large standard deviation in the diameters: the non- drug loaded fibres have mean diameters of 0.43 ± 0.18 µm, while the carmofur- loaded fibres are slightly larger at 0.52 ± 0.09 µm. For both formulations, however, it is clear that good quality and smooth fibres were produced. Similar polydispersity of PVP-based fibres has also been noted in previous work done in the group (Askin 2015).

a b

c d

Figure 5-11. SEM images of DOPC-loaded electrospun fibres and their diameters: a,b) V_PVP-DOPC and c,d) PVP-DOPC-C

200

Experiments were next performed to assess the formation of liposomes after adding the

V_PVP-DOPC and PVP-DOPC-C fibres to water. TEM images (Figure 5-12) clearly show spherical vesicles of 30 ± 9.4nm assembled from V_PVP-DOPC fibres and 45.8 ±

4.5 nm from PVP-DOPC-C.

a) b)

c) d)

Figure 5-12. TEM images from the liposomes self-assembled from a,b) V_PVP-DOPC and c,d) PVP-DOPC-C fibres

5.4.3 Triaxial electrospinning

Subsequently, triaxial electrospinning was assessed. Three flow rates were assessed to see which provided the most stable Taylor cone. The first flow rate explored was 1 mL/h for the shell, with 0.3 mL/h for each core solution, this is as a reduction of 70-80% for the core flow rate has been reported in the literature (Han and Steckl 2013, Khalf and 201

Madihally 2017). However, the Taylor cone formed was not stable and some dripping of the polymeric solution was observed. This led to fusing of the fibres on the collector

(Figure 5-13).

Figure 5-13. SEM images of fibres produced from triaxial electrospinning at a flow rate of 1 mL/h for the sheath and 0.3 mL/h for the core fluids.

The flow rate of the core fluids was therefore decreased to 0.1 ml/h. At this speed the

Taylor cone seemed to be stable but looking at the SEM images of the products two populations of fibres were clearly observed, with a few large fibres and a larger number of very narrow fibres, as observed in Figure 5-14. This suggests that the Taylor cone was not as stable as initially thought (Agrahari, Agrahari et al. 2017).

Figure 5-14. Fibres prepared by triaxial electrospinning using a flow rates of 0.1 ml/h for the core fluids and 1 ml/h for the shell.

202

As the flow rate affects the stability of the cone jet (Khajavi and Abbasipour 2012), to improve the process and to achieve more uniform fibres, the flow rate of the shell was decreased to 0.8 ml/h. This approach seemed to yield good results and the resultant fibres were more uniform as seen in Figure 5-15 and Figure 5-16.

Fibres obtained from triaxial electrospinning using the blank vehicles only (V_3A-F) can be seen in Figure 5-15; the average diameter of these fibres is 1.45 ± 0.48 µm. Fibres appear to be smooth, with no bead-on-fibre morphology. The internal structure is composed by two different fluids and can be appreciated in Figure 5-15 (c,d).

a) b)

c) d)

Figure 5-15. a) SEM image of V_3A-F; b) size distribution of the fibres in V_3A-F; c,d) TEM images of V_3A-F.

203

The drug-loaded 3A_F fibres can be seen in Figure 5-16. The average size of the fibres is 1.18 ± 0.27 µm. The fibres are smooth, cylindrical and relatively monodisperse. A clear core-sheath structure can be seen in the TEM images in Figure 5-16 (c,d); where three compartments can clearly be identified. An exterior shell contains two cores, comprising:

1) PVP + DOPC + carmofur and 2) PVP + micelles-HP.

a) b)

c) d)

Figure 5-16.Data on the 3A-F fibres: a) an SEM image ; b) a frequency graph showing the fibre diameters; c,d) TEM images

204

The products obtained after self-assembly of the triaxial fibres can be seen in Figure

5-17. The morphology of the particles appear Janus-like, and most closely resemble snowman-like particles reported in the literature (Zhang, Liu et al. 2013). We assume that in the snowman Janus particles the larger side of the particles corresponds to the liposomes (size: 36.5 ± 5.1 nm), while the smaller particles represent micelles (21.1 ±

3.46 nm). This is consistent with the sizes noted for the micelles and liposomes found above.

500 nm

Figure 5-17. TEM images of the Janus particles self-assembled from triaxial electrospun fibres.

5.4.3.1 Entrapment efficiency and hydrodynamic size

The entrapment efficiency was calculated for the three nanoscale structures resulting from dissolving the electrospun fibres, results are shown in Table 5-2. The EE will depend on the hydrophobicity of the drug, amongst other factors such as the aqueous volume, surface are and preparation method (Ong, Ming et al. 2016). There is a limited amount of research regarding liposomes loaded with carmofur, however as it is a poorly water-soluble drug it is expected to be encapsulated in the lipophilic part of the liposome.

This is contrary to the analogous 5-FU, which is often reported to be entrapped in the aqueous compartment of the liposome. 5-FU encapsulation efficiencies are reported to be between 10-15% (Fresta, Villari et al. 1993, Nii and Ishii 2005). The liposomes

(liposomes-C) created using from the electrospun template (PVP-DOPC-C) possess an

205

EE of 56.2± 3.7 %. The size of liposomes-C determined by DLS was 113.9 ± 1.7 nm, comparable to values obtained from previous work in the group which reported liposomes assembled from fibres entrapping 5-FU to have a size of 133 ± 2 (Askin,

2015). This size is rather smaller than the TEM size calculated above, as a result of DLS measuring the hydrodynamic size (Malvern 2011).

Micelles-HP present a hydrodynamic diameter size of 56.6 nm,and have a high entrapment efficiency of 92.1%. This is similar to the size of micelles encapsulating HP reported by Yang (Yang, Chen et al. 2013) and the entrapment efficiency reported by

(Yang, Chen et al. 2010).

The Janus particles appear to be the largest of the self-assembled structures, with a hydrodynamic size of 189 ± 3 nm. This agrees well with them comprising a composite of liposomes-C and micelles-HP, which have hydrodynamic sizes which add up to approx.

170 nm. The entrapment efficiency is higher in the micelles entrapping HP, than the liposomes entrapping carmofur.

Table 5-2. DLS size and entrapment efficiency of the self-assembled structures. Data shown as mean ± SD(n=3)

Self-assembled DLS size (nm) Template EE% structure and PDI

DOPC-PVP- C Liposomes-C 113.9 ± 1.7 (0.27) 56.2 ± 3.7

V_DOPC-PVP V-liposomes 108.5 ± 1.3 (0.31) NA

M_HP-PVP Micelles-HP 56.6 ± 1.2 (0.39) 92.1 ± 2.3

V_M_HP-PVP V-micelles 62.5 ± 7.1 (0.38) NA

Janus particles 188.8± 2.7 (0.46) C: 41.5 ±7.2 3A-F HP: 79.2 ± 4.7

V_3A-F V-Janus particles 150.5 ± 5.1 (0.51) NA

206

5.4.3.2 FTIR

To gain an understanding of the interactions between the polymers and drugs, FTIR was employed. The spectra of the raw materials are given in Figure 5-18. For DOPC, the

C=O band of the ester group can be detected at a wavenumber of 1733 cm-1, and the C-

O stretch is visible at 1172 cm-1. DOPC also possesses bands at 2919 cm-1 and 2851 cm-1, representing =C-H and –C-H stretches respectively (Zhang, Zhang et al. 2016). In the case of P188, the FTIR spectra show two main absorption bands at 2888 cm-1 (C-H stretching) and 1113 cm-1 (C-O group stretching) (Manikandan, Kannan et al. 2013). The hematoporphyrin spectrum displays N-H stretching at 3311 cm-1, and carboxylate stretches at 1439 cm-1 and 1646 cm-1 (Sardar, Sarkar et al. 2013). The spectra of PVP shows a distinctive C=O band at 1660 cm-1 and a C-N stretch at 1290 cm-1, while carmofur has peaks between 1660-1770 cm-1 and a broad peak at 1501 cm-1 which could all be attributed to C=O groups.

Figure 5-18. FTIR spectra of the raw materials utilised for the triaxial electrospinning process.

207

The spectra of the fibres with and without drug are very similar. This is because the amount of drug in the fibres is relatively small (the drug loading of M_HP_PVP is 0.025 w/w), probably below the detection limit. Comparing the V_M-PVP and M_HP-PVP formulations, the key difference is that the peak at 1113 cm-1 from P188 is more visible in V_M_HP-PVP. In the M_HP-PVP system drug- this peak is less visible - probably because of interactions between the drug and the P188. Also, the peak representing the

P188 C-O group stretching at 1113 cm-1 in the raw material has shifted to 1287 cm-1 in the 3A-F formulation. Meanwhile, in the DOPC containing fibres, the peak at 1733 cm-1 from DOPC can be observed. The C-N stretching in the PVP-DOPC fibres can been seen to have shifted from 1290 cm-1 to 1280 cm-1 in the formulations.

Figure 5-19. FTIR spectra of the triaxial drug-loaded electrospun fibres, M_HP-PVP, 3A-F and DOPC-PVP-C. 208

It is hard to evidence intermolecular interactions from the spectra in Figure 5-19, owing to their complexity, but the shifts in the positions of the C-O and C-N stretches do confirm interactions between the polymers; unfortunately, the interactions between polymer and drug cannot be determined from this technique.

5.4.3.3 XRD

The XRD patterns of the raw materials are presented in Figure 5-20. P188 is a semi- crystalline powder showing two characteristic Bragg reflections at 19º and 23.8º; this agrees closely with the literature (Vidal, Castro et al. 2014). Hematoporphyrin and DOPC show only broad humps in their diffraction patterns, suggesting the materials to be amorphous. Meanwhile, PVP displays a broad halo without any peaks confirming its amorphous nature. On the contrary, the carmofur pattern displays sharp Bragg reflections, suggesting it is a crystalline material.

Figure 5-20. XRD patterns of the raw materials.

209

The XRD patterns of the composite fibres do not show any Bragg reflections, only the broad humps typical of amorphous materials. (Figure 5-21) There is an extra peak at approximately 5o in the spectra of 3A-F if compared with V_3A-F, which is due to the sample holder used for the sample. P188 concentration in the fibres is probably below the level of detection. Hence, it can be concluded that HP remained amorphous and carmofur lost its crystalline form during the electrospinning process, as previously reported in the literature (Ren, Liu et al. 2008).

Figure 5-21. XRD patterns of blank fibres: V_M-PVP, V_3A-F, V_DOPC-PVP and drug-loaded fibres M_HP-PVP, 3A-F, DOPC-PVP-C.

210

5.4.3.4 DSC

To corroborate the information given by XRD, DSC was performed. The thermograms of the raw materials (see Figure 5-22) show that DOPC has no melting point, confirming the amorphous nature of this material. In contrast P188 possess a sharp endothermic peak at around 55 OC, which corresponds to the reported melting point of the material

(El-Habashy, Allam et al. 2016). PVP, HP, and carmofur thermograms can be found in

Chapter 2, Section 2.5.3. As PVP and HP are amorphous materials, its thermograms do not present any endothermic melting point. Carmofur, on the other side, presents a sharp

melting point around 110-115 oC. HP.

ExoUp

Figure 5-22. DSC thermograms of DOPC and P188.

The thermograms of the electrospun fibres are depicted in Figure 5-23. The only visible peak is the P188 melting peak, although this is very small. The broad hump seen between 60-200 oC corresponds to the evaporation of the water absorbed by the PVP, as the polymer is a very hygroscopic material. There is no difference between the vehicle and the drug-loaded fibres. No meting peaks can be observed arising from either 211

carmofur or HP, suggesting that the drug exists in an amorphous form in these formulations, however it is important to take into account that the drug loading of HP is very low, which might be below the limit of detection. These DSC thermograms are consistent with those of other electrospun PVP/DOPC systems previously reported by the group (Askin 2015).

Exo Up Exo Up

Figure 5-23. DSC thermograms of the triaxial fibres.

212

5.4.3.5 In-vitro drug release

The amount of drug released from the three self-assembled materials was explored, and the results are displayed in Figure 5-24. Sustained release is expected from the micelles and liposomes, since the drug is entrapped inside amphiphilic matrices (Kumbar,

Laurencin et al. 2014). It has been widely reported in the literature that micelles can retain drugs in their interior for long periods of time. For example, micelles have been reported to achieve sustained release of doxorubicin for 35 h (Deng, Jiang et al. 2012), and another study reported micellar formulations releasing just 30% of fluorescein sodium and trypan blue at the end of 60 days (Liyan, Jianxiang et al. 2007). Meanwhile,

DedoDur (a morphine commercialized liposomal formulation) has been proven to give drug release over up to 62 h (Alam and Hartrick 2005).

Figure 5-24. Drug release from single fluid fibres and triaxial fibres. Mean average shown (n=3) with error bars representing S.D.

213

It can be seen that the encapsulated drug is released at the same rate, regardless of whether is coming from single-component or Janus nanostructures. The release of HP from micelles is slower than the release of carmofur from liposomes, with around 95% of the carmofur content released from the liposomes after 8 h, while approximately 60% of hematoporphyrin loading is released after 12 h.

The reasons for the difference in the release rates can be explained because small non- polar molecules such as carmofur (MW: 257 Da) are able to move across the lipid bilayer of the liposomes fairly rapidly to escape into solution (Papahadjopoulos, Nir et al. 1972).

In addition, DOPC is an unsaturated lipid, meaning there are C=C double bonds in the tail; these bonds create a “kink” which reduces the strength of the hydrophobic interactions among the tails and makes the structure more fluid and permeable (Freeman

2011). To create micelles, copolymers with short hydrophobic tails (e.g. P188) need to be used, resulting in a tightly packed core enhancing hydrophobic interactions between the drug and the hydrophobic part of the copolymer. This makes release of the drug more difficult (Imran, Shah et al. 2018). Also, HP’s molecular weight (599 Da) is double that of carmofur, and will also contribute to the slower release of HP into the aqueous medium.

The difference in release patterns between carmofur and hematoporphyrin can be advantageous as it might offer a good combination for anticancer therapy: a treatment can be envisaged where carmofur initially kills most of the cancer cells, and then the drug-resistant cells are eradicated by photodynamic therapy via the hematoporphyrin.

The similarity of the drug release profiles between the Janus self-assembled structure and the micelles and liposomes alone means that tri-axial electrospinning does not affect the size or drug release of the self-assembly of these vesicles.

214

5.4.3.6 Cell viability studies

The next step in characterising the formulations was to perform in-vitro cell viability studies; this was undertaken using the Cell-titre Glo luminescence assay previously described in Chapter 2. Hematoporphyrin, like RB, is a photosensitiser which needs light to get activated. The results of cytotoxicity assays can be seen in Figure 5-25.

Figure 5-25. Cell viability of A549 cells exposed to the self-assembled structures. Results are shown as mean ± S.D. (n=3) with three replicates each

215

All of the carmofur-loaded formulations perform equally, and significantly reduce A549 cell viability. The free drug control (which is not encapsulated in any vehicle) performs on a par with the other formulations, suggesting that the self-assembled structures do not have a significant effect on the delivery of carmofur or hematoporphyrin to the A549 cells after 24h. The vehicle does not seem to have any difference in viability compared to the untreated cells control, indicating that the excipients do not have any effect on the viability of the cells.

Carmofur is known for being cytotoxic activity towards cancer cells (Domracheva,

Muhamadejev et al. 2015): this is clearly seen in Figure 5-25. To activate the phototoxicity of HP, red light was shone into the plates. No difference is seen in viability whether applying light or not. This is at first sight surprising as HP should have phototoxicity when activated by light. However, these findings are believed to arise as a result of the experimental protocol used here: other studies have irradiated the cells and incubated them for further 24 h, while our measurements were done immediately after irradiation.

In addition, literature studies also have irradiated the cells for a longer period of time (60 min), or used different light sources and wavelengths (Kim, Lim et al. 2014, Kim, Lee et al. 2018, Yuan, Li et al. 2018). Hence, further experiments are needed to identify the lack of cell death seen in this work and different drug loading should be experimented with.

The cytotoxicity seen with the Janus particles thus seems to be solely from the encapsulated carmofur.

216

5.5 Conclusions

Fibres composed of PVP, DOPC, and carmofur and PVP fibres encapsulating HP-loaded micelles were first prepared by single fluid electrospinning. Triaxial electrospinning was then used to fabricate core-sheath fibres with a Janus core, the two sides of which comprised PVP-DOPC-C and PVP with HP-loaded micelles, while the sheath fluid was composed of PVP. The resultant three-fluid fibres were then used as a template to create

Janus particles composed of liposomes encapsulating carmofur and micelles encapsulating HP.

Liposomes, micelles and Janus particles created from the fibres were visualized using

TEM to confirm the morphology of these structures. The size was also calculated by using DLS. The liposomes possess a larger hydrodynamic diameter than the micelles, and Janus particles presented the greatest size (approximately equal to the sum of the two components). The calculated drug entrapment efficiencies were higher than 50% when using the single fluid fibres, while the Janus particles presented EEs of approx.

42% and 79% for carmofur and HP, respectively. Drug release from the self-assembled structures was also studied, and a release up to 12h was achieved from the micelles.

The release of HP was slower than that of carmofur, and the release does not seem to vary whether the drug is freed from the triaxial or monolithic fibres. Cytotoxicity studies were also performed, where the fibres performed with the same cytotoxicity as the pure drug carmofur, however the results indicated that the photocytotoxicity of HP was null.

Overall, this chapter provides proof-of-concept that three-fluid electrospinning can be used to generate self-assembled Janus particles with the two sides loaded respectively with a chemotherapeutic agent and photosensitzer. These are worthy of further exploration for the generation of novel anti-cancer medicines.

217

5.6 References

Adhikari, U., A. Goliaei, L. Tsereteli and M. L. Berkowitz (2016). "Properties of Poloxamer Molecules and Poloxamer Micelles Dissolved in Water and Next to Lipid Bilayers: Results from Computer Simulations." J Phys Chem B. 120(26): 5823-5830. Aggarwal, U., A. K. Goyal and G. Rath (2017). "Development and characterization of the cisplatin loaded nanofibers for the treatment of cervical cancer." Mater Sci Eng C Mater Biol Appl 75: 125-132. Agrahari, V., V. Agrahari, J. Meng and A. K. Mitra (2017). Chapter 9 - Electrospun Nanofibers in Drug Delivery: Fabrication, Advances, and Biomedical Applications. Emerging Nanotechnologies for Diagnostics, Drug Delivery and Medical Devices. Boston, Elsevier: 189-215. Akbarzadeh, A., R. Rezaei-Sadabady, S. Davaran, S. W. Joo, N. Zarghami, Y. Hanifehpour, M. Samiei, M. Kouhi and K. Nejati-Koshki (2013). "Liposome: classification, preparation, and applications." Nanoscale Research Letters 8(1): 102-102. Alam, M. and C. T. Hartrick (2005). "Extended-Release Epidural Morphine (DepoDur™): An Old Drug with a New Profile." Pain Practice 5(4): 349-353. Aldred, E. M., C. Buck and K. Vall (2009). Chapter 39 - Chemotherapy. Pharmacology. Askin, S. (2015). Liposomal Drug Delivery Systems from Electrospun Polymer Nanofibres, UCL School of Pharmacy: 1-44. Bangham, A. D., M. M. Standish and J. C. Watkins (1965). "Diffusion of univalent ions across the lamellae of swollen phospholipids." J. Mol. Biol 13(1): 238-IN227. Barenholz, Y. (2012). "Doxil(R)--the first FDA-approved nano-drug: lessons learned." J Control Release 160(2): 117-134. Bozzuto, G. and A. Molinari (2015). "Liposomes as nanomedical devices." Int. J. Nanomed 10: 975-999. Brown, J. M. and A. J. Giaccia (1998). "The Unique Physiology of Solid Tumors: Opportunities (and Problems) for Cancer Therapy." Cancer Research 58(7): 1408-1416. Bulbake, U., S. Doppalapudi, N. Kommineni and W. Khan (2017). "Liposomal Formulations in Clinical Use: An Updated Review." Pharmaceutics 9(2). Chen, F. Y., J. J. Yan, H. X. Yi, F. Q. Hu, Y. Z. Du, H. Yuan, J. You and M. D. Zhao (2014). "TNYL peptide functional chitosan-g-stearate conjugate micelles for tumor specific targeting." Int J Nanomedicine 9: 4597-4608. Chen, J., B. F. Miller and A. V. Furano (2014). "Repair of naturally occurring mismatches can induce mutations in flanking DNA." eLife 3: e02001. Cholkar, K., G. Acharya, H. M. Trinh and G. Singh (2017). Chapter 1 - Therapeutic Applications of Polymeric Materials. Emerging Nanotechnologies for Diagnostics, Drug Delivery and Medical Devices. Boston, Elsevier: 1-19. Deng, C., Y. Jiang, R. Cheng, F. Meng and Z. Zhong (2012). "Biodegradable polymeric micelles for targeted and controlled anticancer drug delivery: Promises, progress and prospects." Nano Today 7(5): 467-480. Domracheva, I., R. Muhamadejev, M. Petrova, E. Liepinsh, A. Gulbe, I. Shestakova, G. Duburs and P. Arsenyan (2015). "1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC) increases Carmofur stability and in vitro antiproliferative effect." Toxicology Reports 2: 377-383.

218

Dong, Y. D., E. Tchung, C. Nowell, S. Kaga, N. Leong, D. Mehta, L. M. Kaminskas and B. J. El-Habashy, S. E., A. N. Allam and A. H. El-Kamel (2016). "Ethyl cellulose nanoparticles as a platform to decrease ulcerogenic potential of piroxicam: formulation and in vitro/in vivo evaluation." Int J Nanomedicine 11: 2369-2380. Freeman, S. (2011). Biological science—lipids, membranes, and the first cells. . The molecule of life. B. Cummings, Pearson Canada. Fresta, M., A. Villari, G. Puglisi and G. Cavallaro (1993). "5-Fluorouracil: various kinds of loaded liposomes: encapsulation efficiency, storage stability and fusogenic properties." Int J Pharm 99(2): 145-156. Gabizon, A., R. Catane, B. Uziely, B. Kaufman, T. Safra, R. Cohen, F. Martin, A. Huang and Y. Barenholz (1994). "Prolonged circulation time and enhanced accumulation in malignant exudates of doxorubicin encapsulated in polyethylene-glycol coated liposomes." Cancer Res 54(4): 987-992. Han, D. and A. J. Steckl (2013). "Triaxial Electrospun Nanofiber Membranes for Controlled Dual Release of Functional Molecules." ACS Appl Mater Interfaces 5(16): 8241-8245. Imran, M., M. R. Shah and Shafiullah (2018). Chapter 10 - Amphiphilic block copolymers–based micelles for drug delivery A2 - Grumezescu, Alexandru Mihai. Design and Development of New Nanocarriers, William Andrew Publishing: 365-400. Kabanov, A. V., E. V. Batrakova and V. Y. Alakhov (2002). "Pluronic block copolymers as novel polymer therapeutics for drug and gene delivery." J Control Release 82(2-3): 189-212. Kedar, U., P. Phutane, S. Shidhaye and V. Kadam (2010). "Advances in polymeric micelles for drug delivery and tumor targeting." Nanomedicine 6(6): 714-729. Khajavi, R. and M. Abbasipour (2012). "Electrospinning as a versatile method for fabricating coreshell, hollow and porous nanofibers." Scientia Iranica 19(6): 2029-2034. Khalf, A. and S. V. Madihally (2017). "Recent advances in multiaxial electrospinning for drug delivery." Eur J Pharm Biopharm 112: 1-17. Kim, D., M. H. Lee, M. A. Koo, B. J. Kwon, M. S. Kim, G. M. Seon, S. H. Hong and J. C. Park (2018). "Suppression of T24 human bladder cancer cells by ROS from locally delivered hematoporphyrin-containing polyurethane films." Photochem Photobiol Sci 17(6): 763-772. Kim, J., W. Lim, S. Kim, S. Jeon, Z. Hui, K. ni, C. Kim, Y. Im, H. Choi and O. Kim (2014). "Photodynamic therapy (PDT) resistance by PARP1 regulation on PDT-induced apoptosis with autophagy in head and neck cancer cells." J Oral Pathol Med 43(9): 675- 684. Kim, W. and S. S. Kim (2011). "Synthesis of biodegradable triple-layered capsules using a triaxial Kumbar, S., C. Laurencin and M. Deng, Eds. (2014). Natural and Synthetic Biomedical Polymers, Elsevier Science. Li, J., X. Wang, T. Zhang, C. Wang, Z. Huang, X. Luo and Y. Deng (2015). "A review on phospholipids and their main applications in drug delivery systems." Asian J. Pharm 10(2): 81-98. Liyan, Q., Z. Jianxiang, Y. Meiqiu, J. Yi and Z. Kangjie (2007). "Reverse self-assemblies based on amphiphilic polyphosphazenes for encapsulation of water-soluble molecules." Nanotechnology 18(47): 475602. 219

Malvern. (2011). "Dynamic Light Scattering." Retrieved 30/07/2018, from http://www.biophysics.bioc.cam.ac.uk/wp- content/uploads/2011/02/DLS_Terms_defined_Malvern.pdf. Manikandan, M., K. Kannan and R. Manavalan (2013). "Compatibility studies of camptothecin with various pharmaceutical excipients used in the development of nanoparticle formulation." Int J Pharm Pharm Sci 5(4): 315-321. Nii, T. and F. Ishii (2005). "Encapsulation efficiency of water-soluble and insoluble drugs in liposomes prepared by the microencapsulation vesicle method." Int J Pharm 298(1): 198-205. Noguchi, H. and M. Takasu (2001). "Self-assembly of amphiphiles into vesicles: a Brownian dynamics simulation." Physical Review E 64(4): 041913. Oerlemans, C., W. Bult, M. Bos, G. Storm, J. F. W. Nijsen and W. E. Hennink (2010). "Polymeric Micelles in Anticancer Therapy: Targeting, Imaging and Triggered Release." Pharm Res 27(12): 2569-2589. Papahadjopoulos, D., S. Nir and S. Ohki (1972). "Permeability properties of phospholipid membranes: Effect of cholesterol and temperature." Biochim. Biophys. Acta 266(3): 561- 583. Parisi, O. I., F. Puoci, D. Restuccia, G. Farina, F. Iemma and N. Picci (2014). Chapter 4 - Polyphenols and Their Formulations: Different Strategies to Overcome the Drawbacks Associated with Their Poor Stability and Bioavailability. Polyphenols in Human Health and Disease. San Diego, Academic Press: 29-45. Ren, J., W. Liu, J. Zhu and S. Gu (2008). "Preparation and characterization of electrospun, biodegradable membranes." J. Appl. Polym. Sci. 109(5): 3390-3397. Sardar, S., S. Sarkar, M. T. Z. Myint, S. Al-Harthi, J. Dutta and S. K. Pal (2013). "Role of central metal ions in hematoporphyrin-functionalized titania in solar energy conversion dynamics." Phys Chem Chem Phys 15(42): 18562-18570. Sessa, G. and G. Weissmann (1968). "Phospholipid spherules (liposomes) as a model for biological membranes." J Lipid Res. 9(3): 310-318. Trinh, H. M., M. Joseph, K. Cholkar, R. Mitra and A. K. Mitra (2017). Chapter 3 - Nanomicelles in Diagnosis and Drug Delivery∗. Emerging Nanotechnologies for Diagnostics, Drug Delivery and Medical Devices. Boston, Elsevier: 45-58. Vidal, N. L. G., S. G. Castro, S. F. S. Bruni, D. A. Allemandi and S. D. Palma (2014). "Albendazole solid dispersions: Influence of dissolution medium composition on in vitro drug release." Dissolution Technologies 21(2): 42-48. Weissig, V., T. K. Pettinger and N. Murdock (2014). "Nanopharmaceuticals (part 1): products on the market." Int J Nanomedicine 9: 4357-4373. Yang, Y.-T., C.-T. Chen and T. Tsai (2013). "Absorption and fluorescence spectral properties of hematoporphyrin in liposomes, micelles, and nanoparticles." Dyes Pigm 96(3): 763-769. Yang, Y.-T., C.-T. Chen, J.-C. Yang and T. Tsai (2010). "Spray-Dried Microparticles Containing Polymeric Micelles Encapsulating Hematoporphyrin." The AAPS Journal 12(2): 138-146. Yu, D.-G., C. Branford-White, G. R. Williams, S. W. A. Bligh, K. White, L.-M. Zhu and N. P. Chatterton (2011). "Self-assembled liposomes from amphiphilic electrospun nanofibers." Soft Matter 7(18): 8239-8247. Yu, D.-G., X.-Y. Li, G. R. Williams, J.-H. Yang and X. Wang (2018). Complex nanostructures from multi-fluid electrohydrodynamic atomization. 220

Yuan, S., J. Li, X. Xu, W. Chen, C. Chen, K. Kuang, F. Wang, K. Wang and F. Li (2018). "Underlying mechanism of the photodynamic activity of hematoporphyrininduced apoptosis in U87 glioma cells." Int J Mol Med. 41(4): 2288-2296. Zhang, P., Z. Zhang, H. Fan, S. Tian, Y. Chen and J. Yan (2016). "Waterborne polyurethane conjugated with novel diol chain-extender bearing cyclic phosphoramidate lateral group: synthesis, flammability and thermal degradation mechanism." RSC Advances 6(61): 56610-56622. Zhang, Y., H.-R. Liu and F.-W. Wang (2013). "Facile fabrication of snowman-like Janus particles with asymmetric fluorescent properties via seeded emulsion polymerization." Colloid and Polymer Science 291(12): 2993-3003. Zhenyu, L. and W. Ce (2013). One-Dimensional nanostructures. One-Dimensional nanostructures, Springer-Verlag Berlin Heidelberg: IX, 141.

221

6. Conclusions and future work

6.1 Conclusions

The aim of this thesis was to develop advanced polymer-based delivery systems for chemotherapeutic agents. This was achieved using electrohydrodynamic techniques, including electrospinning, electrospraying, and spray drying.

Chapter 2 described the fabrication of Janus particles and fibres using a side-by-side spinneret. These structures encapsulated carmofur, which acted as a chemotherapeutic agent, and rose bengal (RB) or haematoporphyrin (HP), which acted as photosensitisers.

The polymer used was polyvinylpyrrolidone (PVP). Firstly, the fabrication of Janus fibres encapsulating HP and RB with a side-by-side spinneret was explored, and once these were successfully achieved HP was substituted for carmofur to create a photochemotherapy formulation. Electrospraying was also used to fabricate Janus particles encapsulating carmofur and rose bengal.

The bicomparmental nature of the particles and fibres was confirmed by scanning electron microscopy (SEM); the Janus fibres had a diameter of 1.28 ± 0.48 µm and the particles of 0.51 ± 0.2 µm. They contained the active ingredients in the amorphous form, as was confirmed by differential scanning calorimetry (DSC) and X-ray diffraction (XRD) data. Both the particles and fibres gave rapid drug release over 2 h, due to the hydrophilicity of PVP. The cytotoxicity was tested in two cell lines: non-cancerous human dermofibroblasts (HDF) and lung cancer (A549) cells. The results showed that the Janus particles provided selective cytotoxicity towards lung cancer cells, and a synergistic

222

effect between the chemotherapeutic and the photodynamic active agents was observed.

After developing Janus fibres using PVP, which possesses a very rapid dissolution rate, a drug delivery system that achieved controlled drug release was sought. In this case,

SLNs were chosen as a DDS and were created based on sacrificial templates. Thus the aim of Chapter 3 was to explore the functional performance of solid lipid nanoparticles

(SLNs) self-assembled from electrospun nanofibres and electrosprayed particles. These were prepared using a benchtop electrospinning kit. Formulations were prepared based on PVP and tristearin, loaded with indomethacin and 5-fluorouracil (5-FU). Drug-loaded fibres and particles were successfully prepared, and the fibres obtained had an average diameter size of 150 nm - 3 µm, while the particles were around 163 - 135 nm in size.

XRD and DSC data confirmed that indomethacin and 5-FU were incorporated into the fibres and particles in an amorphous state while tristearin was present as a semi- crystalline component. SLNs of sizes ranging from 121-143 nm for IMC loaded SLN, 88-

143nm for 5-FU loaded SLN and 98-125 nm for unloaded SLN were generated upon addition of the precursor fibres or particles to water. Interestingly, the SLN which were assembled from the fibres displayed a slower drug release than those assembled from the particles. Both appeared to give sustained in vitro drug release and proved to be as cytotoxic as the raw drug. Electrospraying and electrospinning offer a facile alternative to prepare SLN which could be used for a wide application in the pharmaceutical field.

After the creation of SLNs was achieved using benchtop electrospinning/electrospraying,

Chapter 4 investigated scaling up the production of PVP and tristearin based formulations loaded with IMC or 5-FU. It explored the viability of producing sacrificial templates able to self-assemble SLNs using medium-speed and high-speed electrohydrodynamic techniques, as well as comparing them with spray drying. The benchtop electrospinning kit used in Chapter 3 employed a flow rate of 1 ml h-1, whereas

Chapter 4 used medium speed electrospinning apparatus at a flow rate of 10 ml h-1 and 223

a high speed electrospinning set-up with a flow rate of 1000 ml h-1. The spray drying technique was also used at a flow rate of 10 mL min−1. It was observed that during the scale up of the electrospinning and electrospraying processes, the fibres and particles created using higher flow rates showed a larger average diameter and more morphological deformities. The spray drying process created larger particles than those generated by electrospraying (5-6µm cf 135-163 nm). XRD and DSC confirmed that the drugs were encapsulated in an amorphous form in the PVP-GTS matrix, whereas GTS was semi-crystalline.

In all cases, the self-assembled SLNs proved to be suitable lipophilic drug carriers for indomethacin and 5-FU, showing the potential for the sustained release of these ingredients over a period greater than eight hours. During the permeation experiments, it was observed that the drug cargo of SLN accumulate inside cancer cells. Clear changes in the size of the particles in IMC-loaded SLN suspensions were observed after storage at room temperature for 6 months, while SLNs assembled from the fibres or particles proved to have similar sizes regardless of whether they were assembled from fresh formulations or systems which had been aged. The findings were not as stark with the 5-FU formulations, where a clear trend could not be observed.

The next step after achieving the creation of SLNs, which gave exended release when compared to the PVP materials fabricated in Chapter 2, was to explore triaxial electrospinning to create fibres that serve as a template to create self-assembled snowman Janus particles. Therefore, in Chapter 5, the creation of Janus particles encapsulating haematoporphyrin and carmofur, to create a photochemotherapy formulation, was undertaken. Firstly, poloxamer micelles entrapping the photosensitiser

(HP) were fabricated using the thin-film evaporation method, and after assessing the formation of the micelles by transmission electron microscopy (TEM) these were electrospun using PVP as a matrix. No differences in morphology after electrospinning were observed. An electrospun sacrificial template of 1,2-dioleoyl-sn-glycero-3- 224

phosphocholine (DOPC) -PVP and carmofur was also prepared, and upon hydration found to create self-assembled liposomes. The self-assembled liposomes were then examined by different techniques (e.g. DLS, TEM). Micelles had a size of approx. 90 nm, whereas liposomes were observed to have a bigger size of 113 nm.

Triaxial electrospinning was then undertaken, using the PVP-poloxomer-HP and PVP-

DOPC-carmofur solutions as the two cores and a blank PVP solution as the shell. This resulted in fibres with a diameter of 1.18 ± 0.27 µm, and two cores which were clearly visible in TEM.

The triaxial fibres were then added to water and the assembled structures were examined by TEM to observe their morphology. Snowman Janus nanoparticles were observed, where one compartment is expected to consist of a micelle entrapping HP and the other compartment a liposome entrapping carmofur. Drug release was explored using PBS, where the release of the drug from the micelles and liposomes, did not seem to have significant differences with the drug release from the Janus particles. Cytotoxicity studies proved that the fibres had the same cytotoxicity as pure carmofur, but were inconclusive when it came to HP. Further studies need to be done to confirm the phototoxicity of HP.

6.2 Future work

Chapter 2

The next step in taking this work forward consists of focusing on treating a particular type of cancer. Photodynamic Therapy (PDT) is FDA approved for just two types of cancer: esophageal and non-small lung cancer. Secondly, the size of some of the particles achieved by EHDA methods were of <1μm, which are probably suitable for delivery to the lungs. Formulations could thus be prepared containing lung cancer chemotherapeutics such as cis-platin, paclitaxel, docetaxel and gemicitabine. The

225

photosensitizer employed could also be changed to more effective compounds such as

Photofrin® or phthalocyanines.

The cytotoxicity assessment of Janus fibres encapsulating rose Bengal and carmofur

(PVP-RBC-F) should also be performed and compared to PVP-RBC-P to ascertain if there are any significant differences between particles and fibres. Different types of polymers should be considered to achieve a sustained drug release. The formulations might also be modified to target overexpressed receptors on cancerous cells, which will improve the selectivity of the materials. Another fruitful avenue to explore would be testing the Janus fibres and particles in more types of cells to explore if the high selectivity towards cancer cells can be extrapolated to all cancer cells or to only certain type of cells; for example, the cytotoxicity experiments in this work were completed using adherent cells, thus non-adherent cells should be studied. In vivo experiments would also be informative here.

Chapter 3

For future work, the inclusion of two drugs, such as 5-FU and rose bengal or hematoporphyrin, into the formulations could be probed to investigate the possibility of creating a photo-chemotherapy. Methods to increase the encapsulation efficiency of 5-

FU into the SLN can also be explored; using carmofur can have advantages when working with SLN as it is a hydrophobic drug, optimisation of the electrospinning and the electrospraying process should be attempted. Another approach when using hydrophobic drugs is the addition of cholesterol as it seems to increase the loading efficiency of lipid based carriers

Further, other methods to create SLN should be assessed to determine if there are any differences between the SLN assembled from electrospun fibres or electrosprayed particles and those generated using other, more traditional, methods. The use of

226

chloroform in this chapter could be considered a disadvantage, thus other solvents should be explored.

Chapter 4

The production of the 5-FU formulations with the high-speed electrospinning apparatus should be undertaken to allow a detailed comparison of all the scaled-up formulations.

Accelerated aging studies should be carried out on the formulations to assess storage stability, and the effects of lower temperature storage of the SLN suspension (e.g. at 0-

4 oC) determined. Permeation assays testing all the fabricated formulations should be performed, both before and after storage.

In vitro studies to elucidate factors such as uptake mechanisms, killing efficiency, cytotoxicity mechanisms and selectivity, could usefully be performed to provide a better understanding of the biological effects of the formulations.

Chapter 5

Future work for this chapter would first consist of whether if the micelles and liposomes composing the self-assembled snowman Janus particles, can be replaced with other structures such as SLN to create Janus particles. Systems with different drug loadings should also be produced and their in vitro cytotoxicity elucidated. In addition, a better understanding of the photocytotoxicity of the formulations could be investigated using a bleaching assay. This will give more information about the formulation’s efficacy, for instance enabling us to determine if the photosensitiser is aggregating in an aqueous medium. Finally, confocal microscopy can be used to observe the behaviour of the liposomes, SLN or nanoparticles with the cancer and non-cancer cells.

227

Appendix I: Additional experimental data for Chapter 2

a) Solid state properties of Janus particles with RB and HP

Figure A1-1. XRD patterns for the raw materials (hematoporphyrin & rose bengal) and electrospun fibres PVP+HP, PVP+RB and the Janus fibres.

PVP-RB-C-F

Exo Up Exo

Figure A1-1. DSC data for PVP fibres loaded with rose bengal and hematoporphyrin.

228

b) Physical form characterisation of Janus particles loaded with RB and carmofur

Figure A1-2. FTIR spectra of single fluid and Janus particles containing carmofur and rose bengal

Figure A1-3. XRD patterns of the electrosprayed particles PVP-RB-P, PVP-C-P, PVP- RBC-P and PVP-P.

229

Figure A1-4. DSC data for carmofur, rose bengal , PVP-RB-P, PVP-C-P, PVP-RBC-P and PVP-P.

230