MOLECULES USING ELECTROSPRAYEDeprints.qut.edu.au/74514/1/Nathalie_Bock_Thesis.pdfMicrospheres Loaded with Growth Factors (Poster) - vi - List of Postgraduate Conferences Australian

DELIVERY OF THERAPEUTIC

MOLECULES USING ELECTROSPRAYED

POLYMERIC PARTICLES FOR

APPLICATIONS IN TISSUE ENGINEERING

Nathalie Bock

M.Sc.

Submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

Institute of Health and Biomedical Innovation (IHBI)

Science and Engineering Faculty (SEF)

Queensland University of Technology (QUT)

July 2014

- i -

Abstract

The delivery of growth factors (GFs) from tissue engineered scaffolds is an emerging

strategy for guiding cells towards enhanced regeneration of tissues. Dosage,

however, is critical and GF delivery profiles that mimic natural release profiles are

the holy grail of delivery therapies. Despite this target, currently available products

deliver supraphysiological doses of GFs, generating health concerns associated with

possible tissue formation outside the targeted site and even potential tumour

development. Biodegradable polymeric carriers represent a suitable vehicle for GF

delivery upon matrix degradation, but whilst this is a promising concept in theory, in

practice, processing difficulties arise when encapsulating GFs due to harsh

processing conditions, which in turn may affect GF bioactivity. In this thesis,

electrospraying is hypothesised to be a superior technique to efficiently encapsulate

and reproducibly deliver active GFs from biodegradable polymeric microparticles.

The possibility of dry encapsulation of GFs allows preservation of GF bioactivity

and the identification of key processing parameters enables tailoring of particle size

and morphology, critical in dictating GF release patterns. Firstly, electrospraying was

used to develop polycaprolactone (PCL)- and poly(lactic-co-glycolic acid) 85:15

(PLGA)-based particle formulations containing a model protein, serum albumin (SA)

for optimisation. Following this, the encapsulation of vascular endothelial growth

factor (VEGF) and bone morphogenetic protein-7 (BMP-7), both GFs with proven

effects in bone tissue regeneration, were investigated. The use of poly(ethylene

glycol) (PEG) as an additive within electrosprayed particle formulations was proven

to be efficient in micronising proteins prior to dry encapsulation, and protecting the

bioactivity of GFs. The addition of PEG required optimisation of key variables,

which have been identified to be the electrospraying flow rate coupled with the

polymer solution properties, including concentration and molecular weight. Such

tailoring had a strong effect on particle size distributions, shown to be the most

determinant factor in controlling release profiles, in particular burst release.

Significant difficulties arose during in vitro characterisation due to GF interactions

with other GF molecules and the polymer matrix; this had a marked effect on GF

- ii -

quantification. The use of surfactants in solution was able to partially address this

issue, but may not be sufficient to enable full characterisation of electrosprayed

polymeric particles loaded with GFs. Cells assays were appropriate in assessing GF

bioactivity after various processing steps involved in electrospraying, and showed

that both VEGF and BMP-7 were highly bioactive, even after extended contact with

organic solvent (83% and 98% bioactivity, respectively). When electrosprayed

particles containing BMP-7 were placed in contact with pre-osteoblast cells in vitro,

significant osteogenic differentiation was observed up to three weeks. Finally, melt

electrospun meshes were used as a substrate onto which direct electrospraying of

loaded particles was undertaken, and they were shown to provide an ideal structure

with high porosity and pore size, enabling homogeneous coating throughout the

structure. Electrosprayed particles were shown to be non-toxic in contact with

fibroblast and osteoblast cells and the composite constructs (meshes plus particles)

elicited a positive effect in contact with osteoblast cells, essential pre-requisites for

tissue engineering (TE) applications. This PhD project has contributed new

knowledge for the fabrication and characterisation of electrosprayed particles loaded

with GFs and presents an innovative scaffold design for GF delivery. These findings

are important first steps in applying electrospraying technology in the field of TE,

and demonstrating much promise for the future of GF delivery strategies.

Keywords: albumin, bioactivity, bone morphogenetic protein, controlled release,

drug delivery, electrospraying, encapsulation, growth factor, in vitro characterisation,

melt electrospinning, microfibres, microparticles, microspheres, polycaprolactone,

poly(ethylene glycol), poly(lactic-co-glycolic acid), protein-polymer interactions,

scaffold, vascular endothelial growth factor, tissue engineering.

- iii -

List of Publications

Manuscripts Published

Chapter 2: Electrospraying of Polymers with Therapeutic Molecules: State of the Art

Bock N., Dargaville T. R., Woodruff M. A. (2012)

Progress in Polymer Science 37(11): 1510-1551

Chapter 3: Electrospraying, a Reproducible Method for Production of Polymeric

Microspheres for Biomedical Applications

Bock N., Woodruff M. A., Hutmacher D. W., Dargaville T. R. (2010)

Polymers 3(1): 131-149

Chapter 4: Controlling Microencapsulation and Release of Micronised Proteins

using Poly(Ethylene Glycol) and Electrospraying

Bock N., Dargaville T. R., Woodruff M. A. (2014)

European Journal of Pharmaceutics and Biopharmaceutics - DOI:

10.1016/j.ejpb.2014.03.008

Chapter 6: Composites for Delivery of Therapeutics: Combining Melt Electrospun

Scaffolds with Loaded Electrosprayed Microparticles

Bock N., Woodruff M. A., Steck R., Hutmacher D. W., Farrugia B. L.,

Dargaville T. R. (2014)

Macromolecular Bioscience 14(2): 202-214

- iv -

Scaffolds for Growth Factor Delivery as Applied to Bone Tissue Engineering

Blackwood, K. A., Bock N.*, Dargaville T. R., Woodruff M. A. (2012)

International Journal of Polymer Science - DOI: 17494210.1155/2012/1749

Manuscript Submitted

Chapter 5: Growth Factors Loaded into Electrosprayed Microparticles: Detection

and Bioactivity Discrepancies with In Vitro Assays.

Bock N., Dargaville T. R., Kirby G. T. S., Hutmacher D. W., Woodruff M. A. (2014)

* Co-first author

- v -

List of International Conferences

23rd

Annual Australian Society for Biomaterials and Tissue Engineering

(ASBTE) Conference

AUSTRALIA, Lorne. April 2014.

Bock N., Woodruff M. A., Hutmacher D. W., Dargaville T. R.

Delivery of Growth Factors using Electrosprayed Polymeric Microparticles for

Applications in Bone Tissue Engineering (Oral)

European Society for Biomaterials (ESB) Conference

SPAIN, Madrid. September 2013.

Bock N., Farrugia B. L., Hutmacher D. W., Dargaville T. R., Woodruff M.A.

Polymer Composite Constructs for Drug Delivery: Combining Melt Electrospun

Scaffolds with Electrosprayed Loaded Microparticles (Oral)

ESB Conference

IRELAND, Dublin. September 2011.

Bock N., Dargaville T. R., Hutmacher D. W., Woodruff M. A.

Controlled Release of Bioactive Molecules from PCL Microspheres Produced Using

Electrospraying Technologies (Poster and Short Oral)

ASBTE Conference

NEW ZEALAND, Queenstown. April 2011.


Electrospraying, a Reproducible Method for Production of Polymeric Microspheres

for Protein Delivery (Oral)

Tissue Engineering and Regenerative Medicine International Society – Asia

Pacific (TERMIS-AP) Conference

AUSTRALIA, Sydney. September 2010.


Electrospraying, a Reproducible and Non-Toxic Method for Production of

Microspheres Loaded with Growth Factors (Poster)

- vi -

List of Postgraduate Conferences

Australian Society for Medical Research (ASMR) Conference

AUSTRALIA, Brisbane. May 2014.


Growth-Factor Loaded Electrosprayed Microparticles for Targeted Bone Tissue

Regeneration (Oral)

Institute of Health and Biomedical Innovation (IHBI) Inspires Conference

AUSTRALIA, Brisbane. November 2013.


Delivery of Therapeutic Molecules using Electrosprayed Polymeric Particles for

Tissue Engineering (Oral)

Royal Australian Chemical Institute (RACI) Queensland, Polymer Group

Student Symposium



Delivery of Therapeutic Molecules using Electrosprayed Polymeric Particles for

Tissue Engineering (Oral)

IHBI Inspires Conference




for Protein Delivery (Poster)

RACI Queensland, Polymer Group Student Symposium

AUSTRALIA, Brisbane. August 2011.



for Protein Delivery (Short Oral)

- vii -

ASMR Conference

AUSTRALIA, Brisbane. May 2011.



for Protein Delivery (Poster)

IHBI Inspires Conference

AUSTRALIA, Gold Coast. November 2010.


Electrospraying, a Reproducible Method for Production of Microspheres (Poster)

- viii -

List of Scholarships and Awards

Scholarships

Supervisor Scholarship (2013) – $3,000 pa

Funded by A/Prof. Maria A. Woodruff and Dr. Tim R. Dargaville at Queensland

University of Technology (QUT)

Deputy Vice-Chancellor (DVC)'s Initiative Scholarship (2011-2013) – $6,000 pa

Funded by QUT at QUT

Australian Postgraduate Award (APA) Scholarship (2011-2013) – $22,860 pa

Funded by Dept of Education, Science and Training at QUT

Built Environment and Engineering (BEE) Faculty Living Allowance (2010) –

$22,500 pa

Funded by the Medical Devices Domain of the Institute of Health and Biomedical

Innovation (IHBI) and the Tissue Repair and Regeneration Program at QUT

Awards

ASMR Postgraduate Finalist ASMR Postgraduate

Conference

2014

Travel Award 23rd

ASBTE Conference 2014

Judge’s Prize for Best Oral Presentation,

Runner Up

IHBI Inspires Postgraduate

Conference

2013

Higher Degree Research Student of the

Month Award

Science and Engineering

Faculty (SEF), QUT

2013

PhD Career Start Award, Nominated Women in Technology 2013

Best Student Oral Presentation, 1st place 21

st ASBTE Conference 2011

Travel Award 21st ASBTE Conference 2011

Best Short Oral Presentation, Runner Up RACI Queensland,

Polymer Group Student

Symposium

2011

Outstanding Higher Degree Research

Student of the Month Award

Built Engineering

Environment (BEE)

Faculty, QUT

2010

- ix -

Table of Contents

Abstract ................................................................................................................................ i

List of Publications ............................................................................................................ iii

List of International Conferences ........................................................................................ v

List of Postgraduate Conferences ...................................................................................... vi

List of Scholarships and Awards ..................................................................................... viii

Table of Contents ............................................................................................................... ix

List of Abbreviations ....................................................................................................... xiii

Statement of Original Authorship .................................................................................... xvi

Acknowledgements ......................................................................................................... xvii

CHAPTER 1: GENERAL INTRODUCTION .................................................................... 1

1.1 Overview ...................................................................................................................... 1

1.2 Research Problem ........................................................................................................ 5

1.3 Aims and Outline of the Thesis .................................................................................... 5

1.4 Notes ............................................................................................................................ 6

CHAPTER 2: LITERATURE REVIEW: ELECTROSPRAYING OF POLYMERS

WITH THERAPEUTIC MOLECULES: STATE OF THE ART ...................................... 7

2.1 Abstract ........................................................................................................................ 9

2.2 Keywords ..................................................................................................................... 9

2.3 Introduction .................................................................................................................. 9

2.4 The Technique of Electrospraying ............................................................................. 12

2.4.1 Electrospraying Principles ...............................................................................12

2.4.2 Fabrication Techniques ....................................................................................13

2.5 Control of Particle Characteristics with Electrospraying Parameters ........................ 21

2.5.1 Importance of Electrospraying Parameters ......................................................21

2.5.2 Tailoring of Electrosprayed Particle Characteristics .......................................34

2.6 Electrospraying and Drug Release Characteristics .................................................... 43

2.6.1 Choice of Molecules ........................................................................................43

2.6.2 Loading and Encapsulation ..............................................................................47

2.6.3 Molecule Dispersion ........................................................................................51

2.6.4 Release Kinetics ...............................................................................................54

2.6.5 Denaturation ....................................................................................................64

2.6.6 Bioactivity ........................................................................................................66

2.6.7 In Vivo Performance ........................................................................................69

2.7 The Use of Electrosprayed Particles in Electrospun Scaffolds .................................. 72

2.7.1 Electrospun Nanofibres and Drug Delivery .....................................................72

2.7.2 Electrospun Nanofibres and Particles for Drug Delivery ................................74

2.8 Conclusions ................................................................................................................ 89

2.9 Acknowledgements .................................................................................................... 89

- x -

CHAPTER 3: ELECTROSPRAYING, A REPRODUCIBLE METHOD FOR

PRODUCTION OF POLYMERIC MICROSPHERES FOR BIOMEDICAL

APPLICATIONS .................................................................................................................. 91

3.1 Abstract ..................................................................................................................... 93

3.2 Keywords ................................................................................................................... 94

3.3 Introduction ............................................................................................................... 94

3.4 Experimental Section ................................................................................................. 98

3.4.1 Materials .......................................................................................................... 98

3.4.2 Microsphere Production .................................................................................. 98

3.4.3 Physical Characterisation .............................................................................. 100

3.4.4 Biological Effect of Microspheres ................................................................ 100

3.5 Results and Discussion ............................................................................................ 101


3.5.2 Biological Effect of Microspheres ................................................................ 109

3.6 Conclusions ............................................................................................................. 111

3.7 Acknowledgements ................................................................................................. 112

3.8 References and Notes .............................................................................................. 112

CHAPTER 4: CONTROLLING MICROENCAPSULATION AND RELEASE OF

MICRONISED PROTEINS USING POLY(ETHYLENE GLYCOL) AND

ELECTROSPRAYING...................................................................................................... 113

4.1 Abstract ................................................................................................................... 115

4.2 Keywords ................................................................................................................. 115

4.3 Introduction ............................................................................................................. 115

4.4 Experimental Section ............................................................................................... 118

4.4.1 Materials ........................................................................................................ 118

4.4.2 Particle Fabrication ....................................................................................... 118


4.4.4 In Vitro Characterisation ............................................................................... 121

4.5 Results and Discussion ............................................................................................ 122


4.5.2 In Vitro Characterisation ............................................................................... 136

4.6 Conclusions ............................................................................................................. 142


CHAPTER 5: GROWTH FACTORS LOADED INTO ELECTROSPRAYED

MICROPARTICLES: DETECTION AND BIOACTIVITY DISCREPANCIES WITH

IN VITRO ASSAYS.. ......................................................................................................... 145

5.1 Abstract ................................................................................................................... 147

5.2 Keywords ................................................................................................................. 147

5.3 Introduction ............................................................................................................. 147


5.4.1 Materials ........................................................................................................ 151

5.4.2 Particle Fabrication ....................................................................................... 151

5.4.3 Particle Characterisation ............................................................................... 152

- xi -

5.4.4 In Vitro Characterisation ................................................................................152

5.4.5 Statistical Analysis .........................................................................................156

5.5 Results ...................................................................................................................... 156

5.5.1 Particle Microstructure ..................................................................................156

5.5.2 GF Encapsulation Efficiency .........................................................................157

5.5.3 GF Recovery through In Vitro Processing .....................................................158

5.5.4 In Vitro GF Release .......................................................................................159

5.5.5 In Vitro GF Bioactivity ..................................................................................160

5.5.6 In Vitro Microparticle 2D Culture .................................................................164

5.6 Discussion ................................................................................................................ 167

5.6.1 GF Quantification with In Vitro Assays ........................................................167

5.6.2 The Use of Surfactants in In Vitro Assays .....................................................168

5.6.3 The Use of Stabilisers in Microparticle Formulations ...................................169

5.6.4 Bioactivity of GF through In Vitro Processing ..............................................171

5.6.5 GF Delivery in an In Vitro 2D Culture ..........................................................172

5.7 Conclusions .............................................................................................................. 173

5.8 Acknowledgements .................................................................................................. 174

5.9 Supporting Information ............................................................................................ 174

5.9.1 Culture Conditions for GF Bioactivity Assessment .......................................174

5.9.2 Particle Microstructure ..................................................................................177

5.9.3 In Vitro GF Release .......................................................................................178

5.9.4 The Effect of Freeze-Drying on BMP-7 ........................................................178

5.9.5 In Vitro Microparticle 2D Culture .................................................................178

CHAPTER 6: COMPOSITES FOR DELIVERY OF THERAPEUTICS:

COMBINING MELT ELECTROSPUN SCAFFOLDS WITH LOADED

ELECTROSPRAYED MICROPARTICLES .................................................................. 181

6.1 Abstract .................................................................................................................... 183

6.2 Keywords ................................................................................................................. 183

6.3 Introduction .............................................................................................................. 183


6.4.1 Scaffold Fabrication .......................................................................................185

6.4.2 Physical Characterisation ...............................................................................186

6.4.3 In Vitro Characterisation ................................................................................187

6.4.4 Biological Evaluation ....................................................................................189

6.5 Results and Discussion ............................................................................................. 190

6.5.1 Fabrication and Physical Characterisation .....................................................190

6.5.2 Protein Release and Polymer Degradation ....................................................195

6.5.3 Biological Effects ..........................................................................................202

6.6 Conclusions .............................................................................................................. 204

6.7 Supporting Information ............................................................................................ 205

6.7.1 Electrospraying Setup ....................................................................................205

6.7.2 Particle Size Distributions and Morphologies ...............................................205

6.7.3 Morphology of Composite Scaffolds .............................................................207

- xii -

6.7.4 Glass Transition ............................................................................................ 208

6.7.5 Molecular Weight, Polydispersity, Mass ...................................................... 208


CHAPTER 7: SUMMARY AND FUTURE DIRECTIONS .......................................... 211

BIBLIOGRAPHY .............................................................................................................. 217

- xiii -

List of Abbreviations

ALP alkaline phosphatase

ANOVA analysis of variance

AV applied voltage

BDNF brain-derived neurotrophic factor

BDP beclomethasone dipropionate

BMP bone morphogenetic protein

bSOD superoxide dismutase

C6 coumarin-6

CD circular dichroism

Chi chitosan

CLSM confocal laser scanning microscopy

CS chondroitin sulphate

Da Dalton unit (1 Da = 1 g/mol, 1 kDa = 1,000 g/mol )

DCM dichloromethane

DD direct dissolution

DDPS drug delivery particulate systems

DMEM Dulbecco's modified Eagle medium

DMF N,N-dimethylformamide

DMSO dimethyl sulfoxide

DOX doxorubicin

DSC differential scanning calorimetry

DTAB didodecyltrimethylammonium bromide

EBM-20 endothelial media

ECGS endothelial cell growth supplement

ECM extracellular matrix

EE encapsulation efficiency

EGF epidermal growth factor

ELISA enzyme-linked immunosorbent assay

ELP elastin-like polypeptides

EtOH ethanol

EX extraction

FCS foetal calf serum

FDA Food and Drug Administration

FDAC fluorescein diacetate

FITC fluorescein isothiocyanate

FGF fibroblast growth factor

FR flow rate

FTIR Fourier transform infrared

GC group contribution

GPC gel permeation chromatography

GF growth factor

GFR growth factor recovery

H&E hematoxylin and eosin

HA hydroxyapatite

HFIP hexafluoro-2-propanol

HMDS hexamethyldisilazane

HUVEC human umbilical vein endothelial cell

HRP horseradish perodixase

HSP Hansen solubility parameter

- xiv -

HyA hyaluronic acid

HV high voltage

ID internal diameter

IGF-1 insulin-like growth factor-1

IL-1α interleukin 1 alpha

KOW octanol/water partition coefficient

LC loading capacity

LF lung fibroblasts

L:G lactide:glycolide

MPHB methylparahydroxybenzoate

MSC mesenchymal stem cell

µBCA micro-bicinchoninic acid

µCT micro-computed tomography

MW molecular weight

MWD molecular weight distribution

NGF nerve growth factor

NHEF human epidermal fibroblasts

NHEK human epidermal keratinocytes

NIH National Institutes of Health

o/o/w oil-in-oil-in-water

o/w oil-in-water

PAA poly(amidoamines)

PBS phosphate buffer saline

PCL polycaprolactone

PDGF platelet-derived growth factor

PDI polydispersity index

PEG poly(ethylene glycol)

PEO poly(ethylene oxide)

PEUU poly(ester urethane) urea

PI propidium iodide

PLA polylactide

PLACL poly(L-lactic acid)-co-polycaprolactone

PLGA poly(lactic-co-glycolic acid)

PLL poly(ε-carbobenzoxy-L-lysine)

PLLA poly-L-lactide

PMMA poly(methyl methacrylate)

pNP p-nitrophenol

pNPP para-nitrophenyl phosphate

PPE-EA polyamino ethyl ethylene phosphate

P/S penicillin/streptomycin

PS20 polysorbate 20, Tween 20®

PSU polysulfone

PTMC poly(trimethylene carbonate)

PU polyurethane

PUU polyurethaneurea

PVA poly(vinyl alcohol)

PVC poly(vinyl chloride)

RED relative energy difference

Rg, radius of gyration

RHOB rhodamine B

RHOBOEP rhodamine B octadecyl ester perchlorate

RS release study

SA serum albumin

SD standard deviation

SDS sodium dodecyl sulphate

- xv -

SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis

SE standard error

SEM scanning electron microscope

SMC smooth muscle cell

s/o/o solid-in-oil-in-oil

s/o/w solid-in-oil-in-water

SS salbutamol-sulfate

TE tissue engineering

TEC tissue-engineered construct

TET tetracycline hydrochloride

TFE 2,2,2-trifluoroethanol

TGF-β transforming growth factor beta

TPP tripolyphosphate

TTC tip-to-collector

UV ultraviolet

VEGF vascular endothelial growth factor

vWF von Willebrand factor

w/o water-in-oil

w/o/w water-in-oil-in-water

XPS X-ray photoelectron spectroscopy

XRD X-ray diffractometry

% wt Weight per weight percentage

% wt/v Weight per volume percentage

- xvi -

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution. To the

best of my knowledge and belief, the thesis contains no material previously

published or written by another person except where due reference is made.

Signature:

Date:

QUT Verified Signature

- xvii -

Acknowledgements

I thank my supervisors, Dr. Tim Dargaville and A/Prof. Mia Woodruff, for their

support, encouragement, help and advice throughout my PhD. In particular, I am

grateful for the freedom and trust they granted me since the start, letting me bring the

electrospraying technology in the picture and supporting all of my ideas, both

intellectually and financially. This attitude was very valuable, helping me grow as an

autonomous and confident researcher. Many thanks also go to Prof. Dietmar

Hutmacher, as an inspirational supervisor associated to this project, helping me to

think ‘outside the box’ while envisioning the big picture.

The execution of this PhD would not have been possible without the great facilities

available at IHBI and QUT and their administrative and technical facilitators; they

hold a great part in making the work pleasant and fruitful. Numerous sources of

funding which are mentioned hereafter are also part of this PhD success, but in

particular I would like to thank the Australasian Society of Biomaterials and Tissue

Engineering for funding my travel to their annual conference in New Zealand in

2011, which was a highlight of this PhD with excellent scientific brainstorming and

remarkable good fun with fellow scientists.

I also thank the members of the Regenerative Medicine group, Biomaterials and

Tissue Morphology group, Tissue Regeneration and Repair group and IHBI

community for advice, help, inspiration and companionship. In particular, thanks to

Brooke Farrugia, for being an amazing friend, attentive listener, and perspicacious

advisor in both my personal life and laboratory undertakings, always there for me

when things hit a rough patch.

Coming from Europe and doing a PhD in Australia was not the easiest decision to

make, but I thank my family for understanding and supporting this choice. Here is

the place and time to acknowledge the education I received from them all, but in

particular from my parents, Pascale and Patrick, and grand-mother Simone (dec.),

which brought me here today. I thank them for trusting and supporting my choices,

encouraging me achieve challenges that sounded impossible for women in their

- xviii -

times. Also thanks to my sister and goddaughter, Mélanie and Espérance, for helping

in keeping this whole doctorate thing not too serious.

Finally, I want to express my most profound gratitude to my partner Luigi, for his

unconditional love, patience, support, encouragement and scientific advice over the

last ten years. Going through a PhD together and holding a common passion for

materials (thanks to EEIGM) made the experience incredibly rich and gave me a lot

of strength. I thank him for always believing in me in any situation and making me

believe that I could achieve anything. Having a child together during this PhD was

something I could not have done without Luigi’s precious qualities and he made the

journey easy, natural and marvellous.

Last, but not least, I thank my son Owen for being the biggest joy in my life. His

good nature and all the wonderful moments we spent together made this time in

Australia a priceless treasure that I will cherish forever, thank you!

- 1 -

Chapter 1: General Introduction

1.1 OVERVIEW

Towards the end of the past millenary, the most significant science discoveries have

paved the way to a progressive but tremendous improvement in the quality of life of

the human species. From the discovery of the cell in the 17th

century by Hooke,

followed by the principle of mass conservation by Lavoisier in the 18th

century, the

19th

century was then the cradle of electromagnetism and thermodynamics laws,

opening the horizons of physics, chemistry and materials sciences, which were

completely rethought throughout the 20th

century. Biological and medical

breakthroughs overflew the last half of the 20th

century, with the discovery of the

structure of DNA, various vaccines and other key discoveries in molecular biology.

At the dawn of the 21st century, the map of genetic information for humans has been

completed, providing insights into one of the major challenges of this new millenary;

the diagnosis and treatment of diseases.

Health concerns are an increasing burden in today’s world, where an ageing

population with unhealthier lifestyle is most likely to face injury, trauma,

degenerative diseases, or tumours in the course of their lifetime. Thanks to the

progresses of science, it is now possible with the medical technologies available

today, to evaluate, treat, augment or replace faulty tissues, organs or functions in the

body by the implantation of materials intended to interface with biological systems

[1]. These materials are termed ‘biomaterials’ and imply a non-toxic response from

the human body. However, while the quality of life of many people has been

improved by biomaterials, most materials are still inert and non-resorbable, made of

metallic, polymeric, or ceramic materials and used as temporary or permanent

prosthesis, which will never perform as well as the natural, original tissue or organ.

Transplantation has shown to be another alternative to tissue repair but is associated

with limited availability and immunological rejection, when coming from another

donor. This increasing need for better therapies has driven the commencement of

tissue engineering therapies in the 1980s [2].

Chapter 1 General Introduction

- 2 -

The concept of tissue engineering (TE) was first introduced by Langer and

Vacanti and refers to a relatively new and interdisciplinary field that applies the

principles of engineering sciences and life sciences [3]. Tissue engineering aims to

develop constructs that promote the repair, restoration or regeneration of cells or

tissues by combining a matrix, often bio-resorbable, with living cells or therapeutic

molecules [4]. The strategies for TE include the growth of functional tissues in vitro

or the regeneration of tissues in vivo and they are aimed at treating skeletal tissues

like bone and cartilage, neural tissue, muscle tissue in the heart, smooth and skeletal

areas, liver tissue and skin [5].

Growth factors (GFs) are central molecules during natural tissue formation and

repair. Cells secrete these polypeptidic molecules in situ to modulate cellular

activities, and a complex and orchestrated delivery of numerous GFs have been

shown to direct tissue growth [6]. GFs are either released on external stimuli for

immediate signalling or are embedded in the extracellular matrix (ECM) and further

released as the ECM degrades [7]. The GFs initiate their action by binding to specific

receptors on the surface of target cells [6] and this action can further activate various

responses, such as directing the phenotype of certain cells, triggering their

proliferation, activating macrophage action and inducing morphogenesis [5, 7]. For

example, while vascular endothelial GF (VEGF) and platelet-derived GF (PDGF)

induce angiogenesis (the formation and maturation of blood vessels) fibroblast GF

(FGF), keratinocyte GF (KGF), interleukin 1 alpha (IL-1α) are involved in skin

regeneration, and bone morphogenetic proteins (BMPs) such as BMP-2 and BMP-7

and transforming GF beta (TGF-β) are key molecules in bone regeneration. The

presence of GFs can be traced during the formation or repair of every tissue within

the body [5] and act in a concentration- and time-dependent manner, often requiring

minute quantities to elicit biological activity [7]. In the case of tissue repair, it is the

repair progress that triggers and controls the timing and location of GF release [5].

Tissue engineers are thus faced with the significant challenge of providing systems

that mimic the natural GF production in a spatial and temporal manner [5].

Current technologies enable GFs to be recombinantly constructed by host

organisms in vitro, at very high cost, but in clinical practice, only a few GFs have

been authorised by regulatory federal agencies, such as the American Food and Drug

Administration (FDA). Approved products include PDGF for the treatments of

diabetic foot ulcers [8] and BMP-2 and BMP-7 for oral maxillofacial applications

Section 1.1 Overview

- 3 -

and spinal fusion [9]. Those products use supra-physiological amounts delivered

from a biodegradable carrier, to obtain a substantial healing response, since GFs have

a short half-life in solution. For example, two products from the market leaders in

orthopaedic GF delivery, Medtronic and Olympus, provide several milligrams of

recombinant BMP-2 (INFUSE®) and BMP-7 (OP-1®), respectively, reconstituted

and added to a collagen sponge immediately prior to implantation into a bone defect

site, in the clinic. Only nanogram levels are actually required to stimulate cells,

hence the massive dose of BMPs present in the sponges, that diffuse away from the

site within minutes of implantation, create increased cost and may lead to

complications [7-11]. BMPs are a potent stimulator of bone formation [12] and

because every cell in the body presents a BMP receptor, there is a serious concern

that BMPs could cause bone to form outside the fusion area (ectopic bone

formation), in places where it may be harmful. These safety concerns were feverishly

brought to light in 2011 with an entire issue of the Spine journal dedicated to the

BMP debate, where it is generally acknowledged that these dangers are associated

with high doses and off label use [13-15].

Several strategies have been implemented to deliver physiologically relevant

quantities of GFs by using polymeric systems intended to encapsulate and deliver

GFs in a sustained and controlled manner, while at the same time protecting the GFs

from their environment [6, 7, 16, 17]. These systems include biodegradable

microspheres, hydrogels, porous soft scaffolds and three dimensional (3D) hard

scaffolds, and can be made from natural or synthetic polymers [17, 18]. The choice

of matrix is important to the success of the tissue-engineered construct and often

fibre-based scaffolds have been used; nanofibres in particular can mimic the

architecture of natural tissue constituents like collagen [19]. Fibre scaffolds, in

general, provide favourable chemical and topographical structures for cells to attach

and proliferate but also favourable physical properties, such as porosity and

interconnectivity for cell infiltration and maturation into a specific tissue [20, 21].

Their lack of biological stimulation has been addressed by tissue-inductive coatings

or incorporation of GFs directly into the fibres, available for cellular uptake upon

matrix degradation [22-24]. However, GFs should ideally be sustainably released

over many weeks, while the fibre construct can still maintain its structural function

[25]. The addition of GFs directly into fibres also leads to deteriorated mesh

properties which, taken collectively, may lead to non-optimum constructs. To


- 4 -

overcome these limitations, the addition of a separate release system to a fibre

scaffold represents a more suitable solution, allowing the tailoring of the release

system separately from the scaffold [24].

The use of biodegradable microparticles, in particular, has shown a lot of potential

for GF delivery applications [17, 18]; they somewhat mimic cells by releasing

smaller but more effective amounts of GFs, closer to physiological doses [18]. Many

techniques exist for producing GF-loaded particles with emulsion/evaporation-based

methods being the most extensively used [26]. However, to date, very few of these

techniques have been effectively translated to the clinic. This lack of translational

research is mainly attributed to shortfalls associated with conventional production

methods, including GF degradation during processing, mostly due to the use of

organic solvents and emulsions [11]. Electrospraying is rapidly emerging as a

potentially superior technology for the production of polymeric particles containing

therapeutic molecules, able to reduce denaturation of proteins and drugs [27].

Electrospraying also affords enhanced regulation over particle size/morphology,

which is essential in controlling release profiles. While the concept of

electrospraying is relatively simple, involving electrohydrodynamic atomization of a

polymer solution [28], understanding and optimising the technique is still in its early

years in respect to biological loading, with less than 100 reports in the last 20 years.

Most studies focus on small molecule drugs for antibiotic, anti-cancer, anti-

inflammatory and asthma treatments and few reports are on proteins, and even less

on growth factors (less than 10 [29-33]) mostly for angiogenic and chrondrogenic

applications. Proteins and growth factors are, however, fundamentally different

structures from small molecule drugs, with an increased degree of complexity, due to

their polypeptidic structure, and may trigger different behaviours when loaded and

delivered from electrosprayed particles. Importantly with electrospraying, the

technique allows for dry encapsulation of GFs, which is now recognised to better

maintain GF activity during processing and delivery [34, 35].

The development of electrosprayed particles reproducibly loaded with GFs

relevant to bone tissue may be an appropriate alternative to the current products in

the market, by delivering physiological doses of active BMPs over time, in a

reproducible way. The versatility of electrospraying may enable tailoring of specific

release profiles for different GFs and applications, while delivering fully active GFs

to the injury site, in turn lowering the initial dose needed. This should result in a

Section 1.2 Research Problem

- 5 -

decrease of the associated complications arising when patients receive a ‘critically

too high’ dose of GFs, enhancing the treatments of musculoskeletal disorders. The

use of a fibre scaffold as a carrier for this release system may be particularly

beneficial for this application, where a soft mesh containing GF-loaded particles may

drape the surface of a defected bone for disease treatment or may be integrated in a

bone substitute for bone regeneration, following injury or trauma.

1.2 RESEARCH PROBLEM

In this thesis, it is hypothesised that the electrospraying technology may be used to

produce biodegradable microparticles encapsulating and delivering growth factors

relevant in bone tissue engineering.

1.3 AIMS AND OUTLINE OF THE THESIS

The research problem of this thesis will be addressed by the following aims:

understanding and tailoring the processing parameters involved in

electrosprayed particle formation,

developing electrosprayed particle formulations for reproducible and efficient

GF encapsulation,

characterising GF-loaded formulations for in vitro release and bioactivity,

investigating the potential of loaded electrosprayed microparticles used in

association with a porous fibre scaffold in vitro, as a suitable construct for

tissue engineering.

In this thesis, Chapter 1 introduces the background, rationale and aims of delivering

growth factors using electrosprayed polymeric particles for applications in tissue

engineering. Chapter 2 reviews the literature on this topic, reporting the principles

of the electrospraying technique and presenting the characteristics of particles loaded

with therapeutic molecules, from a physical, pharmaceutical and biological

perspective. In addition, the use of electrosprayed particles in fibre scaffolds for

tissue engineering is reported. Chapters 3 to 6 consist of the experimental chapters

that cover different aspects of the thesis topic. The first objective is the

understanding of reproducible microparticle production via electrospraying, hence

Chapter 3 details the production and optimisation of GF-free electrosprayed


- 6 -

polycaprolactone (PCL) microparticles. Key parameters are identified with regards to

reproducibility and control of particle size and morphology. Furthermore,

cytocompatibility of optimised microparticles is assessed in vitro. The next step

involves encapsulation of proteins, as model GFs, into electrosprayed microparticles,

which is the focus of Chapter 4. This chapter investigates the use of PCL and

poly(lactic-co-glycolic acid) (PLGA) in association with an additive; poly(ethylene

glycol) (PEG), for the encapsulation of a model protein, serum albumin (SA). The

influence of PEG on the physical and in vitro characteristics of loaded microparticles

is discussed. In Chapter 5, the optimised encapsulation technique with PEG is

extended to encapsulating GFs relevant in bone tissue engineering; namely VEGF

and BMP-7. The release, detection and bioactivity of GFs are assessed in vitro with

an emphasis on GF interactions with the environment. Finally, the application of

loaded electrosprayed particles in tissue engineering is addressed in Chapter 6. This

final experimental chapter describes the fabrication of a new type of composite

scaffold comprising PCL microfibres produced by melt electrospinning and

electrosprayed PLGA microparticles loaded with SA. The composites are

characterised by their physical and degradation properties, and for their ability to

support cell growth in vitro. To conclude, Chapter 7 summaries the findings of the

thesis and provides an outlook on the research topic, looking to the future.

1.4 NOTES

This thesis is presented by publications, where chapters 2, 3, 4 and 6 have already

been published and chapter 5 is under review. Due to the layout required for thesis

by published papers, a few alterations from the published versions have been made

for ease of the reader. These changes include:

renumbering of figures and tables,

change of American spelling to British spelling,

compilation of all references into one combined list for the entire thesis,

homogenisation of abbreviations and units for consistency,

minor additions to emphasise specific points relevant to the thesis topic.

- 7 -

Chapter 2: Literature Review:

Electrospraying of Polymers with

Therapeutic Molecules: State of the Art

Nathalie Bock1,2,3

, Tim R. Dargaville1, Maria A. Woodruff

2

Published in Progress in Polymer Science, Volume 37, Issue 11, 2012, Pages 1510-

1551.

© 2012 Elsevier Ltd. All rights reserved.

Statement of contribution of co-authors for thesis by published papers

Contributors Statement of contribution

Nathalie Bock Searched and read the literature

Designed the review outline

Wrote the manuscript

Tim R. Dargaville* Involved in the conception of the project

Provided feedback on manuscript

Maria A. Woodruff* Involved in the conception of the project


1 Tissue Repair and Regeneration Group

2 Biomaterials and Tissue Morphology Group

3 Regenerative Medicine Group

Institute of Health and Biomedical Innovation, Queensland University of Technology,

60 Musk Avenue, Kelvin Grove, QLD 4059, Australia

Chapter 2 A Literature Review on Electrospraying Applied to Tissue Engineering

- 8 -

The authors listed above have certified* that:

1. they meet the criteria for authorship in that they have participated in the

conception, execution, or interpretation, of at least that part of the publication in

their field of expertise;

2. they take public responsibility for their part of the publication, except for the

responsible author who accepts overall responsibility for the publication;

3. there are no other authors of the publication according to these criteria;

4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the

editor or publisher of journals or other publications, and (c) the head of the

responsible academic unit, and

5. they agree to the use of the publication in the student’s thesis and its publication

on the QUT ePrints database consistent with any limitations set by publisher

requirements.

Principal Supervisor Confirmation

I have sighted email or other correspondence from all Co-authors confirming their

certifying authorship.

Section 2.1 Abstract

- 9 -

2.1 ABSTRACT

The encapsulation and release of bioactive molecules from polymeric vehicles

represents the holy grail of drug and growth factor delivery therapies, whereby

sustained and controlled release is crucial in eliciting a positive therapeutic effect. To

this end, electrospraying is rapidly emerging as a popular technology for the

production of polymeric particles containing bioactive molecules. Compared with

traditional emulsion fabrication techniques, electrospraying has the potential to

reduce denaturation of protein drugs and affords tighter regulation over particle size

distribution and morphology. In this article, we review the importance of

the electrospraying parameters that enable reproducible tailoring of the particles’

physical and in vitro drug release characteristics, along with discussion of existing in

vivo data. Controlled morphology and monodispersity of particles can be achieved

with electrospraying, with high encapsulation efficiencies and without

unfavourable denaturation of bioactive molecules throughout the process. Finally, the

combination of electrospraying with electrospun scaffolds, with an emphasis on

tissue regeneration is reviewed, depicting a technique in its relative infancy but

holding great promise for the future of regenerative medicine.

2.2 KEYWORDS

Electrospraying, microparticles, encapsulation, drug delivery, controlled release,

tissue engineering, electrospinning.

2.3 INTRODUCTION

The need for controlled delivery of therapeutic molecules has prompted the

investigation of polymeric particles as biodegradable reservoirs which are designed

to degrade at a determined rate, thereby releasing their encapsulated molecules for

sustained and site-specific delivery [17, 18]. This approach could potentially

overcome the limitations of bolus delivery and has drawn much research attention in

the last decades, particularly in the fields of cancer therapies, hormonal treatments,

asthma delivery, and tissue engineering, for which tailored and multiple-molecule

delivery is necessary for therapeutic effect [36]. Many techniques exist for producing

these drug delivery particulate systems (DDPS) with emulsion/evaporation-based


- 10 -

methods being the most extensively used [26]. In this context, the term ‘drug’ refers

to any type of molecule that has a therapeutic effect. Coacervation, spray-drying,

nanoprecipitation and microfluidics are additional techniques each presenting their

own specific advantages and they are broadly described in the literature [37-40].

However, to date, very few of the DDPS generated using these techniques have been

effectively translated to the clinic, with few devices being commercialised each year.

This lack of translational research is mainly attributed to several shortfalls associated

with these production methods [11]. For instance, there are issues surrounding

molecule degradation (such as denaturation of proteins) and instability during the

processes. In emulsion techniques, the aqueous/organic interface and shear stresses

are the first source of limitation [35]. Moreover, entrapped molecules differ in terms

of therapeutic function and physicochemical properties, demonstrating a different

degree of stability and sensitivity to stress. In other techniques, prolonged exposure

to organic solvents and residual traces of solvents or other processing agents in the

final DDPS are of concern. Such factors can affect the nature and stability of the

encapsulated therapeutic molecules, limiting their performance both in vitro and in

vivo, and thus limiting their clinical use. Furthermore, different applications require

different therapeutic molecule release profiles matching the need of a specific treated

tissue, and ideally mimicking the in vivo release profiles generated by the cells from

such tissues. For this to happen, it is critical to have a thorough grasp of the complex

interplay of fabrication parameters which govern the resultant particle characteristics.

Particle size and morphology for example ultimately dictate the degradation, and

hence release profiles from DDPS, although it should be noted that tight control is

currently limited in the traditional fabrication techniques.

One approach to overcome these drawbacks is the technique of electrospraying.

Although electrospraying is a well-established technique in the field of mass

spectrometry and ink-jet printing, it has only been applied to the loading of

therapeutic molecules in the last 20 years and its understanding and optimisation are

still in their relative infancy with respect to biological loading [27, 28, 41]. Briefly,

in electrospraying, a high voltage is applied to a liquid infused through a capillary

nozzle. The electric charge generated on the droplet competes with the surface

tension of the droplet, causing the droplet to break up in nano- to micro-droplets,

which undergo solvent evaporation. The resulting dried particles can then be

collected [42]. Therapeutic molecules can be incorporated into the polymer solution

Section 2.3 Introduction

- 11 -

prior to electrospraying resulting in loaded particles. There are numerous advantages

to electrospraying including the following: the use of an emulsion is optional but not

required; there is no use of high temperature such as in spray-drying; there is no

further drying step required since particles are instantaneously dried during the

process; and there is an enhanced control over the size distribution of particles with

the possibility of producing quasi-monodisperse particles [43]. The latter is

particularly desirable in drug delivery since monodispersity provides more

controlled, and hence reproducible release profiles, which may in turn be more easily

customised for a desired application [44]. Furthermore, in the specific case of

nanoparticles, size affects cellular uptake and thus uncontrolled size distribution may

lead to different biological responses [45, 46]. Control of size is thus of paramount

importance when producing loaded polymeric particles and electrospraying is a

technique which can provide such control over and above that achieved with

traditional techniques, when appropriate parameters are used [47].

Electrospraying also holds potential to reduce denaturation by limiting exposure

to organic solvents and is highly versatile in terms of the choice of polymers,

apparatus, and therapeutic molecules. For instance, if the therapeutic molecule is

highly sensitive to solvents, such as enzymes and DNA molecules, coaxial

electrospraying may be employed. In this way, core-shell capsules are formed and

the protein resides in the core of the capsule in an aqueous solution while the

polymer matrix composes the shell of the capsule [48]. Finally, although

electrospraying through one nozzle has a low throughput, the flexibility of the

technique would enable the use of several nozzles in parallel for a multiplexed

system, ideal for scale-up [49, 50]. To date, therapeutic molecules such as antibiotics

(ampicillin [51], rifampicin [52]), anti-cancer agents (paclitaxel [53-60], doxorubicin

[61], suramin [58], cisplatin [62]), inhalation drugs (beclomethasone dipropionate

(BDP) [63], salbutamol-sulfate (SS) [64]), anti-inflammatory drugs (celecoxib [65],

budesonide [66], naproxen [67]), drugs for hormonal treatments (β-oestradiol [68],

tamoxifen [69, 70]), model proteins (serum albumin (SA) [71-74]) and growth

factors (GFs) (insulin-like GF-1 (IGF-1) [29], vascular endothelial growth factor

(VEGF) and platelet-derived GF (PDGF) [30]) have been loaded in electrosprayed

particles and these studies will be discussed hereafter.

Here we present a comprehensive review of the current state of the art in

electrospraying technology for the controlled release of therapeutic molecules from


- 12 -

polymeric particles. We review the methods used for producing electrosprayed

particles and encapsulating therapeutic molecules, including important

considerations to enable both the physical properties and in vitro drug release

profiles of the particles to be tailored and optimised. The focus of the review is on in

vitro data since very little in vivo data is available yet in the literature, although

discussion of existing in vivo data is also provided. The various applications of

electrospraying with electrospinning technologies, with an emphasis on tissue

engineering, are also reviewed, for a portrayal of the latest techniques used to

produce scaffolds in the diverse and fascinating field of regenerative medicine.

2.4 THE TECHNIQUE OF ELECTROSPRAYING

2.4.1 Electrospraying Principles

Electrospraying is a method of liquid atomization, also known as

electrohydrodynamic atomization. The principle of electrospraying is based on the

theory of charged droplets; stating that an electric field applied to a liquid droplet

exiting a capillary is able to deform the interface of the droplet [28]. The electric

charge generates an electrostatic force inside the droplet which competes with the

surface tension of the droplet, forming the Taylor cone, characteristic of a charged

droplet. Eventually, the electrostatic force, generated by the use of high voltage on

the capillary, is able to overcome the surface tension of the droplet. The excess

charge then needs to be dissipated and smaller charged droplets on the micro to

nano-scale are ejected from the primary droplet, thus reducing its charge without

significantly reducing its mass. Due to Coulomb repulsion of the charges, the

droplets disperse well and do not coalesce during their flight towards the collector

[43]. Several spraying modes can occur during electrospraying; the most desired

being the single cone-jet mode, due to its stability and reproducibility [42].

The various theories of electrospraying physics have been summarised elsewhere

with reviews on the recent advances and applications of the technology [27, 28]

however limited literature exists pertaining to theoretical and practical inclusion of

bioactive molecules in this process. Briefly, the two major parameters that

characterise the electrosprayed aerosol are the size of droplets and electric charge.

The latter is difficult to determine, due to parasitic electrical discharge, although the

Section 2.4 The Technique of Electrospraying

- 13 -

maximum surface charge of a droplet, q, has been identified as a function of the

surface tension, γ, and radius of droplet, R, expressed in Equation 2.1 [75]:

√ (2.1)

From the surface charge, the Rayleigh limit, LR, can be identified, which

determines the charge leading to droplet break-up (Coulomb fission) and is

expressed in Equation 2.2, where ε is the permittivity of the surrounding medium:

( ) (2.2)

Coulomb fission is an unwanted phenomenon by which the charged

electrosprayed droplet emits a cloud of small highly charged droplets, called

‘offsprings’. This occurs if the droplet holds more charge than the Rayleigh limit, as

determined by electrical stresses and surface tension. Droplets produced by

electrospraying are highly charged, usually close to half of the Rayleigh limit [28].

Similarly, the jet break-up mechanism is shown to be dependent on the ratio of

the electrical normal stress over the surface tension stress. It is dependent on the

viscosity and surface charge as in the Rayleigh limit (Equation 2.2), but also on the

acceleration of the jet. With increasing flow rate, the current increases and the stress

ratio of the jet increases, above a threshold value whereby the jet starts to whip,

leading to the production of heterogeneously sized droplets. Ideally, a sufficient

stress ratio value must be employed to allow for jet break-up, but still a minimal

value must be obtained for limiting droplet break-up for production of monodisperse

and homogeneous particles [42].

2.4.2 Fabrication Techniques

The electrospraying setup can be simple and inexpensive; a polymer solution is

loaded into a syringe fitted with a conductive nozzle, and infused at a desired rate

generally implemented by a syringe pump. The nozzle is subjected to high voltage

(in the order of kilovolts and mostly positive) and various types of collectors, often

grounded or more rarely negatively charged, are placed at a distance ranging from a

few centimetres to several tens of centimetres from the nozzle. Once the droplets are

ejected from the Taylor cone according to the theory of charged droplets, solvent

evaporation leads to the progressive contraction and solidification of droplets

resulting in solid polymeric particles impacting onto the collector. While particles are


- 14 -

generally assumed to be dry or proven to contain residual solvent falling within the

limit of safety standards [76], many studies also use subsequent vacuum treatment to

ensure all residual solvent is removed. In the context of loading, the biologically

active molecule (biomolecule) is generally mixed into the polymer solution before

electrospraying; this approach is covered in section 2.4.2.2.

2.4.2.1 Electrospraying Apparatus

Electrospraying and drug loading characteristics can be tailored by changes in the

choice and configuration of the equipment. One type of apparatus involves the use of

nozzle-ring devices (Figure 2.1A) which are placed inside glass chambers and

subjected to a stream of air/nitrogen (Figure 2.1B). This setup is sometimes referred

as the ‘Delft type’ (from the Technical University of Delft, The Netherlands) [60]. A

potential difference is generated between the nozzle and a ring placed around the

nozzle [55, 59, 60, 71]. Usually the high voltage is applied on the nozzle and the

lower voltage on the ring, respectively. The use of a ring stabilises the

electrospraying process [60], enabling better control over the desired spraying pattern

[59]. For instance, in the single cone-jet mode, more uniform particles are produced

[56]. The use of a ring is recommended when using water as the solvent since a

stable cone-jet mode is harder to achieve with water [61]. A corona discharge is

generated by a grounded needle placed opposite the charged nozzle in order to

discharge the highly charged droplets. Particles can be collected through filters,

transported by an air/nitrogen flow applied in the chamber [59], or collected around

the grounded needle in a Petri dish [71]. The use of a chamber reduces solvent

evaporation rate and smaller particles may be produced [56], however, yield is

lowered in this configuration due to deposition of particles in the glass wells of the

chamber (where up to 30% can be deposited) before collection in the filter [59, 73].

Consequently, this setup is not recommended for loading of molecules where losses

cannot be afforded. However it can be optimised by improving vacuum aspiration

and efficient discharging of particles [59] to reach up to 80% yield. Furthermore, the

reduction of solvent evaporation rate generated by using an enclosed chamber can

lead to smoother microparticle morphologies due to enhanced polymer relaxation

and thus better organisation of polymer chains within the evaporating droplet [77],

which, in turn, allow more homogenous particle degradation and release.


- 15 -

An alternative method for collection involves electrospraying loaded droplets into

a liquid, within a beaker containing an immersed grounded collector [72, 73, 78] or a

wire wrapped around the beaker [64] (Figure 2.1C-D). Collection media include

distilled water [72, 73], ice-water/methanol [79], anhydrous ethanol [58], or 70%

ethanol supplemented with surfactants (such as 0.01% to 0.1% (v/v) Polysorbate 80

(Tween 80®) [64]), to lower the surface tension of the solution and prevent the

aggregation or coalescence of particles [39]. However, it should be noted that high

surfactant concentrations (such as > 0.1% of Tween 80®) have been shown to

broaden the size distribution of particles which reduces consistency between batches

[64]. Stronger solvents such as acetone may also be used, in order to neutralise

residual solvent from the spraying solution [78]. After collection in the liquid,

particles can be further filtered and dried. The major disadvantage of this collection

technique is the loss of surface-adsorbed drugs which may be desorbed into the

media. There is, therefore, no burst release of biomolecules (from the surface of

electrosprayed particles) seen with these systems, and a proportional amount of

molecules is lost, which again is a concern for loading efficiency and cost. An

alternative is to use a collection media in which the particles have poor solubility, as

seen for polylactide (PLA) particles electrosprayed into 70% ethanol, preventing the

leakage of the drug [64]. The use of additives in the collection media has also been

utilised for cross-linking of serum albumin-loaded chitosan capsules electrosprayed

into an aqueous tripolyphosphate solution, to improve the mechanical properties of

capsules [72]. Agglomeration in solution is a potential issue with hydrophobic

polymers when electrospraying in aqueous solutions. Coating is one approach to

enable better stabilisation of individual particles as seen for poly(lactic-co-glycolic

acid) (PLGA) particles electrosprayed into a poly(vinyl alcohol) (PVA) solution [50].


- 16 -

Figure 2.1. Electrospraying apparatus. (A) Formation of the Taylor cone in a nozzle-ring setup. (B)

Electrospraying via a nozzle-ring setup inside a glass chamber under air flow [60]. (C,D) Coaxial

electrospraying in a solution with size measurement by laser optical spectrometer [78]. (E) Single and

multiplexed electrospray setup on grounded collectors [49]. Adapted from [49, 60, 78] with

permission. 2006, 2010 Elsevier Science Ltd. [49, 60]; 2006 Royal Society [78].

When solid matrices, such as hydrogels, have been used to entrap particles, as

seen in cancer treatment where the containment of particles may be necessary, loaded

electrosprayed microspheres could be electrosprayed for a second time, from an

aqueous solution containing alginate, directly into a calcium chloride, CaCl2,

solution. The instantaneous gelation resulted in calcium-cross-linked hydrogel

macrobeads that held the microspheres within the matrix. Low voltages were used in

A B

D

E

Inner solution feed

Outer solution feed

Droplets irradiated by laser

Collection liquid

C

PC

Camera

Stirrer

High voltage supply

Syringe pump 2Outer liquid

Syringe pump 1Inner liquid

Syringe pump

High voltage supplies

To particle collector

HeaterDischarge electrode

Air inlet

ReservoirLiquid

Liquid Liquid

• Reservoir•Nozzle chip• Spacer• Extractor

• Additional extractor

• Voltage metre • Voltage metre

• Collector • Collector

V V

HV3

HV2

•Metallic nozzle

HV HV1

Glass chamber


- 17 -

this context so that the dripping mode of electrospraying occurred, generating

macrobeads with millimeter sizes. This mode is usually unwanted when

electrospraying nano/microparticles due to the macro-size outcome, but it does

present an interesting alternative for generating larger particles such as hydrogel

macrobeads that act as holding matrices. Again, the use of a surfactant such as

Tween 80® in the alginate solution is recommended so that the highly hydrophobic

microspheres stay uniformly suspended during dripping. According to gelation time,

CaCl2 concentration and microsphere loading, different release kinetics may be

obtained with this setup [53].

The most common collector for electrospraying polymer solutions containing

biomolecules remains a conductive and grounded collector such as an aluminium or

copper substrate [51, 52, 61, 63, 77] (Figure 2.1E). The use of a conductive substrate

restricts the deposition of particles to the charged area, limiting losses and does not

require any subsequent washing or filtering step.

In practice, despite electrospraying enabling better control over size and

morphology of particles compared to the traditional fabrication techniques, it is not

without associated drawbacks, including the low-throughput of the technique and

yields in the order of milligrams/hour [28]. This can be overcome with multiple

electrospray sources as seen in Figure 2.1E. An extractor is essential in this type of

setup to minimise interference between sources and to localise the electric field.

Morphology and size of microparticles were similar to that of the single setup and

particle production could be increased from milligrams to grams per hour using 19

parallel nozzles [49].

2.4.2.2 Encapsulation of Biomolecules

Conventional medication via oral or bolus administration typically does not provide

spatially or temporally controlled release of therapeutic molecules. The short half-

lives in solution of most of these molecules also imply that they lose their bioactivity

quickly following ingestion or implantation, or are rapidly cleared by the metabolism

in the body [11]. Such shortfalls require high doses of therapeutic molecules to be

used, resulting in increased cost and possible complications due to levels potentially

toxic for cells and tissues [9].

In recent years, the encapsulation of therapeutic molecules has become a powerful

tool for delivering controlled amounts to target cell populations and tissue sites, with


- 18 -

minimal signal propagation to non-targeted cells and tissues. Encapsulation can be

obtained by the processing of biodegradable polymers which maintain integrity and

relative long-term biological activity of therapeutic molecules. Polymeric devices

can finally provide an exposure for extended periods ranging from hours to months

by gradual polymer degradation allowing a specific release pattern of biomolecules

for treatment [7].

Several methods can be employed for the encapsulation of biomolecules (also

referred as drugs) into electrosprayed polymeric particles, as shown in Figure 2.2.

The resultant particles may be categorised into two distinct groups:

- particulate systems, where the drug is intimately distributed within the polymer

structure;

- capsules, where the shell is made of the polymer while the aqueous drug solution

is located in the core.

Capsules may be obtained by coaxial electrospraying shown in Figure 2.2, where

the aqueous core solution and organic shell solution are extruded independently

through two concentric nozzles leading to the electrospraying of particles with a

distinct core-shell structure. The bi-component syringe may be connected via tubing

to separate syringes with independent flow using two syringe pumps [28, 48].


- 19 -

Figure 2.2. Different methods of drug incorporation within polymeric particles through monoaxial

electrospraying (by aqueous nanoprecipitation, emulsification, and solid dispersion) for production of

particulate systems, and coaxial electrospraying for production of capsule systems.


- 20 -

Particulate systems may be obtained by monoaxial electrospraying where the drug

is mixed with the polymer solution before electrospraying commences, shown in

Figure 2.2. In the course of electrospraying the solvent evaporates from the droplet

and the drug remains entrapped within the polymer structure, ideally randomly

distributed. The drug can be mixed in its solid state, where it is directly dispersed in

the polymer solution and vortexed before electrospraying. The drug may also be

dissolved in an aqueous solution before mixing with the polymer solution, by

emulsification or nanoprecipitation as shown in Figure 2.2. Emulsions are widely

used in traditional encapsulation methods with the water-in-oil-in-water (w/o/w)

double emulsion being the most common used, since it provides access to a wide

range of particle sizes by adjusting the conditions of the process. For electrospraying,

a single water-in-oil emulsion (w/o) may be performed where hydrophilic molecules

are first dissolved in water before encapsulation. Different surfactants may be added

to tailor the encapsulation efficiency (EE) and release profiles [71]. However, the

interface between the organic and aqueous phases may result in protein denaturation

and aggregation, which is the main drawback of all emulsion-based methods [35,

77]. Nanoprecipitation, on the other hand, avoids the denaturation problem since

high shearing rates and interfaces are absent. However it can lead to agglomeration

and is not suitable for hydrophilic biomolecules due to leakage in the aqueous phase

[40]. Solid dispersion may thus remain the most attractive option in monoaxial

electrospraying, with no or limited denaturation and high versatility of drugs that

may be incorporated (both hydrophilic and hydrophobic, small molecule and protein

drug types).

Advantages of coaxial electrospraying include high drug encapsulation

efficiencies within the capsules and the assurance that the drug has minimum contact

with the organic solvent from the polymer solution, meaning less risk of drug

degradation. Nevertheless, the biomolecules remain in aqueous solution within the

capsules before delivery, which happen when the shell starts to degrade and channels

open for release. This is an issue since the stability of some biomolecules in the

aqueous state is known to be lower compared to the dry state, which may

consequently result in loss of bioactivity [71]. Nevertheless coaxial electrospraying

supposedly allows better control over release kinetics due to an increased number of

variable parameters [48, 80, 81].

Section 2.5 Control of Particle Characteristics with Electrospraying Parameters

- 21 -

More complex devices such as tri-needle coaxial devices can allow for more drugs

to be loaded within separated layers of the capsule for sequential and multiple release

[79, 82]. This can also be achieved with normal coaxial electrospraying with loading

of a second drug in the polymer core. However it was previously shown that the

polarity of the drug is of importance and capsules that contain a hydrophobic drug in

the core and a hydrophilic drug in the shell can easily be made, whereas the opposite

configuration is more difficult to achieve [58].

2.5 CONTROL OF PARTICLE CHARACTERISTICS WITH

ELECTROSPRAYING PARAMETERS

2.5.1 Importance of Electrospraying Parameters

Although electrospraying is accepted as a technique which can produce particles with

monodisperse size distributions and reproducible morphologies by controlling the

electrospraying parameters [49, 83], producing particles with very specific

requirements remains challenging due to the large number of variables involved in

the process and their complex inter-dependence. The primary pre-requisite for

reproducible electrospraying and monodisperse size production is the stable cone-jet

mode, for which the working window can be found by tailoring the field strength,

conductivity and flow rate of the polymer solution. Morphology and size can be

further controlled by adjusting additional parameters, such as the polymer

concentration and molecular weight, the solvent vapour pressure, the flow rate, the

electrospraying distance and chamber environment.

2.5.1.1 Polymers

2.5.1.1.1 Polymer Types

Currently, the most common synthetic biodegradable polymer used in the field of

drug delivery and also most commonly used in electrospraying is poly(lactic-co-

glycolic acid). This aliphatic polyester is approved by the American Food and Drug

Administration (FDA) and it is widely used in several medical devices (sutures,

grafts, prostheses) and drug delivery devices [7]. PLGA degrades mainly through

hydrolysis, which distinguishes it from natural biodegradable materials such as fibrin

and collagen which are actively degraded enzymatically in the presence of cells. The

natural products arising from degradation (lactic and glycolic acid) are then cleared


- 22 -

by metabolic pathways [84]. In contrast, the degradation of PLGA via hydrolysis

results in the generation of acidic species which can provoke inflammation of tissues

and generate problems for long-term stability when encapsulating bioactive

molecules. For instance, in vitro simulations of the polymer microclimate of PLGA

microparticles produced using traditional methods revealed a highly acidic

environment (pH < 3), further triggering unfolding of encapsulated serum albumin

(SA), resulting in peptide bond hydrolysis and non-covalent aggregation [85]. Anti-

acid excipients such as magnesium carbonate (MgCO3) or magnesium hydroxide

(Mg(OH)2) can be used in the microparticle fabrication process to buffer the pH [86].

SA structural losses and aggregation were indeed prevented for over one month with

Mg(OH)2 from PLGA microparticles and this strategy was further employed for

delivering angiogenic basic fibroblast GF and bone-regenerating morphogenetic

protein-2 [85]. However no studies so far have tested these anti-acids in

electrosprayed particles, where microclimate pH was mentioned but not addressed

[71], although PLGA remains the most utilised polymer for electrosprayed particles

[52-59, 71].

Polylactides (PLAs) have similar properties to PLGAs but they afford a more

crystalline structure responsible for a slower degradation [58]. Select studies have

chosen pure PLAs over the PLGA copolymers for monoaxially electrosprayed

particles [60, 64, 73, 74], or simultaneously in coaxial electrospraying with PLA as

the core and PLGA as the shell to ensure that the drug in the core was not released

prematurely by choosing a core of slower degrading material than the shell [58, 87].

Very different molecular weights (MW), such as 2 kDa [64] and 175 kDa [73], have

been chosen when electrospraying PLAs. MW has a great influence on degradation

and thus subsequent release of encapsulated biomolecules; PLA 2 kDa degrades

quicker in vivo and is more soluble than higher MW PLAs [64].

The biodegradable polyester polycaprolactone (PCL) is an interesting candidate

for drug delivery and has also been used in electrospraying [55, 59, 66, 68, 71].

Compared to PLGA and PLA, PCL is semi-crystalline with a melting temperature of

approximately 60°C and a glass transition temperature around -60°C (compared to

40 to 65°C for PLGA/PLA) conferring superior viscoelastic properties and easy

formability [88, 89]. PCL is also FDA-approved and various drugs have been

encapsulated in PCL microspheres and nanospheres since PCL is highly permeable

to small drug molecules. Due to its crystallinity and lower ester concentration, PCL


- 23 -

presents the advantage of a less acidic environment being generated during

degradation as compared to PLGA-based polymers [68, 90]. Nevertheless, the high

hydrophobicity of PCL remains a concern for encapsulation of hydrophilic

substances such as peptides, enzymes and other proteins [18].

Although all the aforementioned biodegradable polyesters, PLGA, PLA and PCL,

have generated considerable interest in the last decades as potential matrices for drug

delivery, overall concerns remain, particularly with regard to slow degradation and

hydrophobicity (in the case of PCL) and acidic environment generation (in the case

of PLGA and PLA) leading to possible instability, aggregation and structural

changes of the loaded drug/protein. The introduction of functional groups can

provide these polymers with tunable crystallinity and enhanced hydrophilicity. The

description of such functional polymers and use so far in the field of drug delivery

has been recently summarised in the review by the group of Hennink [91].

An elegant approach to improve the utility of PCL is to copolymerise with a more

hydrophilic commoner. For example, PCL has been functionalised with hydrophilic

components such as polyamino ethyl ethylene phosphate (PPE-EA), in order to

improve hydrophobicity. The amphiphilic block copolymer, PCL-PPE-EA, was

indeed shown to encapsulate SA more efficiently than PCL alone [92]. During the

w/o emulsion procedure, when the protein is introduced into the polymer solution,

micelles are formed around the protein with the hydrophilic part of the polymer

(PPE-EA) in contact with the protein. Such micelle-derived electrosprayed particles

encapsulating SA were 3 µm in diameter and exhibited a linear release profile for 20

days whereas no protein was released from PCL only particles. Unfortunately, the

formulation and processing parameters of PCL particles loaded with SA were not

described in the study and the release data was not normalised to the amount of

loaded protein, rendering the assessment of the system delicate [92].

Natural polymers have also been electrosprayed, including elastin-like

polypeptides (ELP) [61, 93], a bioresponsive biopolymer that can be dissolved in

water, an advantage compared to polyester-based polymers that require organic

solvents to dissolve them. ELP are inspired by the amino acid sequence of natural

elastin and can be synthesised by recombinant DNA methods, allowing for a control

over the ELP sequence and thus over its biofunctionality [94]. Chitosan is another

natural polymer that has been electrosprayed. Chitosan comes from the alkaline

deacetylation of chitin and its main advantage is that it is hydrophilic, which


- 24 -

improves facilitation of drug-polymer interactions compared to hydrophobic

synthetic polymers [72]. Its performance as a drug delivery system is affected mainly

by its molecular weight and degree of deacetylation, while its cationic nature allows

for ionic cross-linking for improved material properties [51, 72]. Compared to

synthetic polymers, the degradation products of chitosan are amino sugars, which are

easily metabolised by the body [51], therefore there is no concern of an acidic

microclimate being generated by chitosan particles. So far, the use of chitosan in

electrospraying has been used limited to encapsulating SA [72], ampicillin [51], an

antibiotic to treat bacterial infections, doxorubicin [95], an anti-cancer agent, and

insulin [96]. In the case of SA and doxorubicin, the microparticles were

electrosprayed in a tripolyphosphate (TPP) solution, a non-toxic biocompatible

cross-linking agent ideal for chitosan [72, 95].

Miscellaneous polymers that have also seen use in electrospraying applications

include poly(amidoamines) (PAA)-cholesterol conjugates, for encapsulation of

tamoxifen [70]. Along with their amphiphilic character – due to the presence of

cholesterol, and low molecular weight (13 kg/mol), PAA-cholesterol conjugates are

likely to produce nanosized particles with a low degree of polymer chain

entanglements, thus providing rapid drug release rates (within hours).

Polyvinylpyrrolidone, a water-soluble polymer, has also been used for self-assembly

of nanoparticles including tristearin, a lipophilic excipient and naproxen, an anti-

inflammatory drug [67]. Although the versatility of electrospraying allows the use of

many types of polymers, only a restricted number of polymers have been tested so

far for encapsulation of biomolecules. Many more polymers remain to be

investigated, for providing a higher degree of diversity in terms of physical and drug

release characteristics of particles, as well as possibly enhanced drug-polymer

interactions.

2.5.1.1.2 Solvents

Organic solvents are required to solubilise polymers prior to electrospraying. The

most widely used solvent for electrospraying particles loaded with drugs is

dichloromethane (DCM), a chlorohydrocarbon with the lowest boiling temperature

(40°C) of the common solvents used in electrospraying. Other solvents include (by

increasing boiling temperatures): acetone [74], chloroform [52], ethanol [64],

acetonitrile [55], 1,2-dichloroethane [73, 74], acetic acid [51, 72], and N,N-


- 25 -

dimethylformamide (DMF) [74], which may be used alone or in combination. The

boiling temperature of a solvent is the temperature at which the vapour pressure

equals the ambient atmospheric pressure and it is representative of the solvent’s

volatility. Solvents with low vapour pressure (high boiling temperatures) are

vapourised less easily than solvents with high vapour pressure (low boiling

temperatures) and are thus less volatile.

This means that polymer diffusion is reduced in electrosprayed droplets from

solvents with high vapour pressures, where solvent evaporation occurs at a higher

rate. This affects the size and morphology of particles and it was previously shown

that an increase in boiling point, corresponding to a decrease in volatility, correlated

with a decrease in particle size with smoother surfaces generated for solvents with

boiling temperatures above 140°C (such as DMF, 146°C) [97]. A greater particle size

and more textured surfaces can be seen with solvents with low boiling temperatures

such as chloroform (61°C) [83] and dichloromethane (40°C) [98]. This is due to fast

solvent evaporation, where less time is available for polymer chains to contract and

re-arrange within the evaporating droplet exposed to electric field. Faster evaporation

can also result in the formation of pores [94] and even hollow particles [60, 99].

Importantly, it was shown that a decrease in vapour pressure weakens the forces

of polymer chain entanglements [100]. Therefore, the Coulombic repulsion is able to

overcome the surface tension of evaporating droplets, possibly leading to the ejection

of small and highly charged offspring droplets. This was seen with PLGA particles

where the addition of 30% DMF to chloroform reduced the vapour pressure from 21

kPa to 15 kPa and induced a bimodal size distribution, made of primary and

offspring droplets. The use of a co-solvent with low vapour pressure is thus not

recommended to obtain monodisperse particles [100].

It must also be noted that different polymers have different interactions with

solvents, affecting polymer chain entanglements and final morphology of particles.

Both polymer concentration and molecular weight greatly dictate these interactions

[101].

2.5.1.1.3 Polymer Solutions

When electrospraying polymer solutions, electrosprayed droplets undergo solvent

evaporation and polymer diffusion simultaneously. Chain entanglements occur

during these processes and are responsible for the final morphology of particles. In


- 26 -

electrospraying, uniform microparticles and smaller droplets are favoured by limiting

chain entanglements [102]. The number of entanglements per chain in solution,

(ne)sol, can be expressed with the polymer volume fraction φ, the average molecular

weight Mw and the average entanglement molecular weight Me according to Equation

2.3:

( )

(2.3)

It was previously shown that electrospraying occurred for 1 entanglement per

chain ((ne)sol = 2) whereas 2.5 entanglements per polymer chain ((ne)sol = 3.5) would

lead to formation of fibres; a process known as electrospinning [102]. Beaded fibres

can form for intermediate values of (ne)sol. Me is primarily a function of chain

geometry and corresponds to the average molecular weight between entanglement

junctions. Me is readily available for more than 70 polymers but in the absence of

experimental values, it can be theoretically estimated by employing the entanglement

constraint model used by Shenoy et al. [102].

Polymer concentration plays an important role in the entanglement regime which

dictates particle or fibre formation and is an essential parameter to control in order to

optimise the electrospraying process. The critical chain overlap concentration, Cov, is

known as the point when the average distance between chains is on the same order as

their size and is inversely proportional to the intrinsic viscosity [η], as shown in

Equation 2.4 [103]:

(2.4)

When the concentration C is below Cov, there are no chain entanglements and the

regime is known as the dilute regime (Figure 2.3A). Above Cov, the concentration is

large enough for chains to overlap but not sufficient to generate a significant degree

of entanglement. The regime is the semi-dilute unentangled regime, and some

entanglement is observed (Figure 2.3B) although not desirable since particles have

the ability to deform during evaporation, leading to inferior, non-reproducible

morphology. Such a regime can be used for the production of electrosprayed films,

another type of delivery device useful in some therapies such as chemotherapy.

Multiple layers of polymers encapsulating various drugs can thus be made by


- 27 -

electrospraying in either the dilute or semi-dilute unentangled regime, allowing for a

controlled release of therapeutic molecules tailored by the thickness of the films [57].

b

Figure 2.3. Left column: Physical representation at the molecular level of various entanglement

regimes obtained for different polymer concentrations. Rg is referred as the radius of gyration.

Adapted from [103] with permission. 2005 Elsevier Science Ltd. Right column: Examples of

corresponding scanning electron micrographs of dried PCL microparticles. PCL concentration in

chloroform was: (A) 5%, (B) 7.4%, (C) 8.7%, (D) 9.6% wt/v. Electrospraying conditions were 26 G

for needle gauge, 20-25 cm for tip-to-collector distance, 0.5 mL/h for flow rate and 10 kV for voltage.

The molecular weight of PCL on average was 130 kg/mol with a polydispersity index of 1.45. Scale

bar is 10 µm.

Polymer chain

Rg

(A) Dilute regime C < Cov

(B) Semi-dilute unentangled regime Cov < C < Cent

(C) Semi-dilute moderately entangled regime Cent < C < 3Cov

(D) Semi-dilute highly entangled regime C > 3Cov


- 28 -

For electrospraying of particles, the regime of choice is the semi-dilute moderately

entangled regime. It happens for Cent, the crossover from the semi-dilute unentangled

regime to the semi-dilute moderately entangled regime, where a significant degree of

entanglement is observed and dense, solid and reproducible particles can be

produced (Figure 2.3C). However, for C/Cov>3, molecular cohesion is generally too

high for electrospraying and beaded fibres or fibres are electrospun, corresponding to

the semi-dilute highly entangled regime (Figure 2.3D). For optimal particle

electrospraying, it is thus essential to work above Cent but not overcome C/Cov>3.

The molecular weight and molecular weight distribution (MWD) do affect Cov due to

differences in intrinsic viscosity, and it was demonstrated that an increase in MW

reduces the C/Cov ratio, narrowing the working window of the semi-dilute moderately

entangled regime, thus narrowing the range of appropriate concentrations for

reproducible electrospraying. On the other hand, for broader MWD, the ratio C/Cov

required to obtain the semi-dilute highly entangled regime was shown to be higher

than 3, broadening the working window of the semi-dilute moderately entangled

regime where reproducible electrospraying can be obtained [103].

2.5.1.2 Processing Parameters

2.5.1.2.1 Spraying Modes

Different spraying modes can take place in the course of electrospraying and they

vary according to the field strength and flow rate of the polymer solution. The

magnitude of the field strength is a key to reproducible spraying patterns [59] and its

variation leads to different spraying modes, starting from the dripping mode and

moving to cone-jet modes with increasing applied voltage [28]. When sufficient

voltage is applied to the droplet to form the Taylor cone (corresponding to the

change-over from dripping mode to cone-jet mode), the ejection of small and highly

charged droplets assumes the form of a cone which proportionally increases with an

increase in the tip-to-collector distance. This single cone-jet mode seen for moderate

field strengths is stable and fairly consistent from one replicate to the next [68].

Conversely, when increasing the field strengths, multiple cone-jets are formed, which

are unstable and unpredictable, and importantly can vary throughout the course of

electrospraying [59]. Such modes can be found in all types of electrospraying setups

and are also observed for the nozzle-ring setup when increasing the potential

difference between the nozzle and ring [55]. The multiple cone-jet mode needs to be


- 29 -

avoided so that only targeted areas are sprayed, in order to ensure a high yield of

particles. This is especially important when loading expensive molecules, where

minimal loss is desired.

One strategy to obtain the single cone-jet mode is to lower the electrical

conductivity and surface tension of the solution [68]. When incorporating therapeutic

molecules to the polymer solution, the stable mode can be maintained by decreasing

the protein concentration and the loading since the electrical conductivity increases

with increasing protein concentration, as has been shown for serum albumin [74,

104]. As a result, the stable single cone-jet mode region shrinks and shifts to a lower

flow rate for higher protein concentration (Figure 2.4A-B). On the other hand,

increasing the viscosity of solutions (by increasing polymer concentration for

instance) results in a shift of the cone-jet mode to higher voltages, as seen in Figure

2.4C. This is because of the lower conductivity of more viscous solutions: a stronger

electric field should be applied to overcome the surface tension and liquid viscosity

to form the cone-jet [68].

Figure 2.4. Mode selections maps to obtain different electrospraying modes, for (A) 5.5 mg/mL and

(B) 20 mg/mL as-prepared serum albumin solution. In the case of an unstable jet, a clear mode

classification was not possible. Microdripping and spindle both refer to undesirable electrospraying

modes [104]. (C) Cone-jet mode maps for different PCL solutions [68]. Adapted from [68, 104] with

permission. 2005 Springer [104]; 2010 Royal Society [68].

It is very important to keep in mind that only in the stable cone-jet mode is the

production of narrowly dispersed particles possible. Only then can the size and

morphology of particles be controlled by carefully changing other parameters.

2.5.1.2.2 Electrical Conductivity

Since electrospraying depends on the electrostatic attraction of charged particles to a

grounded or oppositely charged collector, the electrical conductivity, K, of the

polymer solution is an important parameter when optimising the process. Along with

Flow rate (µL/min)0 15 30 45

6.5

9.5

12.5

15.5

PCL 10 wt%

PCL 5 wt%

PCL 2 wt%

C

45

5.2

6.4

15 30Flow rate / 10-11 m3/s

Unstable jet

Unstable cone-jet

Spindle

Unstable jet

Stable cone-jet

Microdripping4

A

Ap

plie

d v

olt

age

(kV

)

15 30 45

Unstable jet

SpindleUnstable jet

Microdripping

Unstable cone-jet

Stable cone-jet

5.2

6.4

Flow rate / 10-11 m3/s

4

B


- 30 -

flow rate, electrical conductivity provides a powerful means to control the

electrosprayed particle size, as demonstrated by the scaling laws from Gañan-Calvo,

where a higher conductivity leads to a decrease in size [105].

An increased conductivity of a solution implies that more charge is carried by the

electrospraying jet. In general, a low electrical conductivity is preferred to obtain

quasi-monodisperse particles [100] since a higher conductivity may favour elongated

particles or even fibres if the polymer concentration is high enough [106].

Correlating with viscosity, stable electrospraying is known to be achieved only when

viscosity is high or conductivity is low [107]. Changes in electrical conductivity can

be obtained by changing the electrospraying solvent or using co-solvents, although

this latter case may be detrimental to size distribution and morphology of particles

[74, 100]. Organic solvents are generally less conductive than aqueous solvents and

their conductivity can be increased by the addition of electrolytes, such

didodecyltrimethylammonium bromide (DTAB) [56] or ammonium hydroxide [64],

which can increase conductivity by orders of magnitude. For instance the

conductivity of a 5% wt/v PLGA solution in acetonitrile containing 10% wt

paclitaxel was shown to increase from 0.51 μS/cm to 116.5 μS/cm by the addition of

2 mM DTAB. This led to a particle size decrease from around 1.2 μm to 355 nm

[56]. Compared to pure solvents, it must be kept in mind that the addition of a

polymer will most likely decrease the electrical conductivity, although remaining in

the same order of magnitude [108].

When the electrical conductivity of the solution is lower than 0.01 µS/m, it is

likely that insufficient current can flow, and the liquid cannot be electrosprayed,

although too a high conductivity value leads to unstable electrospraying [59]. The

bending instability of the jet becomes more important when more charges are present

due to increased conductivity, leading to a wider deposition of particles on the

collector. With higher electrical conductivity, the Coulombic repulsion forces are

higher and compete with the viscoelastic forces of the solution, disentangling more

easily the polymer network which is being formed during electrospraying. In other

words, increasing conductivity makes it easier for the solution to be broken up into

smaller droplets. Therefore for the same polymer dissolved at the same concentration

in a higher conductive solvent (or the same solvent but supplemented with organic

salts), disentanglement may take place during electrospraying, in turn reducing the

final particle size. Furthermore, if the Coulombic repulsion forces are sufficiently


- 31 -

high to overcome the entanglement forces, then Coulomb fission occurs before

strong entanglements can form, and smaller offsprings are ejected from the primary

droplet. This will provide a bimodal size distribution, with particles presenting

various types of morphologies, mostly unwanted and further discussed in the section

2.5.2. Low electrical conductivity may thus be more favourable for electrospraying

of quasi-monodispere microparticles.

When nanoparticles are required, increasing conductivity may be a good means of

reducing particle size, although sufficient viscosity needs to be ensured so that

entanglement forces remain higher than Coulomb forces, and the ejection of

offspring droplets is avoided. Higher flow rates can also be used to produce

nanoparticles – if higher salt concentration is used to increase solution conductivity

[56]. In the context of electrospraying emulsions, the organic/aqueous volume ratio is

another significant factor influencing the electrical conductivity whereby addition of

water to the organic phase significantly increases the electrical conductivity of the

resulting emulsions [73].

2.5.1.2.3 Flow Rate

After the selection of polymer solutions, flow rate is arguably the second most

important parameter in electrospraying and together with the solution parameters

(polymer MW, concentration, solvent, and conductivity) can control polymer

entanglements and Coulomb fission [49]. Flow rate thus has consequences for both

the morphology and size of particles and must be judiciously chosen since both these

characteristics will influence the drug dispersion within the polymer matrix,

ultimately affecting drug release.

Firstly, it is essential to use a flow rate that allows for complete solvent

evaporation, which is not possible with high flow rates. Particles are partially

solvated when they impact the collector leading to a deformed and non-consistent

morphology [55, 83]. Furthermore, too high flow rates can confer a bimodal or

polydisperse character to the size of electrosprayed particles. This is explained by the

processes involved in solvent evaporation from the charged droplet, based on φRay,

the polymer volume fraction in a droplet at the Rayleigh limit, and expressed in

Equation 2.5:

(

)

(2.5)


- 32 -

where Q is the liquid flow rate (m3/s), I the current, εair the permittivity of air, γ the

surface tension of solution in ambient air, and d the initial droplet diameter. I/Q and d

can be determined as a function of Q and the polymer solution properties.

Considering φent as the critical entanglement polymer volume fraction and φov as

the critical chain overlap polymer volume fraction, it was shown by Almería et al.

that for φov < φRay < φent in the semi-dilute unentangled regime mentioned previously,

the polymer network can preserve some droplet integrity, but is not strong enough to

preserve the particle from deforming via stretching during the fission process, while

droplets are stabilised from such rupture when φRay > φent [49]. For φRay < φov, the

droplets behave like a pure liquid and there are no entanglements, leading to the

ejection of offspring droplets from the primary droplet, a consequence of Coulomb

fission. In order to obtain a spherical morphology and monodispersity, it is important

that sufficient entanglements are present before the Rayleigh limit is reached so that

the droplet cannot be disrupted by Coulomb fission, ensuring that φRay > φent.

According to Equation 2.5, flow rate has a significant influence on φRay and it was

shown for the morphology of PLGA particles that larger flow rates decreased φRay,

generating non-spherical morphology with the possible formation of offspring

droplets and extruded fibres when concentration was sufficiently high [49].

In addition to offspring droplets that can form, high flow rates can also generate

secondary droplets, even in an entangled network, due to the perturbation amplitude

of the electrospraying jet, which is increased with increased flow rates [100]. This

can be explained by the phenomena occurring when the droplets are ejected from the

Taylor cone. Initially, a filament unites 2 droplets, but it is further broken up by the

charge. Once broken from the farthest droplet, the filament flows back to the nearest

droplet from the cone, and monodisperse particles can be achieved, for relatively low

flow rates. At increased rates, there is more distance between evaporating droplets.

Thus the filament may not reach the former droplet anymore and instead it breaks,

forming a secondary smaller droplet. At even higher flow rates, a filament between

primary and secondary can form, which, being unable to reach back to the primary

droplet, turns into a satellite droplet (even smaller than secondary droplets) [42]. If

the solvent has a high evaporation rate, it is even possible that the filament remains

frozen, leading not only to polydisperse sizes but also leading to elongated particles

[49].


- 33 -

The same rules apply for coaxial electrospraying where the inner and outer flow

rates may strongly affect the properties of electrosprayed capsules. Usually the flow

rate of the core solution is much lower than the shell solution [77], resulting in

uniform sizes. In contrast, for a shell flow rate slower or equal to that of the core

flow rate, irregular morphologies are obtained, while for increasing ratios of

shell:core flow rates, the shell becomes thicker [80]. This presents a useful tool for

the tailoring of release kinetics.

2.5.1.2.4 Other Parameters

Effect of Gauge

The diameter of a needle is commonly expressed in gauge (G), each gauge size

arbitrarily correlating to multiples of 0.001 inches [109]. For the electrospraying of

particles loaded with bioactive molecules, these diameters range from 18 G (internal

diameter (ID) of 1.27 mm) [73, 74] to 29 G (ID of 0.33 mm) [56]. Prior to

electrospraying, beveled needles are typically shortened and given a flat end for

homogeneous spraying, although characterisation of the needle tip, while important,

is often overlooked. The effect of gauge has little effect on morphology or size of

particles. For instance when comparing the size of PCL particles made with 21 G

versus 26 G, the average size was equivalent for both gauges, however the size

distribution was slightly broader for the bigger gauge (21 G) with standard deviation

(SD) of 3.42 while SD was 2.40 for 26 G, suggesting that a smaller gauge can

produce a narrower size distribution [83]. A similar result was observed in the 20 to

26 G range when electrospraying ampicillin-loaded chitosan nanoparticles, where the

use of the 20 G led to sputtering only, 22 G led to a mixture of particles and

sputtering, while 24 and 26 G led to spherical particles with no sputtering and with

reduced polydispersity for the smallest gauge (26 G) [51].

Effect of Voltage

The main incidence of voltage is on spraying modes as described previously in

section 2.5.1.2.1. Within the single cone-jet mode, size is not significantly affected

by voltage where only a slight decrease in size is observed when voltage is increased

[52]. Morphology however will be changed as stated by Shenoy et al., since as the

voltage is increased, the morphology changes from spherical particles to elongated

particles or beaded fibres to eventually only fibres if concentration is sufficiently

high [102]. This is due to more charge acting on droplets with increased voltage,


- 34 -

leading to stretching and elongation of droplets. It is therefore recommended to use

moderate voltages that allow for the single cone-jet mode to take place while

maintaining the spherical morphology of particles.

Effect of Tip-To-Collector (TTC) Distance

The lower limit of distance is determined by electric discharge. A small TTC

distance can impair full solvent evaporation and consequently, wet microspheres

impact the collector, leading to collapsing, coalescence and broad size distributions

[51]. Increasing the distance leads to more spherical morphologies since polymer

chains have sufficient time to diffuse within the droplet [83] and thus also reduced

polydispersity. At constant voltage, a decrease in the TTC distance leads to an

increase in the strength of the electric field, thus leading to a decrease in particle size

[108]. Depending on the type of solvent used, an increase in TTC distance may also

be detrimental for morphology as shown with polyacrylonitrile microspheres in DMF

where at 10 cm, the evaporation rate of DMF allowed round spheres to be formed,

while at 20 cm, the round spheres collapsed into half-hollow spheres. The authors

stated that the evaporation rate was excessive. However this result was not clearly

evident in the images shown by the authors, and no explanation was proposed [110].

2.5.2 Tailoring of Electrosprayed Particle Characteristics

2.5.2.1 Morphology

The morphology of electrosprayed particles is controlled by solvent evaporation and

polymer diffusion [49]. The polymer solution thus plays a determinant role in these

mechanisms, where the nature of polymer (solubility, molecular weight,

concentration) and solvent (vapour pressure, miscibility, conductivity of solution)

coupled with the solution flow rate form the levers of morphology tailoring [83]. As

explained previously in the section 2.5.1.1.3, regarding polymer solutions,

concentration and molecular weight can dictate the entanglement regime taking

place, leading to reproducible and solid electrosprayed particles, when a certain

degree of chain entanglement is obtained. Therefore in most studies, morphology is

initially linked to polymer concentration and molecular weight, where a decrease in

concentration or an increase in molecular weight induces non-spherical morphologies

such as shell-like, wrinkled, hollow particles, beaded fibres or particles with tails

(Figure 2.5A-B). However, as seen in the section 2.5.1.2.3, flow rate also has a

significant influence on morphology through φRay, the polymer volume fraction in a


- 35 -

droplet at the Rayleigh limit (expressed in Equation 2.5), where it was shown that

φRay needs to be greater than φent for obtaining spherical morphology. Once this is

ensured, subsequent flow rate alteration may lead to various morphologies; while

lower flow rates may cause beaded fibres to form for too viscous solutions, larger

flow rates may disentangle the charged droplet leading to non-spherical morphology.

Figure 2.5. Importance of electrospraying parameters. (A-B) Effect of various parameters on the

morphology of PLGA particles: (A) Effect of concentration and molecular weight (flow rate (FR) = 1

mL/h and applied voltage (AV) = 10 kV). (B) Effect of concentration and liquid flow rate (MW = 38

kDa and AV = 10 kV) [100]. (C) Relationship between mean particle size, electrical conductivity and

viscosity of PCL solutions (FR = 10 µl/min, AV = 10 kV). Square with solid line: mean size; circle

with dotted line: conductivity. (D) Polydispersivity index of PCL particles produced in the cone-jet

region as a function of flow rate (AV = 10kV). Square line: PCL 2% wt (viscosity: 2.6 mPa.s);

triangle line, PCL 5% wt (viscosity: 4.6 mPa.s); diamond line: PCL 10% wt (viscosity: 11.1 mPa.s)

[68]. (E) Dependence of the size and shape of microcapsules made by coaxial electrospraying on the

flow ratio between outer and inner solutions [80]. (F) Diagram depicting the influence of parameters

on particle diameter (↑: increase) [48]. Adapted from [48, 68, 80, 100] with permission. 2009 John

Wiley and Sons [100]; 2010 Royal Society [68]; 2008 American Chemical Society [80]; 2009 Elsevier

Science Ltd. [48].

These theories were confirmed for serum albumin-loaded PLGA microparticles

when polymer concentration (in DCM) was decreased from 10 to 6%, and wrinkled

particles were obtained instead of dense and spherical particles [71]. Similarly, this

trend was observed in paclitaxel-loaded PLGA microparticles where the morphology

changed from spherical to shell-like shapes when decreasing polymer concentration

from 8 to 6% [55], while a decrease from 2 to 1% in ampicillin-loaded chitosan gave

abnormal shapes instead of spherical particles seen for 2% [51]. Hollow particles

were seen when decreasing the concentration of aqueous elastin-like polypeptides

from 1 to 0.5% while spherical spheres were initially obtained at higher

Tailed particles, fibres

Spherical particles

Debris2

4

6

8

0

10

PLG

A c

on

cen

trat

ion

(wt%

)

0.5 1 1.5 2

Flow rate (mL/h)

Particle size

Small particle sizes with:

•Electric field strength ↑

•Fluid conductivity ↑

•Solvent volatility ↑

•Surface tension ↑

•Distance to collector ↑

Large particle sizes with:

•Polymer concentration↑

•Molecular weight ↑

•Flow rate↑

•Needle diameter ↑

•Viscosity↑

A B

F

Me

an p

arti

cle

siz

e (µ

m)

Viscosity (mPa s)

3

1

5

0 5 10 Co

nd

uct

ivit

y (µ

S/m

)

2

0

4

C

PLG

A c

on

cen

trat

ion

(wt%

)

4

8

12

16

020 40 60

Tailed particles, fibres

Spherical particles

Debris

PLGA molecular weight (kDa)

5

3

1

7

Par

ticl

e d

iam

ete

r (µ

m)

0:1

Feed ratio (inner:outer)1:1 1:2 1:3 1:4

E

Po

lyd

isp

ers

ivit

y (%

)

20

5

35

Flow rate (µL/min)0 15 4530

D


- 36 -

concentrations [61]. The effect of polymer concentration was also considered with

MW where a higher MW (70,200 g/mol versus 17,800 g/mol) provided tailed

structures or beaded fibres instead of spherical particles for both concentrations [61].

As explained previously, this is a consequence of high viscosity with stronger

entanglements taking place for high MW polymer chains, impairing full jet break-up.

In coaxial electrospraying, flow rates of both core and shell solutions are

determinant for reproducible morphology and size of capsules, where inner flow

rates are required to be lower than outer flow rates. Decreasing the inner flow rates

led to a thicker capsule shell and reduced particle size (Figure 2.5E) [80]. When

loading paclitaxel and suramin in the PLGA shell and poly-L-lactide (PLLA) core of

microcapsules, respectively, it was concluded that a Qcore ranging between 1.0 and

2.0 mL/h and a constant Qshell of 2.0 mL/h may maintain a stable cone-spraying mode

and consequently result in uniform and smooth microspheres with varied core sizes

[58, 87].

In electrospraying of emulsions, a decrease in the organic/aqueous phase volume

ratio (from 20:1 to 6.7:1) led to a degeneration of the spherical shape of particles

with a more wrinkled surface. This was tentatively explained by a corresponding

decrease in viscosity which would hinder the shrinkage of droplets during

evaporation [73].

Similarly, loading of biomolecules into the particles affects the morphology and

wrinkled particles were observed for a 30% - and above - loading of rifampicin in

PLGA particles [52]. This is a consequence of the difference in molecule types:

PLGA is a linear macromolecule whereas rifampicin is a small organic molecule. It

was stated that the addition of rifampicin decreased the diffusion coefficient of

solutes and weakened the intermolecular entanglements of PLGA. When the

concentration of the drug is higher than a critical value, diffusion is slower than

solvent evaporation, resulting in the increase of drug concentration near the front of

the droplets. Drug molecules accumulate and form a layer of semi-solidified skin on

the surface. With further evaporation of solvent, the intermolecular polymer

entanglements in the droplet skin predominate, forcing the semi-solidified skin to

collapse, leading to wrinkles [52].


- 37 -

2.5.2.2 Size

The size of polymeric particles that contain bioactive molecules significantly

influences their therapeutic capabilities. For instance, the release rate increases with a

larger surface to volume ratio of the particles [56] and is dependent on surface

diffusion and degradation. Control of size is therefore essential for tailoring release

properties. The electrospraying technique can produce particles with sizes varying

from tens of micrometers to tens of nanometres [63] and by choosing the right

parameters, low polydispersity can be obtained with relative standard deviations

(RSD) within 2 to 27% of the average size [62, 71]. This is advantageous in drug

delivery since drug distribution within the matrix can be controlled more precisely

with a single known particle size, allowing degradation rates and diffusion of drugs

to be better tailored to fit a desired application [44]. However, this is a constant issue

when microparticles are made from double emulsion fabrication methods where

broad distributions are obtained, ranging from 49 to 110% RSD [50, 64].

When nanoparticles are considered, the control of size and polydispersity

becomes even more important since they can greatly affect cell response mechanisms

where particles are internalised by cells. This involves particles lower than 500 nm

for uptake by epithelia cells for example [46], and lower than 100 nm for

applications such as tumour targeting by the enhanced permeability and retention

(EPR) effect. However, although the electrospraying technique allows the generation

of nanoparticles such as pharmaceutical nanoparticles or non-organic nanoparticles

(for coating for instance), when polymeric carriers were used to encapsulate drugs,

the resulting electrosprayed particles were mainly found to be on the micron to

submicron size. The minimum size reported so far is 116.1 nm with budesonide-

loaded PCL particles [66] and very few studies are found in the 100 to 500 nm range

[56, 61, 64, 70, 95]. This is likely due to polymer chains used in electrospraying of

polymers with drugs since a significant molecular weight and polymer concentration

are needed in order to efficiently encapsulate a drug within the polymeric matrix.

The nanosize of electrosprayed polymer/drug systems is thus less likely and most

systems actually produce micrometric sizes as seen in Table 2.1.

.


- 38 -

Table 2.1. Size and polydispersity (expressed as the relative standard deviation to the average particle

size (RSD) in %) achieved by electrosprayed polymeric particles loaded with various types of

therapeutic molecules. BDP, SS and SA stand for beclomethasone dipropionate, salbutamol-sulfate,

and serum albumin, respectively.

Size

domain

Size range

studied

Size with

lowest

polydispersity

RSD

(%) Polymer

Loaded

molecule Ref.

Below

500 nm

116.1 - 165 nm 165 nm 11.5 PCL Budesonide [66]

- 247 nm 7.2 PAA-

cholesterol Tamoxifen [70]

304.9 - 569 nm 304.9 nm 6 TPP-

Chitosan Doxorubicin [95]

255 - 355 nm 335 nm 18.2 PLGA Paclitaxel [56]

- 370 nm 6.8 Elastin-like

polypeptides Doxorubicin [61]

250 - 500 nm 470 nm 2.1 PLA BDP/SS [64]

500 nm -

1 µm

510 - 630 nm 630 nm 11.1 PLA BDP/SS [64]

- 840 nm 21.4 PLA SA [73]

580 - 910 nm 910 nm 12 PLGA Rhodamine [50]

1 - 5 µm

1.12 - 1.34 µm 1.12 µm 10 PLGA Rhodamine [50]

1.34 µm 10 PLGA Doxorubicin [50]

1.64 - 4.77 µm 3.95 µm 12.9 PLA SA [74]

- 4.13 µm 26.9 PLA Cisplatin [62]

2.3 - 4.4 µm 4.4 µm 7.4 PLGA Celecoxib [65]

Above 5

µm

5 - 5.31 µm 5 µm 15.0 PLA/PLGA

(30/70) Cisplatin [62]

5.4 - 5.7 µm 5.4 µm 16.7 PLGA IGF-1 [29]

6.51 - 12.8 µm 7.5 µm 6.8 PCL Paclitaxel [59]

7.9 - 10.4 µm 8 and 10 µm 15 PLGA SA [77]

5.67 - 9.78 µm 9.39 µm 3 TPP-

Chitosan SA [72]

- 11.4 µm 7.9 PCL Paclitaxel [55]

- 11.76 µm 23.7 PLGA Paclitaxel [53]

- 15 µm 11.3 PLGA Paclitaxel [54]

14.2 - 15.2 µm 15.1 µm 4.6 PLGA Paclitaxel [55]

- 20 µm 15.5 PCL SA [71]

20.3 - 22.1 µm 21.2 µm 7.1 PLGA SA [71]

Monodispersity remains a very important factor for micron sizes, especially for

release properties and Table 2.1 shows a non-exhaustive summary of the various

polydispersities (expressed as the relative standard deviation to the average size of

one formulation) that have been achieved so far by loaded electrosprayed polymeric

particles, per size domain. It can be seen that several similar sizes can provide very

different polydispersivities according to the processing parameters and also depend

case by case on the polymer/drug/solvent selection. It is thus possible to obtain very


- 39 -

low polydispersivity with electrospraying (as low as 2.1% [64, 71]), by tailoring

electrospraying parameters, but this is a complex undertaking. Monodispersity is

obtained with reduced flow rates, increased polymer concentrations (higher

viscosities), reduced conductivity of the electrosprayed solutions and reduced applied

voltages [68, 111]. This applies for the stable-cone jet mode, known to be the only

one able to produce monodisperse particles, and when parameters such as flow rate

are used in the central cone-jet region. Indeed, when flow rate is used close to the

upper and lower limit of the cone-jet region, the polydispersivity index of particles

increases, as seen in Figure 2.5D [68].

The size of electrosprayed polymer particles is greatly influenced by flow rate and

polymer concentration, where increased particle sizes are most significantly obtained

with an increase in flow rate and polymer concentration and a decrease in

conductivity (Figure 2.5C) [83]. However, at increasing flow rates, the size

distribution also becomes broader [63] and formation of secondary droplets can take

place. This leads to a bimodal size distribution which is quite common in

electrospraying and sometimes unavoidable [28, 52, 59, 83]. Some strategies have

been suggested to separate the two size populations by using a steel plate with a 3 cm

circular hole as the grounded electrode, which serves to collect only the primary

droplet population. Often a spatial separation occurs after the droplet break-up where

two regions of electrospray can be seen during the stable cone-jet mode, since

secondary droplets have a larger surface charge density but less mass than primary

droplets. Therefore primary droplets can be found in the inner core of the

electrospraying cone, while secondary droplets get ejected at the periphery of the

cone [52]. By using a plate with a circular hole as a screen on top of the collector,

secondary droplets thus are left behind and only primary droplets are recovered

ensuring monodispersity, although reduced yield may be of concern. Hartman et al.

measured very small currents (31-57 nA) during electrospraying and found that

another way to reduce the frequency of secondary droplets is to lower the current by

lowering the applied voltage or flow rate [42]. However secondary droplets remain

difficult to eliminate completely [52].

These findings are explained by the relationship between size and electrospraying

variables depicted by Hartman et al. in the stable single cone-jet mode and shown in

Equations 2.6 and 2.7 [42]. The droplet diameter, d, can be modeled using various

equations generated by De La Mora and Loscertales [112] in 1994, Gañan-Calvo et


- 40 -

al. in 1997 [105] and Hartmann in 2000 [42] (Equations 2.6 and 2.7) for the single

cone-jet mode. They are functions of the liquid flow rate, Q, the solution density, ρ,

the current, I, the surface tension in ambient air, γ, and the liquid conductivity, K (α

is a constant):

(

)

(2.6)

( )

(2.7)

Particle size is directly proportional to droplet size, where an increase in particle

size is obtained with increasing flow rate and decreasing surface tension, and is

shown to correspond to an increase in polymer content [110]. However, as stated

earlier, an increase in flow rate is also responsible for broader size distributions [63],

thus a compromise needs to be made between particle size and dispersity. This is

explained in section 2.5.1.2.3, where higher flow rates lead to smaller φRay,

eventually falling in the case where φov < φRay < φent where polymer entanglements

are not strong enough to preserve the droplet integrity during electrospraying,

leading to the ejection of offspring droplets (which are around one fifth of the

primary droplets in size [100]). The first pre-requisite of monodispersity of

electrosprayed particles is thus the use of a flow rate that ensures φRay > φent, where

primary droplets cannot be disrupted by Coulomb fission. As stated by Almeria et

al., if increasing particle size is highly desired, while maintaining monodispersity,

increased flow rates could be coupled with higher polymer concentrations so that the

φRay > φent is still validated [49]. However, the use of a higher flow rate may still lead

to the formation of secondary droplets (one fifth to half of the primary droplets in

size [100]) due to increased jet perturbation, in turn providing a bimodal size

distribution.

When involving the polymer concentration parameter, particle size has been

shown to increase compared to the theoretical calculations for PLGA microparticles

containing paclitaxel made from low polymer concentrations (4, 6 and 8% in

acetonitrile), while it was in good agreement for 10%. This was attributed to the non-

spherical shape and high porosity of particles made from lower concentrations [56].

The increase in particle size as a function of the square root of flow rate was also

shown to be sharper for higher polymer concentrations. An increase in concentration


- 41 -

from 5% to 10% PCL in DCM, however, led only to a slight increase in size, for

paclitaxel-loaded particles from around 9 to 13 µm [59], suggesting that flow rate is

more determinant than polymer concentration for directing the size of loaded

particles. Indeed, increasing the flow rate from 4 to 8 mL/h when electrospraying

chitosan solutions significantly increased the size of microparticles at each chitosan

concentration [72]. However the size did not increase significantly with an increase

in concentration from 1 to 2%. In a similar study, an increase in polymer

concentration from 4 to 8% (PLGA 50:50 in acetonitrile) resulted in a limited change

in the particle size while for 10%, the sizes were considerably larger. This was

explained by the low diffusion rate of PLGA chains where a shell of solid PLGA

would form on the surface of the droplets. For lower polymer concentrations in the 4-

8% range, the shell would be thinner but a similar overall size would be obtained,

while for concentrations higher than 8% a high polymer concentration was

established on the surface of the droplet with less solvent evaporation, resulting in a

larger final particle size [56]. All these results underline the strong effect and inter-

dependence of flow rate coupled with concentration on particle size.

Drug loading was also shown to affect microparticle size; loading up to 15.8% of

the anti-cancer drug paclitaxel increased PLGA particle size from around 13 to 15

µm although the number of particles analysed (n) or the technique for size

measurement was not described in the study [55]. The link between loading and size

is however less evident than previous variables. In the case of electrospraying of

emulsions, the size was shown to first decrease and then increase, for a serum

albumin emulsion, when the SA:PLA weight ratio decreased from 1:2 to 1:6. This

was tentatively explained by a decrease in viscosity from less solid mass in the

emulsion causing the initial decrease in size, and the lower conductivity causing the

subsequent increase [73].

As explained in section 2.5.1.2.2, the electrical conductivity is indeed a potent

parameter for controlling particle size where the scaling laws from Gañan-Calvo

showed that a decrease in particle size can be obtained with an increase in

conductivity, according to Equation 2.8 [105]:

(

)

⁄

(2.8)


- 42 -

The use of a solvent with a higher conductivity can thus decrease particle size as

seen when comparing PLGA particles made using acetonitrile and dichloromethane

as the solvents [108] or when adding acetone to a PLA solution of 1,2-dichloroethane

(1:1) [74] where conductivity was increased from 0.058 to 0.412 µS/cm and particles

decreased in size from 4.8 to 1.6 µm. However, it was shown that the use of co-

solvent with increased conductivity broadened the size distribution with a bimodal

character and reduced the spherical morphology of PLGA particles [100] and PLA

particles [74], respectively. The use of organic salts is more effective to increase

conductivity without causing a concomitant deterioration in the initial morphology of

particles as seen with 2 mM of DTAB added to an acetonitrile solution of PLGA and

paclitaxel [56]. In this case, particle size was decreased from 1.2 µm to 355 nm.

Other possible electrolytes include ammonium hydroxide (0.02 to 0.2% (v/v)) [64].

Although conductivity is pivotal in size tailoring, it must be kept in mind that an

increase in conductivity reduces the region of the stable cone-jet mode and hence

standard deviation tends to increase, broadening the size distribution [63]. Again this

is due to a consequent decrease in φRay which may eventually be smaller than φent

where the ejection of offspring, secondary and satellite droplets from the primary

droplets is possible. Such broadening was also observed with increasing surfactant

concentration, thus conductivity of the solution, as seen for 2-16% Pluronic® F-127

in PLGA solution in acetonitrile, although it did not appreciably reduce the average

particle size [56]. However, for budesonide-loaded PCL particles, a decrease of

Tween 20® from 0.005 to 0.001% led to a decrease from 884 to 116.1 nm under

optimal electrospraying conditions [66].

Emulsions comprising organic/aqueous and protein/polymer phases also have

significant impact on particle size [73]. Particle size increased with organic/aqueous

volume phase ratio. This was due to a corresponding increase in viscosity and

decrease in electrical conductivity which makes it more difficult for the solution to

be broken up into smaller droplets in the course of electrospraying, thus increasing

particle size. Such correlation between size, viscosity and electrical conductivity was

also seen for β-oestradiol-loaded PCL particles where an increase of PCL

concentration from 2 to 10 wt% led to a change in viscosity and electrical

conductivity of the PCL solutions from 2.6 to 11 mPa.s and from 3.4 to 0.8 µS/m,

respectively. This resulted in a mean particle size increase from 0.3 to 4.5 µm [68].

Section 2.6 Electrospraying and Drug Release Characteristics

- 43 -

2.6 ELECTROSPRAYING AND DRUG RELEASE CHARACTERISTICS

2.6.1 Choice of Molecules

Most pharmaceutical drugs and proteins are expensive. For this reason the majority

of drug delivery studies are first undertaken with model drugs or model proteins, to

enable optimisation of parameters and characteristics of particles in the first instance,

before loading fragile and expensive drugs/proteins. A non-exhaustive summary of

various drugs and proteins that have been loaded so far in electrosprayed particles is

presented in Figure 2.6.

As far as proteins are concerned, serum albumin has been widely used for this

‘model’ purpose in traditional encapsulation techniques [113], and to some extent in

electrospraying [71-74, 77, 79, 92, 114]. SA is readily available and it offers high

stability and low cost, which is advantageous in the early stages of optimisation. The

molecular mass of SA is 66.4 kg/mol, which is similar in size to some growth factors

(GFs) used for tissue regeneration, providing a more suitable choice than smaller

model molecules. Serum albumins are also extensively used as an excipient, i.e. as an

inactive substance used as carrier for the molecules of interest, since they have the

ability to bind a wide variety of biological molecules, e.g. cationic, anionic,

hydrophilic, hydrophobic substances. Many drugs, such as anti-coagulants and

anesthetics, are also transported in blood while bound to albumin [104]. Serum

albumins have no adverse effect in most biochemical reactions and they have been

shown to assist decreasing the initial burst release occurring in most particulate

systems [115], and stabilise and protect the bioactivity of molecules during the harsh

conditions of encapsulation [116].


- 44 -

Figure 2.6. Structures of some drugs and proteins which have been encapsulated in electrosprayed

particles [117].

Due to its combined advantages, serum albumin has often been selected as a

model protein to be encapsulated in electrosprayed particles, in studies focusing on

release of proteins or drugs. Unfortunately, only a handful of studies can be found

where electrosprayed particles have been loaded with the actual protein of interest

other than SA, in part due to their high cost. It would be valuable for more studies to

progress towards using therapeutic proteins in place of these model systems. Some

examples of studies which do encapsulate therapeutic molecules include growth

factors such as insulin-like growth factor-1 (IGF-1) [29], vascular endothelial growth

factor (VEGF) and platelet-derived growth factor (PDGF) [30]. Growth factors are

essential actors during natural tissue formation. These polypeptides are produced in-

situ by cells and transmit signals to modulate cellular activities [6, 11]. During tissue

growth a complex and orchestrated delivery of several types of GFs occurs and tissue

Anti-cancer drugs

PaclitaxelDoxorubicin

Suramin

Inhalation drugs

Methylparahydroxybenzoate (model drug)

Beclomethasone dipropionate

Salbutamol

Rifampicin

Antibiotics

Ampicillin

Proteins

Bovine serum albumin (model protein)

Insulin-like growth factor-1

Vascular endothelial growth factor

Platelet-derived growth factor-BB


- 45 -

growth is dependent on this delivery. Thanks to the current technologies, GFs can

now be recombinantly produced, albeit at very high cost, and have thus attracted a lot

of interest among tissue engineers. Many drug delivery particulate systems (DDPS)

have attempted to encapsulate and release GFs in a sustained manner, mimicking the

normal in vivo production. For GFs that were encapsulated in electrosprayed

particles, we find IGF-1, PDGF and VEGF, which are mostly involved in

angiogenesis. It must be noted that their molecular mass ranges from around 7 (IGF-

1) to 45 kg/mol (VEGF) and are therefore lower than SA (66.4 kg/mol). Compared to

small drugs, proteins are also prone to denaturation, which is often an issue in DDPS

production where organic solvents are used. The technique of electrospraying

however, offers a reduced contact of proteins with solvents and most importantly

does not require the emulsion step present in the traditional fabrication processes.

The aqueous/organic interface is thus avoided along with its respective shear

stresses, mainly responsible for protein denaturation [35]. Electrospraying may thus

prove to be superior to traditional techniques for loading of proteins, sensitive to

denaturation. This is further discussed in sections 2.6.5 and 2.6.6.

As far as small molecules are concerned, various drugs have been loaded in

electrosprayed particles, finding applications in the fields of inhalation therapies,

antibiotic delivery, cancer treatments and hormonal treatments. In inhalation

therapies, the control and monodispersity of particle sizes obtained with

electrospraying allow for more efficient administration of drugs by a reduction of the

required dose and higher drug availability for treatment. The major drugs that have

been utilised to date by direct electrospraying of the drug solutions or by

encapsulation in PLA particles, are beclomethasone dipropionate (BDP) [63, 64] and

salbutamol-sulfate (SS) [64] (commercially known as ventolin) and are delivered

through bronchodilatators. These small molecules (both less than 1 kg/mol) are used

in the treatment of asthma and other chronic obstructive lung diseases and need to be

inhaled for direct effect on bronchial smooth muscle [63]. BDP and SS have very

different properties; for one they are hydrophobic and hydrophilic, respectively. For

this reason electrospraying represents a superior alternative to traditional techniques

since it does not require for the experimental parameters or setup to be changed,

being renowned as a method suited to molecules that do not process well (such as

those with different solubilities) [64]. Methylparahydroxybenzoate (MPHB) is a

model drug that can be used for mimicking BDP and it has been demonstrated that


- 46 -

the electrospraying of both drug solutions were similar, validating its use as a model

molecule [63].

Antibiotics have also been encapsulated in electrosprayed particles, such as

rifampicin, an anti-tuberculosis drug, [52] and ampicillin [51] which function to treat

bacterial infections (Figure 2.6). Similarly to inhalation drugs, ampicillin and

rifampicin have a small molecular mass of 350 g/mol and 823 g/mol, respectively, an

important contrast with the mass of polymer chains used in electrospraying, ranging

up to hundreds of kg/mol.

Anti-cancer drugs remain the most frequently tested drugs in electrospraying.

Encaspulated anti-cancer drugs include cisplatin [62], paclitaxel (sold commercially

as Taxol®) [53-60, 87], a hydrophobic molecule, suramin [58, 87], and doxorubicin

[50, 61, 95], both hydrophilic and shown in Figure 2.6. In cancer therapies, multiple

and temporal drug delivery is generally required for treatment. However, previous

methods to obtain double-walled microspheres such as the oil-in-oil-in-water (o/o/w)

emulsion require several hours of rapid stirring to create an emulsion, which is

detrimental to the drug, limiting loading, with possible degradation issues and

difficulties in controlling the drug distribution [58]. Coaxial electrospraying is thus

advantageous in this instance since it allows: encapsulation of different types of

drugs in different compartments in one single step; encapsulation of both

hydrophobic and hydrophilic drugs; and tailoring of release (sequential or coupled)

with the tailoring of electrospraying parameters and physical disposition of drugs

within the core and shell [58, 87].

Less commonly in the field of hormonal treatments, sex hormones or drugs have

also been encapsulated in electrosprayed particles. β-oestradiol, a contraceptive and

hypocholesteraemic drug of low molecular weight (272 g/mol) was for instance

encapsulated in PCL particles [68], while tamoxifen (371.5 g/mol), a drug that blocks

the effects of oestrogen in breast tissue was encapsulated in lipid-based particles [69,

70].

Miscellaneous drugs that have been encapsulated in electrosprayed particles

include α-lipoic acid, an agent shown to be effective in treating various diseases

(diabetes, atherogenesis) [81]. Anti-inflammatory drugs, such as celecoxib,

budesonide and naproxen have also been encapsulated in chitosan, PCL, and

polyvinylpyrrolidone particles, respectively [65-67]. Celecoxib is widely used in the

treatment of osteoarthritis but has undesirable properties such as high cohesiveness


- 47 -

and low solubility. The use of electrospraying for encapsulating celecoxib in PLGA

microparticles allowed an increase in celecoxib dissolution rate, which is desired to

improve oral bioavailability [65].

2.6.2 Loading and Encapsulation

2.6.2.1 Definitions and Methods

Electrospraying is an encapsulation process in which efficiency is measured by the

traditional encapsulation efficiency (EE) and loading capacity (LC) parameters

commonly used in the field. Encapsulation efficiency represents the weight of

biomolecules effectively loaded in particles (wLoaded) with respect to the initial weight

of biomolecules available (wTotal) (Equation 2.9). Loading capacity is the weight of

biomolecules effectively loaded in particles (wLoaded) as a fraction of the total weight

of particles (wParticles) (Equation 2.10):

(2.9)

(2.10)

Extraction is the most commonly used process to determine these parameters.

Briefly, particles are dissolved in an organic solvent, usually identical to that used to

initially solubilise the polymer, followed by the addition of an aqueous solution. The

mixture is vortexed to extract the encapsulated biomolecule to the aqueous phase,

eventually followed by centrifugation to separate the oil and water phases. The

aqueous phase is then collected and analysed. In some cases, organic solvent is left to

evaporate before addition of the aqueous phase [55, 57, 59]. Since most studies

encapsulate SA, the standard assay for concentration determination is the micro-

bicinchoninic acid (µBCA) protein assay [59, 71, 77], and sometimes the Bradford

assay [78]. When small molecules containing chromophores or large quantities of

protein are loaded, high-performance liquid chromatography (HPLC) [53, 55, 58, 59,

87] and ultraviolet (UV) spectrophotometer [64] have been used. In all cases,

calibration curves (produced by serial dilutions of the biomolecule in question) allow

the quantification of encapsulated contents.

From the literature, it appears that most studies only undertake one extraction,

with the exception of Nie et al.’s study where a total of three extraction cycles were

performed for suramin recovery [58]. Doing only one extraction is quite restrictive


- 48 -

considering that E(%)=100D/(1+D), where E is the amount extracted and D is the

distribution coefficient. SA is likely to have a relatively high distribution coefficient

into aqueous solutions and may give high extraction values with only one cycle,

however for biomolecules that have a lower distribution coefficient, two or three

extraction cycles are necessary to maximise recovery. In general terms, two or three

smaller extractions are always more efficient than one large one [118].

The choice of solvent for extraction is paramount to success. DCM is widely used

for general extractions and extractions of loaded electrosprayed particles. DCM is an

excellent choice for extraction: it is immiscible with water and is more dense and

volatile, allowing an easy removal by evaporation if required. Its drawback is that

being a chlorinated solvent, like chloroform, DCM has a greater tendency to form

emulsions than non-chlorinated solvents [118]. This might be an issue for full

recovery of biomolecules.

A final important consideration with the extraction process is that it does not

represent the amount of biomolecules effectively encapsulated/loaded in particles but

comprises also non-encapsulated molecules which may be simply adsorbed on the

surface, (which are responsible for the initial burst release often seen with such

systems). This is an issue since very high EE are reported in the literature but there is

rarely sufficient description of quantification of adsorbed/encapsulated molecules.

Some quantification was attempted in a study from Ding et al., where the particles

received an ultrasonic treatment after dispersion in water, followed by freeze-drying.

The EE of particles from this batch was reduced over 18%, corresponding to the loss

of adsorbed molecules on the surface of particles, readily dissolved in water during

the treatment [59]. Some EE/LC numbers are therefore to be considered with caution

if measured by the extraction method, as they are not representative of the real

amount of encapsulated/loaded molecules; this is a real shortfall in most studies.

Interestingly, the determination of EE and LC is also presented via a ‘non-

entrapped’ method proposed by Xu et al. [72, 73]. Particles were centrifuged at

20,000 g at 15°C for 30 min and the amount of free molecules (SA in this case) was

determined in clear supernatant by UV spectrophotometry at 280 nm using the

supernatant of non-loaded particles as a basic correction. LC and EE were calculated

according to Equations 2.11 and 2.12:

(2.11)


- 49 -

(2.12)

where A is the total amount of SA, B is the free amount of SA and C is the particles

weight. A variant of this technique was used by Arya et al. (10,000 g at 12°C for 10

min) [51] and Enayati et al. (β-oestradiol, 4,300 rpm at room T for 45 min) [68]. This

method does not necessarily take into account losses during particle preparation.

2.6.2.2 Influence of Parameters on Loading and Encapsulation

In traditional encapsulation processes, EE and LC are typically affected by the

processing parameters, including particle formation temperature. For example EE in

double emulsion procedures is dependent on the balance between solvent

evaporation rate and immiscibility between water and particle, rate of polymer

precipitation and thus hardening rate of the sphere wall [18]. In electrospraying,

similar variables such as the nature of the polymer, protein/polymer weight ratio,

along with flow rates for instance, are parameters influencing EE and LC, and will be

presented hereafter.

2.6.2.2.1 Loading Capacities

High loading capacities are always desirable for an increased availability of the

therapeutic molecule in targeted areas with minimal use of carrier materials.

Nevertheless, this can induce possible changes of particle morphology that occur

with increased loading, and thus possible alteration of release profiles. In a study

from Hong et al., it was proven that an increase in the loading capacity led to a loss

of sphericity of microparticles [52] which in turn affected the release profiles. The

scaling laws of electrospraying were nevertheless verified with almost no theoretical

variation from non-loaded to loaded particles. The authors stipulated that the

shrinkage and drying processes were responsible for such variation of morphology.

Actual loading capacities in electrospraying have also been shown to be slightly

decreased compared to theoretical loadings. For instance a loss of 20% was seen

when loading paclitaxel in PCL and PLGA microspheres [55]. Loadings also affect

burst release with a higher loading leading to a higher burst release. This was

observed for PLGA microparticles when a loading from 10 to 20% of paclitaxel

almost doubled the burst release for 15 µm-particles [54], and a loading from 10 to

50% of celecoxib increased the burst release from 39 to 54% [65] for 2-4 µm-

particles.


- 50 -

In terms of parameters affecting the loading capacity, an increase in

protein/polymer weight ratio dramatically decreased the loading capacity in the case

of emulsions of SA in PLA solution [73], while it increased for solid dispersion of

SA in chitosan solution [72]. In the case of coaxial electrospraying, drug loading

could be increased by increasing inner protein aqueous solution flow rate or

increasing inner protein concentration [77].

2.6.2.2.2 Encapsulation Efficiencies

Electrospraying is known as a technique which can give high encapsulation

efficiencies (EE), and indeed has been shown to reach 100% EE for doxorubicin and

rhodamine-loaded PLGA particles [50]. Electrospraying also presents the great

benefit that encapsulation of both types of drugs, hydrophilic and hydrophobic, are

efficiently obtained compared to traditional methods [50, 64], mainly since there is

no need of an emulsion step. In emulsion-based methods, the presence of both

aqueous and organic phases may indeed lead to preferential diffusion of the drug to

one phase or the other according to their hydrophilicity/hydrophobicity

characteristics and thus reducing final EE. This is avoided with electrospraying

where emulsions are not required.

EE depends case by case on the combination of drug/solvent/polymer selection,

where the hydrophilic nature of these components plays an important role. This may

be illustrated by considering the nature of the polymer itself, for example the

encapsulation of SA in a more hydrophobic polymer such as PCL has led to 28% EE

compared to 40% for PLGA microparticles electrosprayed in the same conditions

[71]. The use of hydrophilic additives should also be considered and it has been

shown that Pluronic® F-127, a highly hydrophilic copolymer used as a surfactant,

increased encapsulations efficiencies from 53.4% to 76.7% by using 5% and 10% of

Pluronic respectively in PLGA microparticles loaded with SA, by enhancing the w/o

emulsion stability. It was further stated that the use of a probe sonication to form the

emulsion could enhance the EE [71].

An increase in loading generally leads to a decrease in EE, as seen in

electrosprayed PLGA films loaded with paclitaxel, where a 5, 10, 15 and 30%

loading respectively led to 80, 71.9, 66.4 and 63% EE [57]. This was attributed to the

partition coefficient, referring to the equilibrium solubility of the drug in the polymer

against the equilibrium solubility of the drug in the solvent and responsible for


- 51 -

diffusion of the drug into the polymer phase. With paclitaxel being more

hydrophobic and soluble in organic solvents, increasing the loading could have led to

preferential diffusion into the solvent, and thus reduced encapsulation [57]. In a

study from Xie et al., for similar polymer solutions and spraying conditions, an

increase of paclitaxel loading from 8% to 16% slightly decreased EE from 82% to

78% [55] in PLGA microparticles. This was also shown for PLA microparticles

encapsulating SA where EE decreased with increasing SA/PLA ratio and increased

with organic/aqueous phase ratio. This latter may be explained by an increase in

viscosity when increasing the organic/aqueous phase ratio, leading to better

encapsulation [73]. Furthermore, a strong correlation was found between SA/PLA

weight ratio and organic/aqueous phase ratio with respect to encapsulation

efficiencies. Size affects EEs as well, where smaller particle sizes lower the EE [64].

In the case of coaxial electrospraying, higher EE in both core and shell can be

obtained by encapsulating hydrophobic drugs in the core and hydrophilic drugs in the

shell [58]. In monoaxial electrospraying, similar EE were found for hydrophobic and

hydrophilic drugs (54% BDP and 56% SS, respectively) in PLA nanoparticles,

demonstrating the versatility of the technique. Loss of drugs can, however, be an

issue in electrospraying, mainly caused by spreading of the particles to the receiving

vessel walls and other manufacturing equipment. This was measured as 20% for SS-

loaded PLA nanoparticles sprayed in ethanol. When electrospraying in a cross-

linking solution however, an increase in the concentration of the cross-linking agent

(10% against 5%) led to a significant increase in EE of SA [72]. This was due to an

increased intermolecular interaction of the polymer and cross-linking agent when

increasing their respective concentrations, inhibiting the loss of SA into the

collection solution and improving EE. When gelation was incomplete, as seen for

higher flow rates, more SA was lost in the collection solution [72].

2.6.3 Molecule Dispersion

Controlled dispersion of the drug within the polymer matrix is of upmost importance

for consistent release. It was previously stated that drug concentration in a particle

matrix tends to decline as we move outwards from the centre with the increase of the

particle diameter as seen in rifampicin-loaded particles [52]. This is explained by the

diffusion mechanism of solutes, stating that an increase in the droplet size provides a


- 52 -

longer distance and time for diffusion of solutes, leading to a drug gradient within the

particle.

An electrosprayed droplet of polymer contains macromolecules which move and

diffuse during solvent evaporation, providing the final polymer network. The level of

intermolecular entanglement among macromolecules dictates these parameters,

affecting the diffusion rate of macromolecules towards the centre. When adding

small molecules like drugs to this system, they generally diffuse easily towards the

droplet centre due to the absence of intermolecular action. However the

intermolecular entanglement of polymer macromolecules is weakened, leading to a

decrease in the diffusion coefficient of solutes expressed by the Stokes-Einstein

Equation 2.13:

(2.13)

where kB is the Boltzman’s constant, η the viscosity of solvent, T the temperature,

and RH the hydrodynamic radius of solutes [52]. Therefore by increasing the

concentration of small solutes in a polymer droplet composed of big

macromolecules, the diffusion coefficient decreases. Above a critical concentration

value, the diffusion of solutes becomes slower than solvent evaporation and the small

molecules are trapped on the surface of the droplet, leading to a molecule saturated

layer of semi-solidified skin [52, 60]. Such a configuration is not ideal for the

physical and release properties of particles, since with further evaporation of the

droplet, the skin moves towards the droplet centre, leading to particle collapse and a

final wrinkled morphology, which does not lend itself to sustained and reproducible

release properties. This critical concentration value was found to be 30% wt/v in the

case of rifampicin in a PLGA/chloroform system and led to the loss of particle

sphericity and subsequent increased burst release compared to the sustained release

from spherical particles obtained with 10% wt/v loading of rifampicin [52].

Ideal and homogeneous molecule dispersion, which is preferable for sustained

release, is therefore obtained for low loadings, smaller particle size (which limits the

drug gradient effect) and a good balance between the diffusion and evaporation

mechanisms. If high loadings are needed, it is important to control diffusion and

ensure that it does not become lower than evaporation. This can be balanced by using

a slow evaporating solvent (such as DMF).


- 53 -

A few techniques have been used to assess the integration of drugs within the

polymer matrix of electrosprayed particles, although they do not allow for physical

visualisation of drugs within the droplets (although confocal laser scanning

microscopy may be used to look at fluorescently-labelled drugs). Differential

scanning calorimetry (DSC) relies on the fact that if a drug is well dispersed in the

polymer matrix, the melting transition of the drug will be suppressed either partially

or completely [59]. This theory was used for Taxol®-loaded PCL particles where

only a slight heat flow peak of Taxol® was observed in the physical mixture of PCL

and Taxol® [59]. The authors concluded that the drug was well-dispersed within the

matrix. The same theory was used for paclitaxel-loaded PLGA and PCL

microparticles where no peak at all was seen in the 150-250°C temperature range,

while paclitaxel normally has an endothermic peak of melting at 223°C [55]. From

this result, the authors stated that the paclitaxel was in an amorphous or disordered-

crystalline phase of a molecular dispersion or a solid solution state in the polymer

matrix. The same conclusion was made for hydrogel beads encapsulating paclitaxel-

loaded PLGA microspheres [53]. This was further seen in electrosprayed PLGA

films loaded with the same drug [57]. The authors even annealed their samples for 3

days at 60°C to facilitate a higher diffusion rate for dispersed drug molecules, but

still no crystalline peak of paclitaxel was observed for annealed samples, leading to

the conclusion that the drug was in a solid solution state within the matrix, as

compared to a metastable molecular dispersion [57]. A similar theory was used for

BDP and SS-loaded PLA nanoparticles where no melting peak at all was seen for

BDP-loaded particles and only a smaller and broader peak was seen in SS-loaded

particles [64]. The crystallinity of PLA was changed in the presence of both drugs,

namely higher when SS was encapsulated and lower when BDP was encapsulated.

This was tentatively explained by the presence of water, since SS was emulsified

before electrospraying, thus reducing evaporation rate, allowing for more polymer

diffusion and chains re-arrangement, and was thus responsible for the higher value

for crystallinity [64]. The crystallinity of all materials (BDP, SS and PLA) in the

nanoparticle formulations were decreased although the crystalline intensities were

distributed as 80% for BDP and 20% for PLA, or 54% for SS and 46% for PLA.

Importantly, no new peaks were seen in the DSC profiles, indicating no strong

physical or chemical interactions were present between the drugs and polymer. A

similar result was observed for celecoxib-loaded PLGA microparticles, where the


- 54 -

melting endotherms of celecoxib shifted down or disappeared according to decreased

drug content, suggesting that the drug was molecularly dispersed within the PLGA

matrix [65]. The disappearance of the Tamoxifen peak for loaded PAA-Cholesterol

nanoparticles suggested the same result [70].

X-ray diffractometry (XRD) may also be used for determining the physical state

of a drug within polymeric matrices since characteristics of the peaks mark the

degree of crystallisation of the drug with the matrix. XRD previously showed that

paclitaxel was in an amorphous form in the PLGA matrix, even for up to 30% drug

loading, since no peak was seen in the expected range of temperature (200-250°C)

when analysing the polymeric matrix [57].

Analysis of surface chemistry by X-ray photoelectron spectroscopy (XPS) can

also give information regarding the distribution of drugs within microparticles, by

examining the C, N and O element compositions. This technique has been used to

show that paclitaxel was present on the surface layer of PLGA microparticles (with

up to 0.8% atomic mass concentration), a phenomena which is argued to be

responsible for the initial burst release seen in the in vitro release study [55]. XPS

showed that the amount of nitrogen increased with increasing paclitaxel contents (0-

30% loading) in electrosprayed PLGA films, attesting of the presence of the drug.

Confocal laser scanning microscopy (CLSM) is another way of qualitatively

looking at fluorescently labelled biomolecules encapsulated within particles. This

method allows for screening of cross-sections of a loaded particle through the entire

particle, for further 3D reconstruction. This powerful tool proves to be very useful

for visualising in 3D the biomolecule distribution inside the particles after

production, and studying the mechanisms of release from particles. CLSM has been

used successfully for other fabrication methods such as spray-drying [119], however

it has yet to be extensively utilised for visualisation purposes with electrosprayed

particles [50], although it would be very valuable, especially in combination with the

aforementioned analysis techniques for a more thorough characterisation of molecule

dispersion.

2.6.4 Release Kinetics

Polymeric microparticles for controlled drug delivery have been extensively studied

in the last 50 years and various reviews detail their preparation, the factors affecting

the release and the current difficulties faced during processes [17, 18, 26, 35, 120].


- 55 -

Most of these reviews encompass microparticles made from traditional fabrication

techniques and limited information is available on release kinetics from

electrosprayed particles.

In general terms, release occurs through two different mechanisms; passive

diffusion and polymer degradation. Ideally a controlled release system would show a

zero-order release profile, meaning a constant release rate over time. However, the

release profile from particles is usually split in two distinct processes:

1. The initial burst release of molecules contained on and in the surface of the

particle due to the leaching occurring at the outer wall of the particle as it becomes

hydrated [18].

2. The slower and more constant release of molecule from the inner part of the

particle.

Release profiles can be affected by physical and chemical factors: the nature of

the polymer (molecular weight, blending, crystallinity), the nature of the loaded

molecule, its distribution and activity, the morphology of microspheres, their

porosity and size distribution [17, 18]. In electrospraying, similar parameters are able

to tailor release kinetics and they are discussed in the next section. Table 2.2

illustrates various electrospraying studies with the release profiles and corresponding

morphologies.


- 56 -

Table 2.2. Drug loading and release characteristics of electrosprayed particles loaded with various

therapeutic molecules. Adapted from [50-52, 55, 59, 65, 68, 71, 73, 77] with permission. 2006, 2008,

2005, 2011 Elsevier Science Ltd. [50, 52, 55, 59, 65, 73, 77]; 2007, 2009 John Wiley and Sons [51,

71]; 2010 Royal Society [68].

Ref. Polymer Molecule LC EE Size (µm)

MorphologyCumulative release profiles (%) (y axis)

[73] PLA SA 74-91% 23-81% 0.8-4

[71] PLGA 50:50 SA - 20-77% 20-22

[77] PLGA 75:25 SA Lysozyme

0.3-0.8% - 0.8-10

[57] PCL Paclitaxel 1-2% 77-98% 7-11

[62] PLGA 50:50PCL

Paclitaxel 8-16% 78-84% 11-15

[50]PVA coated PLGA 50:50

RhodamineDoxorubicin

0.05% 4-100% 0.6-1.3

[51] Chitosan Ampicillin 50% 80% 0.5

[52] PLGA 80:20 Rifampicin 10-30% - 3-7

[65] PLGA 50:50 Celecoxib 10-50% - 2-4

[68] PCL β-oestradiol 15% 85-89% 0.3-5

0

50

100

0 60 120

RhodamineDoxorubicin

Time (h)

0 Time (h) 120

100

50

0

0

100

50

0 25Time (days)

50% LC30% LC10% LC

5 µm

5% PCL (a)7.5% PCL (b)

0 Time (days) 500

40

20

(a)(b)

10 µm

2 µm

5 µm

40 µm

4 µm

0Time (days)

4 8

20

60

100 (a)(b)(c)

(a) 50% LC(b) 25% LC(c) 10% LC

4 µm

0 10 20 30 40Time (days)

(a)(b)(c)

(a) 2% PCL(b) 5% PCL(c) 10% PCL

20

40

0 Time (days) 350

40

80 (a)6%PCL-3mL/h-8%LC(b)8%PLGA-5mL/h-8%LC

(c)6%PLGA-3mL/h-8%LC(d)8%PLGA-5mL/h-16%LC

0 6030 90

v

0

20

40

Time (h)

SA:PLA=1:6 (a)SA:PLA=1:4 (b)SA:PLA=1:2 (c) (a)

(b)

(c)

v50

100

0 40Time (days)0

(a)(b)(c)

6% PLGA, 0% F1276% PLGA, 5% F12710% PLGA,10% F127

(a )(b)(c)

0

40

80

0 Time (days) 30

(a)

(b)

0.28% LC (a)0.50% LC (b)

3 µm

20 µm

20 µm

(a)

(d)(c)(b)

0


- 57 -

Size

The size of particles which encapsulate bioactive molecules is paramount in tailoring

release profiles. A larger surface area to volume ratio (smaller particles) leads to

faster release since particles are more easily penetrated by fluids, favouring easier

diffusion of drugs and faster degradation of the polymer matrix. However, it is

important to emphasise that it is not size itself that controls the release profiles but it

has more to do with the polymer/drug/solvent selection and processing parameters

that are used in each case, as explained in section 2.5.2.2. Therefore size is a result of

other variables which have an inter-dependent effect and need to be appropriately

correlated to truly control release kinetics.

In a study by Enayati et al. for instance, β-oestradiol-loaded PCL particles had

similar release pattern for mean sizes of 0.34, 0.8 and 4.6 µm, however release was

45, 42 and 36%, respectively, after 45 days, thus showing a reduced release for

increasing particle size [68]. This increase in size was actually due to an increase in

polymer content (2, 5 and 10%, respectively), showing that polymer concentration do

indeed provide bigger particle sizes but may also provide reduced release rates at the

same time. In this study, due to the nature of the polymer (PCL) where degradation is

unlikely to have occurred over a seven-week period, the release was due to diffusion

of β-oestradiol from the particles, proving that drugs loaded in smaller particles

comprising less polymeric bulk material are prone to better diffusion outside the

polymeric matrix and thus enhanced release [68]. However, burst release is also

more likely to happen from smaller particles. This was observed for paclitaxel-

loaded PCL particles where two formulations with similar sizes (9.45 and 9.52 µm)

showed a similar release pattern and amount released, while a slight decrease in size

to 8.68 µm gave the same release pattern but with a higher burst at the beginning

(11% compared to 7% burst within a day). This was again attributed to a higher

polymer concentration (5% PCL in DCM for the smaller size and 7.5% for the larger

size) which influenced the final size, where larger particles had a denser polymer

matrix and thus a reduced rate of diffusion, in turn reducing the initial burst release

[59]. When working with smaller concentrations, size is less affected although

release profiles can still show great differences. For instance, for an increase in

chitosan concentration from 1 to 2%, size was slightly higher, but not significantly

(7.48 and 8.11 µm, respectively), while a higher burst was observed for the reduced

concentration (7 and 2% after 4 days) leading to a final cumulative release of 33%


- 58 -

and 21% for the 1% and 2% concentration formulations, respectively. Although an

increase in polymer concentration is generally shown to increase size and decrease

release rates/burst release at high concentrations, the contrary is seen for loadings.

An increase in loading is generally responsible for increased sizes too, but generates

faster release rates and burst, especially on the submicron scale [72]. This can be

seen for doxorubicin-loaded chitosan nanoparticles for example, where an increase in

loading from 0.25 to 1% doxorubicin increased the resultant particle size from 527.3

to 873 nm and led to 40% burst release within 3 hours while only a 20% burst release

was observed for the smaller sizes [95]. For microparticles, however, size is less

affected by loading although release is affected. For instance, in paclitaxel-loaded

PCL particles, when loading was increased from 7.9 to 15.8%, the resultant size was

very similar (15.2 and 15.2 µm, respectively) although burst release was 10 and 20%

after 1 day, and the final cumulative release reached 57 and 62%, respectively [55].

In the case of emulsions, the organic/aqueous phase factor has little effect on

release profiles although size is significantly affected, as a consequence from

decreased conductivity with increased organic phase, and thus increased size, as

explained in section 2.5.2.2. This was seen for SA-loaded PLA particles where at a

constant ratio of 1:4 SA/PLA, but increased organic/aqueous phase ratio ranging

from 6.7:1 to 20:1, size increased from 0.8 to 1.9 µm but showed similar release

pattern and amount released [73].

Another significant factor affecting release kinetics is the degree of agglomeration

of the particles. The burst release process is mainly diffusion-driven while the second

process providing a slower release is erosion-driven. It was shown by Almería et al.

that the burst release stage was greatly affected by particle agglomeration and

particle size, whereas the slower release part was much less dependent on particle

size [50]. Agglomeration however was shown to affects release kinetics for

hydrophilic biomolecules, since sizes of particle clusters result in orders of

magnitude larger than individual particles. This aggregation compromises the

reproducibility of release profiles and provides less cumulative release than dispersed

particles (Rhodamine B from PLGA electrosprayed particles) [50]. Coating

techniques may thus be used for preventing aggregation when electrospraying in

solution and enables tight control over particle size.


- 59 -

Morphology

Along with size distribution, morphology is another major contributor for controlling

drug release behaviour and like size, morphology is directed by the

polymer/drug/solvent selection and processing parameters [59]. It was indeed shown

that wrinkled particles led to a burst release of 50% of cumulative release of SA form

PLGA particles in the first day, which was not seen for spherical particles with the

same size distribution (21 ± 2 µm average diameter) [71]. This is a direct

consequence of lower polymer concentration used in wrinkled particles (6%)

compared to dense and spherical particles made of 10% PLGA. More pores were

found in wrinkled particles allowing for molecule adsorption instead of

encapsulation. Water penetration is more accessible in porous particles and leads to

the rapid diffusion of adsorbed molecules, responsible for the high burst release. In

denser particles, the rate of water penetration is reduced, allowing for desirable zero-

order release kinetics.

Similarly to polymer concentration, molecular weight is another important factor

for tailoring particles and their release profiles, since both parameters direct the

viscosity of solutions. This was illustrated with PLGA capsules containing IGF-1,

made of low (5-15 kDa) and high MW (40-75 kDa) [29]. The release profiles were

similar, triphasic in nature, but with an initial burst which was more prevalent for the

low MW formulation, for same PLGA concentration and IGF-1 loading. The burst

was 5.5% compared to 7% and led to a final cumulative release of 10 and 12% for

high and low MW formulations, respectively. The morphology of high MW particles

was spherical while the low MW particles displayed an irregular morphology. This

was a consequence of weaker chain interactions in the low MW PLGA where

packing of polymer chains was looser than that of high MW PLGA, which allowed

the encapsulated IGF-1 to diffuse through the polymer more easily [29].

The solvent is another means to control particle morphology, due to different

evaporation rates that lead to more or less porous structures respectively. This will

ultimately condition the release kinetics as well, as seen with PLGA particles

containing doxorubicin, electrosprayed from a 2,2,2-trifluoroethanol (TFE) solution

and TFE-dimethyl sulfoxide (DMSO) mixture (vapour pressures at room temperature

of TFE and DMSO are 0.08 kPa and 10.09 kPa respectively). As expected, PLGA

particles electrosprayed from TFE were more porous than the ones from TFE-


- 60 -

DMSO, leading to 77% compared to 52% of drug released by diffusion (burst),

respectively [50].

Nature of Polymer

It is accepted that degradation of polymeric particles initially occurs in amorphous

regions, followed by a slower degradation of the crystalline regions of particles [18].

Freiberg et al. stated that low crystallinity allows better drug dispersion and

increased drug-polymer interactions while the degree of crystallinity is also

influenced by the drug loading and the concentration and removal rate of organic

solvent [18]. Therefore, the use of polymers with highly crystalline structures such as

PCL enables the production of microparticles with uniform and reproducible

physical characteristics [83], but might be inadequate for optimal drug dispersion and

release characteristics. For instance, in a study from Ding et al., the cumulative

release of Taxol® from electrosprayed PCL particles (65k) did not exceed 37% of the

total amount of encapsulated Taxol® after 10 days of release (tested up to 50 days),

suggesting that a high percentage of drug aggregated after contact with the polymer

[59]. In a similar fashion, a study by Xie et al. showed that PCL microparticles

loaded with 8.1% paclitaxel were able to release only 32% of this load after 30 days

of in vitro incubation while PLGA microparticles loaded with the same amounts and

electrosprayed with the same conditions reached 60% of cumulative release [55].

Regions of high crystallinity and aggregated protein may likely contribute to the

incomplete release of the protein [91].

Responsive polymers such as elastin-like polypeptides can also release

biomolecules, such as doxorubicin, according to pH variation. However, in all cases

(pH; 2.5, 5.5 and 7.5), all systems suffered from burst release were maximum release

was achieved after 15 min only, and therefore they did not provide sustained release

[61].

Nature of Drug

Interactions between loaded molecules and polymers direct the location of molecules

within the polymer matrix (either encapsulated in the core or adsorbed on the surface

of the particle) and affect the kinetics of release [64]. In the case of coaxial

electrospraying for loading of multiple drugs within microcapsules, it was shown that

the nature of the drugs and their location within the microcapsules affected the

release patterns; loading of hydrophilic drug in the shell and hydrophobic drug in the


- 61 -

core provided a sequential release, while the opposite led to the drugs being released

in parallel [58].

The physicochemical affinity of the drug with the polymer system has a great

influence on release kinetics. For similar size distributions and loadings, dramatic

differences can be observed when varying the hydrophilicity/hydrophobicity of the

drug encapsulated. For instance when rhodamine B (RHOB) and rhodamine B

octadecyl ester perchlorate (RHOBOEP) were used as hydrophilic and hydrophobic

drug surrogates respectively, 98% of RHOB was released within 1 day, while only

6% of total RHOBOEP was released after 5 days. This was explained by the strong

affinity of RHOBOEP with PLGA, preventing any initial burst release (RHOB and

RHOBOEP have octanol/water partition coefficients such that log KOW = 1.48 and 8-9

respectively). When compared with doxorubicin (DOX) (log KOW = 1.85), an

intermediate behaviour was observed where 60% of the drug was released in the first

24h and further 20% was released after 5 days. This was attributed to the different

partition coefficients of the two substances within PLGA [50]. The release of RHOB

will occur rapidly by diffusion of molecules inside the polymer matrix while DOX -

having a greater partition coefficient within PLGA - remains entrapped longer in the

hydrophobic porous regions of the matrix.

Additives

The use of additives can greatly affect the release kinetics, such as the use of

poly(ethylene glycol) (PEG), commonly used in traditional encapsulation processes.

Due to its hydrophilicity, PEG increases the degradation rate of the main polymer

matrix by rendering the overall polymer network more hydrophilic, increasing

swelling and thus accelerating release [121].

In co-axial electrospinning, for instance, an aqueous PDGF solution was

encapsulated in a blend of PCL:PEG nanofibres [122]. PEG acted as a porogen and

PDGF release reached 100% in 35 days with a relatively linear release profile, while

less than 1% of PDGF was released from the PCL nanofibres with no PEG in the

shell. The rate of protein release was shown to be controlled by the molecular weight

and concentration of PEG [122]. Johnson et al. showed that the amount of PEG co-

lyophilised with PLGA before encapsulation in discs was the dominating factor in

the rate of nerve growth factor (NGF), allowing modulation of the release [123].

Nevertheless, for particle fabrication methods based on emulsions, the efficiency of


- 62 -

PEG was mainly observed when the therapeutic molecules were lyophilised with

PEG before emulsion. When adding PEG directly in the polymer solution instead,

encapsulation efficiencies and/or release amounts have been shown to be reduced.

For instance, the encapsulation efficiency of transforming growth factors beta (TGF-

β1) in PLGA microspheres was reduced from 83% to 54% for PEG contents of 0 and

5%, respectively, and also showed a decreased cumulative mass of released GFs

[124]. In another study intended to incorporate brain-derived neurotrophic factor

(BDNF) in microparticles, a blend of PLGA:PEG was compared to a blend of

PLGA:PLGA-poly(ε-carbobenzoxy-L-lysine)(PLL)-PEG. The final cumulative

release showed a 7-fold increase for the second blend, showing the potential of PEG

used in a copolymer compared to a blend [116].

These results may serve as a useful guide for the use of PEG in electrosprayed

particles and tailoring of release kinetics.

Loading/EE/In Vitro Release and Processing Parameters

In electrospraying, the drug/matrix ratio and organic/aqueous phase ratio affect EE,

LC and in vitro release in different fashions as shown in a study encapsulating SA in

a PLA matrix [73] and summarised in Table 2.3. As mentioned before, EE decreased

with increasing SA/PLA ratio and increased with organic/aqueous phase ratio.

However opposite results were observed for in vitro release where release was

reduced with increasing organic/aqueous phase ratio and was enhanced by the

increase in the SA/PLA ratio. In the same study it was shown that increasing

SA/PLA ratio dramatically decreased the SA loading [73]. These results show how

complex the optimisation of parameters can be, especially in the case of emulsions.

Nevertheless, the study summarised in Table 2.3 represents only one case and may

not be true for every polymer/drug/solvent selection.

Table 2.3. Influence of organic/aqueous phase ratio and protein/polymer phase ratio on various

parameters. Protein was serum albumin, polymer was PLA 175 kDa. Ratios of organic/aqueous phase

ranged from 6.7:1 to 20:1 v/v. Ratios of protein/polymer ranged from 1:2 to 1:6 wt. ↑ = increase, ↓ =

decrease. *Particle size was shown to initially increase and then decrease in the studied range [73].

Viscosity Electrical

conductivity

Particle

size

Loading

capacity

Encapsulation

efficiency

In vitro

release

↑ Organic/aqueous

phase ↑ ↓ ↑ ↑ ↑ ↓

↑ Protein/polymer

phase ↑ Little effect ↑↓* ↓ ↓ ↑


- 63 -

High drug contents are generally responsible for faster release rates [72]. For

higher drug loadings, initial rate of drug release increases, as seen for 30 and 50%

loading of rifampicin in PLGA-loaded particles while a 10% loading provided zero-

order release profile. It was stated that the drug concentration affects the drug

distribution in the particle matrix, with a gradual increasing gradient of concentration

present from the centre of the particle towards the surface, which was proportional to

the drug concentration [52]. In the case of microcapsules of PLGA containing an

aqueous solution of SA obtained by coaxial electrospraying, a 0.5% loading indeed

led to an increased burst release (almost double) as compared to the 0.3% loading,

although the release rates were identical once passed the burst release [77]. A similar

result was observed for paclitaxel-loaded PLGA microparticles where 10 and 20%-

loaded particles showed similar release kinetics, with only a higher initial burst

release over the first 2 days for the 20% formulation, followed by identical sustained

release kinetics from both formulations for the remaining 28 days [54]. By increasing

the drug content in a similar matrix type, a higher amount of porosity is created in the

matrix, thus the drug diffuses more easily through though the matrix, generating an

increased burst release [65].

Matrix Use

When loading electrosprayed particles into matrices such as hydrogels, different

release kinetics may be obtained by varying the gelation time, the concentration of

the cross-linking agent and particle loading. For example, paclitaxel-loaded PLGA

microspheres (12 µm average size) loaded in alginate macrobeads (1.61-1.68 mm

average size) provided different release kinetics. Although the authors did not show

the release profiles of non-entrapped microspheres, most alginate formulations

provided zero-order release kinetics of paclitaxel over 60 days reaching over 70%

cumulative release for the best formulation (50% microsphere loading, 5 min

gelation time and 1% CaCl2) [53]. The small burst release (maximum of 10%)

observed for the 50% microsphere-loaded formulations was reduced when alginate

beads were increasingly loaded to 80 and 90% of microspheres. However, overall

kinetics were also reduced reaching a maximum of 50% and 22% cumulative release

after 60 days for 80 and 90% microsphere-loaded formulations respectively. The

extent of cross-linking did not show a clear trend, since for the 50%-loaded

formulation, extended cross-linking resulted in lower release profiles while the


- 64 -

opposite was seen for the 80% formulation and it was not significantly different for

the 90% formulation. This indicates that the microsphere loading in the matrix may

be a more determinant factor in release kinetics than the extent of cross-linking [53].

2.6.5 Denaturation

Electrospraying remains a process that employs organic solvents and therefore the

possibility of drug degradation and protein denaturation needs to be assessed and

compared to traditional encapsulation techniques to prove its superiority. So far,

limited studies have addressed this issue, nonetheless they present promising results.

The techniques generally employed are sodium dodecyl sulphate-polyacrylamide gel

electrophoresis (SDS-PAGE), Fourier transform infrared (FTIR), UV, and circular

dichroism (CD) spectroscopy [71, 77, 78, 104].

In the early stages of denaturation assessment by electrospraying, the model

protein SA was directly electrosprayed from an ethanol solution. Structural changes

were assessed by UV and CD spectroscopy showing that electrospraying of the

protein did not result in significant structural changes of SA, particularly at higher

concentrations (up to 20 mg/mL) [104]. When encapsulating the same protein in

PLGA microcapsules, no alteration in the secondary structure of SA was observed as

confirmed by comparing the CD spectra of SA before and after release from

polymeric microparticles [77]. In a study from Xie and Wang, the authors used SDS-

PAGE to investigate the protein integrity of SA released from PLGA (50:50)

microparticles after 38 days and characterised the secondary nature of SA by FTIR

and CD spectroscopy. They found that the released SA was almost identical to native

SA (after 1 day release) and no protein degradation was observed during the 38 days

release [71].

Although promising progress has been made, more studies are required to assess a

greater variety of molecules (drugs, growth factors, enzymes, DNA) in contact with

various organic solvents, and various polymers, since the purity and source of

molecule, and the nature of polymer can also influence the stability of loaded

molecules. Besides, SA remains a very stable protein which is unlikely to suffer from

denaturation. Typically, protein denaturation is potentially a major problem in

encapsulation processes involving organic solvents and it needs to be more

thoroughly assessed for the electrospraying technique. So far, when therapeutically

relevant proteins such as IGF-1, PDGF and VEGF were loaded in electrosprayed


- 65 -

particles, authors discussed the bioactivity of the released proteins by performing

cell-proliferation assays rather than using the typical assays for the assessment of

protein degradation (SDS-PAGE, CD, etc.). Since in both studies the released

proteins were shown to be bioactive, i.e. induced cell proliferation, the authors

correlated their results with denaturation, concluding that the electrospraying

technique was efficient in protecting the growth factors from denaturation [29, 30].

This approach is a nice start to degradation assessment, showing that part of the

released proteins was indeed intact; however it remains a qualitative assessment and

does not conclude quantitatively on potential protein structural changes.

Traditionally, with emulsion techniques, additives such as surfactants, carrier

proteins, sugars, salts, amino acids and polymers are considered to protect the loaded

molecules [125]. Hydrophilic additives such as SA as an excipient [126] and

poly(ethylene glycol) (PEG) [123, 127, 128] have demonstrated good protection of

growth factors in traditional emulsion techniques. However such use has not yet been

seen in electrospraying since there is little focus on denaturation of loaded proteins,

where in most studies, loadings, encapsulation efficiencies and in vitro release

profiles of electrosprayed particles remain the most discussed characteristics of these

systems. This approach is not ideal when one considers that denatured proteins will

ultimately not fulfill their intended function, despite whether they have proven to be

ideally loaded, encapsulated or released. As discussed earlier, bioactivity assessment

is an indirect way of assessing denaturation, although it does not provide thorough

description of structural changes of proteins. Therefore, although the denaturation of

protein drugs seem to be minimal through the electrospraying process, as seen with

bioactivity assays, more extensive denaturation studies are required. To this end, the

use of appropriate additives for the electrospraying technique may be identified to

fight any potential degradation of molecules. To date, only Xie et al. have published

the use of Pluronic® F-127 as an additive to tailor and enhance the protection of SA

in PLGA electrosprayed particles [71].

Another potential disadvantage of electrospraying is the use of electric fields,

since they intensify around the highly charged droplets in the course of solvent

evaporation. Such high fields may induce conformational changes of the bioactive

molecule, leading to denaturation and thus loss of bioactivity. It was indeed proven

that electrospinning of collagen out of fluoroalcohols denatured collagen to gelatin

due to the presence of high voltage [129]. Nevertheless, this hypothesis is


- 66 -

counterbalanced in the literature by the fact that, in electrospraying, the droplet size

is on the micro- to nano-scale, allowing for solvent evaporation to occur over

milliseconds, which is considered too short to have significant effects on

denaturation [44].

2.6.6 Bioactivity

The bioactivity of electrosprayed molecules was first assessed by electrospraying

insulin from an acidic water-ethanol solution. Bioactivity was assessed by comparing

the insulin receptor binding properties from electrospray-processed insulin and

control insulin. No significant differences were observed and the authors further

stated that the electrospraying technique was sufficiently ‘gentle’ not to hinder the

insulin biological activity [44]. Progressing towards the bioactivity of molecules

encapsulated within a polymeric matrix, several types of tests involving different cell

lines are presented, according to the type of encapsulated molecule: protein, anti-

cancer drug, anti-bacterial drug, antibiotic, etc.

The most commonly used model protein encapsulated in electrosprayed particles

is SA. However, SA’s bioactivity after encapsulation is rarely studied. Interestingly,

in a study from Xie et al., PLGA microparticles were used to encapsulate SA,

however lysozyme was used as the ‘model protein’ to study the bioactivity of

entrapped molecules, although SA was the focus of all other characterisations in the

paper [71]. The concentration of released lysozyme from lysozyme-loaded PLGA

microparticles was quantified by characterising the rate of lysis of Micrococcus

lysodeikticus cells by lysozyme after one day of incubation with particles. 92% of

bioactivity was calculated and it was stipulated to be much higher than with

traditional encapsulation methods (30-80%) [71]. In a similar study from the same

authors, the same assay was used for PLGA microcapsules made by coaxial

electrospraying and lysozyme bioactivity reached this time 94.6% after in vitro

release [77]. Although promising, both these studies mainly described the

encapsulation of SA (denaturation, encapsulation efficiencies and release) but used

lysozyme for depicting bioactivity. This creates a gap in characterisation since results

may not necessarily directly translate from one molecule to the other: since SA (a

plasma protein) and lysozyme (an enzyme) have different structures and size (66.4

kg/mol and 14.7 kg/mol respectively), thus they may be affected by the

electrospraying process in a different way, likely leading to different results.


- 67 -

When encapsulating anti-cancer drugs such as paclitaxel, bioactivity is generally

assessed by Coumarin-6 (C6) glioma cells (brain tumour cells) inhibition with cell

cycling analysis and 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-

sulfophenyl)-2H-tetrazolium (MTS) assay [55]. In the study from Xie et al., cell

viability was hardly affected by particle concentration but was significantly

decreased with increasing exposure time (1 to 5 days), showing a delayed cytotoxic

effect of particles, equivalent to Taxol® treatment at day 4 and 5 only [55]. The same

cytotoxicity test was used for electrosprayed PLGA films loaded with paclitaxel [57]

where a decrease of C6 glioma cell viability compared to unloaded films was clearly

seen, while an increase in the loading (5 to 30%) showed only a slight decrease in

cell viability (from around 65% to 52% viability) without being statistically

significant. In another study from Nie et al. where paclitaxel and suramin were

coaxially encapsulated in microcapsules, a continued marginal increase in apoptotic

activity of C6 glioma cells was shown after 9 days, proving the efficiency of the

capsule system to deliver anti-cancer agents in a sustained way. Interestingly, cellular

recovery was observed in free drug treated groups, indicative of the limitations of

systemic drug administration, providing only short and acute exposure due to low

terminal half-life of paclitaxel and suramin [58]. In another study the same authors

showed that apoptotic activity was increased with the delivery systems compared to

the free Taxol® groups over 9 days. They also found an increased apoptotic activity

for their co-delivery system compared to single delivery with the combination

‘suramin in the core’ and ‘paclitaxel in the shell’ (S/P) outperforming the opposite

formulation (P/S). This could be correlated with in vitro release results where the S/P

formulation released higher doses of drugs compared to the P/S formulation [87].

A similar result was observed by measuring in vitro cellular apoptosis from

alginate macrobeads containing electrosprayed PLGA microspheres releasing

paclitaxel. Although the Taxol® control group gave high apoptosis of C6 glioma

cells at day 2, it decreased at day 4 and 6 while beads formulations were giving

increased and significantly higher apoptosis over time, demonstrating the potential of

the delivery system to sustain therapeutic levels of paclitaxel [53].

When loading antibiotics to treat antibacterial infections such as ampicillin or

ripamficin in electrosprayed particles, bioactivity can be assessed by measuring the

zone of inhibition in contact with sensitive bacterial strain such as E. coli DH5α in

the case of ampicillin. When using such test, the bioactivity of ampicillin released


- 68 -

from chitosan particles was proven by a similar inhibition zone for loaded particles

and for the same amount of free drug [51]. However a very small inhibition zone was

also observed with unloaded particles, which the authors attributed to the inherent

antibacterial activity of chitosan. This result may be further investigated to ensure the

non-cytotoxicity of unloaded particles before loading of any therapeutic molecule.

For tissue regeneration and mainly angiogenesis, when growth factors such as

IGF-1, PDGF and VEGF were loaded in electrosprayed particles, their bioactivity

was respectively assessed by a smooth muscle cell (SMC) proliferation assay for

IGF-1 [29] and human umbilical vein endothelial cells (HUVECs) and lung

fibroblasts (LF) proliferation assays for VEGF and PDGF respectively [30]. The

SMC viability was assessed by a (3-(4,5-Dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide (MTT) assay and was shown to be significantly

increased over a 4 week period with exposure of released IGF-1. The results showed

that the bioactivity of IGF-1 was dependent on: the amount of IGF-1 loaded; the

amount of PLGA and its molecular weight. Briefly, IGF-1 demonstrated more

bioactivity for higher PLGA concentration, higher IGF-1 loadings and lower

molecular weight PLGA [29]. The viability of HUVECs and fibroblasts used for

determining the bioactivity of PDGF and VEGF was assessed by the PicoGreen®

dsDNA quantitation kit. The bioactivity of both GFs was shown to be high after two

days in vitro indicating minimal changes to the proteins during the electrospraying

process (around 80-90%). Interestingly, bioactivity decreased to less than 21% after

21 days, which authors attributed to the in vitro conditions that were too harsh for

growth factors, prone to oxidation and pH dependent deamidation in the in vitro

context [30].

When encapsulating therapeutic molecules in polymeric devices, the use of PEG

as an additive is shown to affect the release profiles, but also known to protect the

bioactivity of encapsulated molecules. PEG has not been used yet in electrosprayed

particles, although other polymeric devices have proven its benefits. For instance

Morita et al. indicated that co-lyophilisation of PEG and horseradish peroxidase

before exposure to organic solvents increased the retention of bioactivity [127].

Johnson et al. confirmed this theory by showing significantly more retention of nerve

GF (NGF) when PEG was co-lyophilised before encapsulation in PLGA discs [123].

Co-lyophilisation of PEG with therapeutic molecules may thus be considered in

electrospraying when the solid dispersion method is used for enhanced bioactivity.


- 69 -

2.6.7 In Vivo Performance

Most studies on electrosprayed particles loaded with therapeutic molecules are done

within an in vitro context. This approach is very important so that parameters can be

tailored and optimised in the first instance, before the use of animals to further

validate the optimised formulations. However, in vivo data remains essential for

translation of electrosprayed particles loaded with therapeutic molecules to the clinic.

Owing to electrospraying, as applied to biological loadings, being in its relative

infancy, only limited in vivo data is currently available, although these studies do

show promising results.

Most in vivo studies involve the assessment of tumour treatment by sustained

release of anti-cancer agents such as paclitaxel and suramin [53, 54, 87]. In the study

from Naraharisetti et al. for instance, 10 and 20% wt of paclitaxel were loaded in

PLGA 50:50 particles by electrospraying from a DCM solution, providing final

microparticles of 15.0 µm in diameter within a narrow size distribution of 1.7 µm

[54]. C6 glioma cells were inoculated subcutaneously to BALB/c nude mice and

loaded particles were injected to the tumour in two doses on day 14 and 28 at 0.5 mg

paclitaxel/injection. A control injection of 1 mg of commercial paclitaxel (Taxol®)

was injected only at day 14 for comparison. All the groups showed improved tumour

suppression over the placebo control and cytotoxicity of the microparticles was

evident in the analysis by hematoxylin and eosin staining of the tumour tissue when

compared with the placebo and commercial Taxol® control. Both in vitro release

profiles from 10 and 20%-loaded particles showed similar release kinetics, with an

initial burst release over the first 2 days before sustained release for the further 28

days, with only the burst being higher for the 20% formulation. Such burst release

was shown to be more effective in treating the tumour since the 10% drug-loaded

group performed poorly compared to the 20% drug-loaded group and the Taxol®

control (10% drug-loaded group had to be sacrificed at 14 days due to excessive

tumour volume, and a second injection at 28 days could not therefore be performed)

[54].

In a similar study from Ranganath et al. (Figure 2.7A-C), monodisperse

paclitaxel-loaded PLGA microspheres were obtained by electrospraying, with an

average size of 11.79 ± 2.79 µm and a smooth, spherical morphology [53]. The

loaded microspheres were further loaded in alginate macrobeads (1.61-1.68 mm

average size) (Figure 2.7A) and presented various release profiles according to


- 70 -

gelation time, concentration of the cross-linking agent and microsphere loading

(Figure 2.7B). Two formulations were selected to be implanted in subcutaneous C6

glioma tumour in mice and both showed smaller tumours in comparison to the blank

after 21 days. However, only the formulation with medium cross-linking (M80) was

able to demonstrate significantly smaller tumours compared to the free Taxol®

group, while the highly cross-linked formulation (H80) gave lower tumour formation

but not significantly (Figure 2.7C). Importantly these results were different from the

in vitro results where H80 released more paclitaxel than M80 and in a slightly more

rapid manner [53].

Still within the context of brain tumour treatment, Nie et al. prepared

electrosprayed core/shell capsules by coaxial electrospraying of PLLA for the core

and PLGA 50:50 for the shell (Figure 2.7D-G) [87]. They loaded simultaneously

both paclitaxel and suramin, with either paclitaxel in the core and suramin in the

shell (P/S formulation) which provided a sequential release, or the opposite (S/P

formulation), which provided a release in parallel (Figure 2.7F). Interestingly, in

vitro data showed that the highest apoptotic activity was obtained for the S/P

formulation over 9 days. However when looking at the in vivo results (subcutaneous

inoculation of U87 MG-luc2 xenograft in BALB/c nude mice), the P/S formulation

was best in inhibiting growth of brain tumours after 21 days (Figure 2.7G). The

authors deducted that the presence of a higher released dose of suramin at the early

stage efficiently prevented the excess growth of tumour cells while a subsequent

controlled and sustainable release of paclitaxel could induce the apoptosis of tumour

cells continuously [87].


- 71 -

Figure 2.7. (A) Representative scanning electron microscope (SEM) image of paclitaxel-loaded

PLGA microspheres (large arrow) entrapped in an alginate matrix (small arrow). Scale bar is 200 µm.

(B) In vitro release of paclitaxel from different formulations of microspheres entrapped in the alginate

matrix. L80, M80 and H80 correspond to different degrees of cross-linking (low, medium, high) with

1, 5, 15 min gelation time and 0.5, 1, 2% wt) of CaCl2 concentration, respectively. (C) In vivo

subcutaneous C6 tumour volume profiles of mice treated with different groups for 21 days (n = 5).

The control group had no beads and no drug; the placebo group was implanted with alginate

macrobeads but no drug; the Taxol® group received an injection of 180 µg of Taxol® directly in the

tumour mass; and animals in the H80 and M80 were implanted with 2 mg of alginate beads loaded

with microspheres containing an average amount of 162 µg of paclitaxel per animal [53]. (D-E)

Representative SEM images of PLLA/PLGA capsules loaded with (D) P/S formulation (paclitaxel

(PTX) in the core and suramin (SRM) in the shell) and (E) S/P formulation (SRM in the shell and

PTX in the core). Scale bar is 100 µm. (F) In vitro release of PTX and SRM from P/S and S/P

formulations. (G) In vivo subcutaneous U87 MG-luc2 tumour progression profile over the period of

treatment measured as normalised bioluminescence intensity (n = 5). The blank group received no

injection of drug; the placebo group was implanted with blank particles; the S/O group received

particles loaded with only suramin in the core while particles were loaded only with paclitaxel in the

P/O group [87]. Adapted from [53, 87] with permission. 2009 Springer [53]; 2010 Elsevier Science

Ltd. [87].

The results from both these studies underline the versatility of the electrospraying

technique being able to generate different release profiles according to the processing

parameters (in these cases being the drug loading [54], the extent of matrix cross-

linking [53] and the location of loaded drugs [87]) and generating different in vivo

results. Importantly, an initial burst release before the onset of a linear release was

shown to be more effective in tumour suppression and provided important feedback

for tailoring in vitro release profiles, showing that zero-order release kinetics are not

always desired for efficient therapeutic effect [54]. The importance of undergoing in

vivo studies is also shown to be paramount when looking at the study from

Ranganath et al. and Nie et al., where in vivo results may give different results to

what would be extrapolated from the in vitro data [53, 87].

Cu

mu

lati

ve p

aclit

axel

rel

ease

(%

)

100

60

20

0 20 40 60Time (days)

H80M80L80

Sub

cuta

neo

us

tum

ou

r vo

lum

e (m

m3 )0

1000

2000

3000

Treatment period (days)0 7 14 21

ControlPlaceboTaxol

H80

M80

A B C

Weeks after cell inoculation

0.1

10

100

0 1 2 3

BlankPlacebo

S/OP/OS/PP/S

No

rmal

ised

BLI

sig

nal

st

ren

gth

D

E

Time (days)

Cu

mu

lati

ve r

elea

se (

%)

0

20

40

60

80

100

0 6 12 18 24 30

P/SP/SS/PS/P

F G


- 72 -

2.7 THE USE OF ELECTROSPRAYED PARTICLES IN ELECTROSPUN

SCAFFOLDS

2.7.1 Electrospun Nanofibres and Drug Delivery

Amongst the many scaffolds that have been generated to date in the field of tissue

engineering (TE), some of the most promising are the scaffolds produced which

comprise nanofibre structures [19, 130]. The nanoscale could be argued as being the

most realistic scale to approach when mimicking the architecture of natural tissues.

Nanofibre scaffolds are distinctive compared to scaffolds at the micro- or macro-

scale owing to their similarity to natural extracellular matrices, like collagens, which

are the major protein components of many tissues including skin, tendon, ligament

and bone. Collagen is characterised by a fibrillar structure shown to enhance cell

attachment, proliferation and differentiation in tissue culture [19]. For this reason,

engineers aim to mimic its structure whilst fabricating engineered tissues to closely

resemble the native tissues. In addition, the high surface to volume ratio of nanofibre

scaffolds is highly favourable for drug loading, while its high porosity and

interconnected pores facilitates nutrient and waste exchange during tissue

regeneration. These scaffolds can also be further modified by various 3D surface

modification techniques to incorporate other valuable features of the extracellular

matrix [19]. Nanofibre scaffolds are studied in various areas of TE: in neural TE,

where uniaxially aligned nanofibres can be used to guide the growth of neurons

[131], in bone TE, where nanofibres may be mixed with hydroxyapatite (HA) to

mimic the bone extracellular matrix which is mainly composed of collagen and HA

[21, 131], and in cartilage TE, where nanofibre meshes can support cell spreading

and growth of chondrocytes [21].

Three main techniques have recently emerged in the production of nanofibres:

electrospinning, phase separation and self-assembly. The two first techniques mainly

use polymeric materials due to their ease of processability and capacity to provide a

large variety of cost-effective materials. All methods can produce nanofibres, even

though electrospinning has been shown to generate larger diameter nanofibres on the

upper end of the nano-range of natural collagen, rather considered to be submicron

[130]. Electrospinning remains the most widely used technique for production of

nanofibres, due to numerous advantages when compared to the other techniques; it is

simple, cost effective, reproducible and versatile: a wide range of natural and

Section 2.7 The Use of Electrosprayed Particles in Electrospun Scaffolds

- 73 -

synthetic polymer solutions can be used (collagen, silk fibroin, PCL, PLGA,

polyurethane (PU), poly(methyl methacrylate) (PMMA), etc.) and the method allows

for control of fibre diameter and alignment [19, 21, 130, 132] (Examples include 30-

120 nm for silk fibroin [133], 250-800 nm for PCL [134, 135] and 200-1,000 nm for

PLGA [136]).

Electrospinning is an electrohydrodynamic variant of electrospraying which uses

identical apparatus. Compared to electrospraying, solution electrospinning requires

polymer solutions with higher viscosity, which can be obtained by using higher

molecular weights or most generally higher polymer concentrations that ensure at

least 2.5 entanglements per polymer chain [102]. The regime used for

electrospinning must be in the semi-dilute highly entangled regime where ratio C >

3Cov which can be up to 10Cov for obtaining uniform fibres, and which depends on

the molecular weight distribution of the polymer chains in solution [103].

Nanofibres have been investigated as drug delivery vehicles as well, where drugs

could be dissolved or dispersed in the polymer solution before electrospinning [137]

or by using coaxial electrospinning wherein a secondary polymer solution containing

the biomolecules is electrospun within the core of the forming nanofibre [122, 138].

The tissue-conductive only scaffold then becomes a tissue-inductive scaffold by

releasing bioactive agents capable of inducing specific tissue treatment. Nanofibre

scaffolds applied to drug delivery have predominantly been focused on the loading of

antibiotics and anti-cancer agents [23], and there have been several reports regarding

the incorporation of growth factors into these scaffolds [25, 122, 139-142]. More

details on electrospinning and drug delivery can be found to review by Sill and von

Recum [143].

Although direct incorporation of drugs into nanofibres seems promising, there is

presently a lack of characterisation of these systems. For instance some studies have

been done in vitro, while the understanding of scaffold behaviour and effectiveness

in vivo is essential for clinical applicability of these devices [23]. In addition,

drawbacks such as low reproducibility do not allow sufficient control over drug

distribution and thus insufficient control of the release profiles. This impairs both

reproducible pharmacokinetics and pharmacodynamics. Drug aggregation in solution

is another issue which can lead to denaturation and non-homogeneous distribution

within the scaffold after processing [25, 140]. Importantly, when the scaffold is

responsible for load bearing and drug delivery simultaneously, direct incorporation


- 74 -

of drug within the nanofibres may have adverse effects on the mechanical properties

of the scaffold [144]. This is particularly important for bone applications where

bioactive agents such as growth factors must be released when the load-bearing

implant can still perform its function [145]. For instance, when NGF was directly

incorporated into the electrospinning solution, it resulted in a loss of control of the

mesh properties and in a low loading efficiency (about 3x10-4

%), which was

attributed to differences in charge densities between the GFs and polymer resulting

in a chaotic and instable jet [25]. Based on these factors, it may be concluded that

the loading of bioactive agents in nanofibres may not be ideal for controlled drug

delivery.

2.7.2 Electrospun Nanofibres and Particles for Drug Delivery

The use of loaded microparticles in nanofibre scaffolds was introduced as a response

to the drawback provided by direct encapsulation in nanofibres, where both scaffold

properties and delivery requirements were difficult to attain [146]. Separating the

drug of interest from the scaffold permits the use of a different material for

encapsulation, which allows enhanced properties for the intended function of both

the scaffold and drug reservoir. Among many others, the scaffold material requires

higher mechanical properties, slower degradation and interconnected structures for

cell infiltration, while the microparticles containing the drug need to provide positive

interactions with the drug for high loading and enhanced protection from the

environment, along with tunable degradation for the tailoring of release profiles. The

use of such composites can enhance encapsulation efficiencies but also allows

control over drug distribution within the scaffold, by providing different gradients or

loading patterns within the scaffold. Importantly, in direct encapsulation in scaffolds,

loading is often limited to only one component [58] since bioactive agents can

aggregate and denature after contact with each other when multiple loading is

attempted within the same scaffold material [147]. However, some applications such

as tissue regeneration or cancer therapy require the action of more than one type of

drug or protein being delivered in various fashions (linear, pulsatile, delayed, burst)

according to a programmed cascade triggered by cells for a specific treatment [36].

For instance in tissue regeneration, the formation of a mature vascular network is

known to involve, among others, VEGF-165 and PDGF, both with distinct temporal

actions [148]. By using separate populations of microspheres, independent bio-agents


- 75 -

can be loaded, whose release profiles are tuned with the particle characteristics,

without altering the scaffold characteristics, thus meeting both the delivery and

scaffold requirements.

2.7.2.1 Loaded Particles in Electrospun Nanofibres

The incorporation of loaded nano/microparticles into electrospun scaffolds can be

achieved by using a drug emulsion (Figure 2.8A) [149, 150] or pre-formed

microspheres (Figure 2.8B-D) [144, 151] within the electrospinning solution. In the

first case, an aqueous solution containing a bioactive agent is mixed with an organic

polymer solution, also known as emulsion electrospinning, providing aqueous

reservoirs within electrospun nanofibres. Dong et al. used this technique to

incorporate two distinct populations of nanospheres within fibres, and presented their

findings in a short communication (Figure 2.8B-C) [150]. They first loaded polyvinyl

alcohol particles with serum albumin or epidermal growth factor (EGF) by a single

emulsion process. This involved emulsifying the PVA solution containing SA or

EGF, followed by hardening of the formed nanoparticles, before incorporating in a

polyurethane solution and further electrospinning (Figure 2.8B). In terms of

parameters, increasing the concentration of PVA from 1% wt to 5% wt in the PU

solution led to the formation of larger PVA particles, with an average diameter

increasing from 200 nm to 300 nm. The fibres had an average diameter of 2 µm. The

authors managed to show the distinct populations of nanospheres within the fibres by

labelling SA and EGF with fluorescent dyes (Figure 2.8C). They commented on the

opportunity to control the release of multiple compounds, potentially at distinct rates,

but did not provide more details. No information was given pertaining to loading and

efficiency capacities, bioactivity of released molecules or release profiles [150].


- 76 -

Figure 2.8. Incorporation of loaded nano/microparticles into nanofibres by using (A) drug emulsions

[149] and (B-D) pre-formed microspheres [144, 150]. (A) Schematic overview over the four major

steps of microencapsulation in fibres by emulsion electrospinning [149]. (B) Schematic illustration of

the preparation of polyurethane electrospun fibres containing two distinct populations of

nanoparticles. (C) Overlay of fluorescence image of polyurethane fibres containing PVA/tagged-EGF

and PVA/tagged-SA particles when excited with blue light with fluorescence image when excited

with green light. Scale bar is 20 µm [150]. (D) Preparation of composite scaffolds through a sacrificial

poly(ethylene oxide) (PEO) fibre fraction coupled with a stable PCL fibre fraction (pre-wash). With

PEO dissolution (post-wash), microspheres remained entrapped within the PCL network [144].

Adapted from [144, 149, 150] with permission. 2006 American Chemical Society [149]; 2009 John

Wiley and Sons [150]; 2010 Elsevier Science Ltd. [144].

More consistently, Qi et al. formed loaded PLLA nanofibres by adding PLLA in

an emulsion of calcium (Ca)-alginate microspheres containing SA, providing

homogeneous beads-in-string structures after electrospinning [149]. Although

microspheres had larger diameters than fibres, they were found embedded within the

fibres (Figure 2.9B). The authors supposed that when the emulsion flew through the

capillary, due to the rapid jet elongation, the dispersed phase accumulated in the

centre of the liquid along the fluid direction, allowing microspheres to settle into

fibres rather than on surfaces. However, an increase in the electrospinning voltage,

above 20 kV, led to inferior morphology with a decrease in fibre diameter and

microspheres transforming into spindles. Although the final cumulative release of

microspheres in nanofibres was slightly decreased compared to blank microspheres

(less than 10% difference), the microspheres from the composite provided a lower

C

A

1. Emulsification 2. Dissolution

Core materials

Fibre forming materials

Emulsion

Capillary

Taylor cone

Jet ejection

HV generator

3. Electrospinning4. Harvesting

D

PU solution

A

PU solution

B

B

A mixed with B

Followed by electrospinning

Taylor cone

+

PCL

PEO with microspheres Pre-wash Post-wash


- 77 -

initial burst release as well as a more sustained release over the period of study (120

hours) as seen in Figure 2.9C [149]. Although the delivery characteristics of this

system were promising, no information was given on the scaffold properties and

performance after addition of microspheres.

Figure 2.9. Morphology and release profiles of loaded nano/microparticles embedded in nanofibre

scaffolds by using (A-C) drug emulsions [149] and (D-F) pre-formed microspheres [144]. (A) SEM

micrograph of electrospun PLLA fibres containing pre-made polystyrene microspheres. Scale bar is

2.5 µm (B) SEM micrograph of electrospun PLLA fibres containing Ca-alginate microspheres formed

by w/o emulsion. Scale bar is 10 µm. (C) Release profiles of SA from: Ca-alginate microspheres

(diamonds); fibres shown in B, made at 15 kV (squares); fibres with spindle particles corresponding to

electrospinning at 20kV (triangles) [149]. (D) SEM micrograph of composite scaffolds made through

a sacrificial PEO fibre fraction containing SA- and chondroitin sulphate (CS)-loaded microspheres

coupled with a stable PCL fibre fraction. (E) Release of SA or CS from PLGA microspheres. (K)

Release of both SA and CS from the single composite system containing both SA and CS

microspheres at a 1:1 ratio [144]. Adapted from [144, 149] with permission. 2006 American Chemical

Society [149]; 2010 Elsevier Science Ltd. [144].

In a similar study, chitosan nanoparticles encapsulating naproxen and rhodamine

B separately were made by ionic gelation and mixed into a PCL electrospinning

solution [151]. After electrospinning, nanoparticles were embedded in the fibres and

release rates from fibres were slower than bare nanoparticles. Different release

kinetics could be obtained by incorporating the raw molecule or incorporating the

chitosan nanoparticles containing the molecule in the PCL solution. However, less

final cumulative release was observed in the latter case for loading of rhodamine B

(18% from nanoparticles in fibres versus 70% from fibres after 70 hours) [151].

In addition to the emulsion method and direct incorporation of particles into the

electrospinning solution, an interesting technique was recently proposed, using a co-

spinning process of a sacrificial polymer (poly(ethylene oxide) (PEO)) solution

containing preformed PLGA microspheres, with a PCL solution [144]. Upon

A C

Cu

mu

lati

ve re

leas

e (%

)

0

100

80

60

40

20

1200 Time (h)40 80

E F

B

D


- 78 -

hydration, PEO was removed, leaving the microspheres entrapped between the PCL

nanofibres (Figure 2.8D). The use of sacrificial fibres in electrospinning was

previously shown to increase scaffold porosity and cell infiltration by the same

authors [152]. Using this method for incorporating microspheres was aimed at

mitigating any changes to the scaffold properties, an issue with direct encapsulation

into nanofibres [122]. In order to assess the mechanical properties of scaffolds,

polystyrene microspheres (15.7 µm in diameter) were used as a model microsphere

(since authors argued that PLGA microspheres would have dissolved in the solvent

used for electrospinning) and entrapped either within or between the nanofibres, for

mechanical comparison. When microspheres were included in PCL, it was shown

that both the stiffness and modulus decreased with increasing microsphere density.

However, when the microspheres where entrapped between the fibres, no change in

stiffness was observed for any density, and the modulus were equivalent but only for

low microsphere densities (0.05 mg of microspheres/mL of solution). In terms of

loaded biomolecules, SA and chondroitin sulphate were used for encapsulation in

PLGA microspheres through a w/o/w double emulsion. Very low encapsulation

efficiencies were obtained; 13% and 11% respectively. Release kinetics were

independent from one another and comparable to composites containing only the

single populations. A slightly more sustained release profile was observed for CS in

the scaffold compared to free microspheres (25 days, maximum release of more than

60% in all cases), while the maximum release was reached after only 5 days for SA

with about 20% for free microspheres and 10% for microspheres within the scaffold

(Figure 2.9D-F)) [144].

All these studies underline the increasing interest for incorporating particles in

electrospun scaffolds, although release profiles are not yet optimal and their effects

on cells in both the in vitro and in vivo environments remain to be addressed, along

with the characterisation of mechanical properties of scaffolds after incorporation of

loaded particles.

2.7.2.2 Multiple Electrospraying/Electrospinning

2.7.2.2.1 Concept

As explained previously, the electrospinning/electrospraying processes use simple

apparatus consisting of syringe pumps, collectors and external voltage supplies, and

thus they can be easily manipulated to fit specific requirements (horizontal, vertical,


- 79 -

or angled setups). For this reason, these apparatus may be easily used in

combination, for the production of composites with an increased number of

properties. Early attempts to electrospray whilst electrospinning consisted of side-by-

side capillaries and a flat collector moving on an x-y stage [153]. Although the

scaffolds yielded were 100 µm in thickness after 45 min, the area of stream

convergence was so small that non-uniform integration was obtained. The authors

attributed this problem to a stream repulsion effect from Coulombic forces, which

they limited by locating the nozzles perpendicular to one another and using a rotating

mandrel translating on its axis. Stream repulsion was minimised and the combination

of rotation and translation of the mandrel target provided an ideal integration of both

components (electrosprayed smooth muscle cells and poly(ester urethane) urea

(PEUU) fibres in this case). Using this configuration, 5 × 5 cm construct sheets

ranging from 300 to 500 nm in thickness were created and scaffold thickness could

be controlled by adjusting polymer flow rate or fabrication time. The authors

concluded that this setup may find other applications in the future as a means to

fabricate more uniform composite scaffolds by electrospinning multiple materials or

introducing drug-laden microspheres between fibres, a setup which has indeed been

used in consecutive years for either multiple electrospinning or simultaneous

electrospraying/electrospinning [153].

2.7.2.2.2 Multiple Electrospinning

Compared with single electrospinning, more versatility in properties can be achieved

with multiple electrospinning. The drug delivery characteristic can indeed be

effectively coupled with desired mechanical properties. This was obtained, for

example, by simultaneously electrospinning PLGA fibres loaded with tetracycline

hydrochloride (TET), for antibacterial activity, with PEUU, that maintained the

required elastomeric properties [154]. The use of a rotating mandrel for collecting

these constructs is the best approach in multiple electrospinning. It allows aligned

fibres, a configuration highly desirable in tissue engineering, since it mimics some of

the fibrous musculoskeletal tissues, like tendons and ligaments. Although fewer

studies on multiple electrospraying/electrospinning are available, they also use a

rotating mandrel collector and most of these studies face similar issues relating to

this mandrel approach.


- 80 -

Fundamentally, the main limitation in aligned fibres obtained by collection onto a

rotating mandrel is the high density fibre packing, since fibres are drawn in parallel

to one another. This becomes an issue for cellular infiltration, where in most cases,

cellular and tissue formation are often limited to the surface of the electrospun

construct, impairing the necessary cell growth within the central architecture of the

construct [30]. The use of a second electrospinning apparatus was proposed for

simultaneous electrospinning of sacrificial fibres that, once they were removed from

the scaffold, conferred an increased porosity, beneficial for cell infiltration, while

maintaining the anisotropy of the scaffold. To this end, PCL and PEO were co-

electrospun onto a rotating mandrel, followed by dissolution of PEO into water after

production (Figure 2.10B) [152]. Importantly, cell infiltration and distribution after

three weeks in culture increased in the starting sacrificial fraction when scaffolds

were seeded with mesenchymal stem cells [152]. On the other hand, limited cell

infiltration was reported when a PCL/collagen blend was co-spun with PEO [155].

Certainly the electrospinning parameters have a great influence in the fibre

deposition and fibre characteristics and must be optimised for effective improvement

in cell infiltration. Another approach to improve cell infiltration was proposed which

involved simultaneously electrospinning microfibres and nanofibres, obtained by

melt and solution electrospinning, respectively (Figure 2.10A). Microfibres increased

the porosity of scaffolds to facilitate cellular infiltration and nanofibres gave an

enhanced effect on cell attachment and growth due to the nanoscale features (Figure

2.10C-E). The so-produced PLGA composite scaffolds provided significantly higher

attachment and spreading of both human epidermal keratinocytes and fibroblasts

(Figure 2.10F-H) [156].

Another drawback of electrospun fibres from synthetic polymers, which may be

overcome by multiple electrospinning, is the lack of biological recognition. For some

applications, however, such as vascular grafts, a cell-responsive surface is

paramount. This has been achieved by simultaneously electrospinning PCL and silk

fibroin, for their respective mechanical and cell-conducive properties, onto a rotating

mandrel, conferring anisotropic properties as well [157]. More than a simple

overlapping of nanofibres, double electrospinning provided a high integration of both

types of fibres and the change of mandrel rotation speed may render the anisotropy

tunable. This may be kept in mind when optimising the electrosprayed/electrospun

constructs.


- 81 -

Figure 2.10. Multiple electrospinning. Schematic of (A) co-solution/melt electrospinning on a rotating

mandrel [156] and (B) photo of co-solution electrospinning apparatus [152]. (C-E) SEM images of

three types of nano-/microfibre composite scaffolds: (C) nanoparticle/microfibre scaffold for 1% wt

solution of the nano-component, (D) beaded nanofibre/microfibre scaffold for 9% wt, (E)

nanofibre/microfibre scaffold for 10% wt. Scale bar is 20 µm. (F) Cell numbers of human epidermal

keratinocytes (NHEK) and human epidermal fibroblasts (NHEF) that adhered to two types of

scaffolds after 1 h (means ± SD, n = 4). (G-H) Micrographs of NHEF in (G) PLGA microfibre

scaffold and (H) PLGA nano-/microfibre scaffold (10/90) [156]. Adapted from [152, 156] with

permission. 2010, 2008 Elsevier Science Ltd.

2.7.2.2.3 Applications of Multiple Electrospraying/Electrospinning

The multiple electrospinning devices have recently proven quite promising in

enhancing the typical properties obtained with single electrospinning. In a similar

fashion but different scope, the association of electrospinning with electrospraying

was proposed to provide a 3D structural construct of nanofibres embedded with

electrosprayed particles for varied applications such as drug delivery, coatings or

cellularisation of the constructs. Such a simultaneous process would allow the

composite production in a single step sequence and permit a better integration of

particles within the scaffold. Due to the versatility of both processes, several types of

scaffolds could be easily and quickly achieved while adding extra properties without

affecting the essential properties required for scaffolds.

2.7.2.2.3.1 Drug Delivery

The application of coaxial electrospraying technique in association with

electrospinning applied to drug delivery was first reported in 2009 by Wang et al.

and shown in Figure 2.11[29]. They created a soft tissue-engineered construct (TEC)

PLGA microfibre

PLGA nanofibre

E

Oil circulator

Heat nozzle

Power supply

Collection drum

Syringe pump

MandrelMotor

Fanner

Pump

NeedleShield

F PLGA microfibre

PLGA nanofibre/microfibre

Ce

lls/3

3.7

5 m

m²

NHEK NHEF

G

PLGA microfibreC

PLGA nanoparticle

H

A

B PLGA beaded nanofibre

D PLGA microfibre


- 82 -

with anisotropic structure, able to deliver growth factors for the survival of cells

which were often subjected to hypoxia and a nutrient starvation microenvironment in

the context of TECs [158]. The co-spinning technique enabled simultaneous

electrospinning of polyurethaneurea (PUU) nanofibres and electrosprayed PLGA

microcapsules of an IGF-1 gelatin solution, obtaining the direct assembly of a

scaffold onto a rotating mandrel collector (Figure 2.11A). Results showed that the

release profile and bioactivity of IGF-1 were dependent on: the amount of IGF-1

loaded (tested with 50 and 150 µg/mL); the amount of PLGA (tested with 5 and 10%

wt) and the molecular weight (tested with high MW (40-75 kDa) and low MW (5-15

kDa)). The release profile was triphasic with an initial burst release attributed to the

imperfect core-shell structure of the microcapsule (Figure 2.11E). It was

hypothesised that during the travel of the microcapsules toward the collector, the

inner part of the shell may have solidified slower than the outer part, allowing the

leakage of the aqueous IGF-1 into the shell, becoming trapped there. The increased

release occurring after 3 weeks was attributed to the release of accumulated acid

from PLGA bulk, creating pores that allowed the encapsulated IGF-1 to quickly

diffuse out. Bioactivity was maintained over the 4-week study period and the cell

growth on all loaded scaffolds was assessed in vitro for a 7-day culture period under

normal conditions and under hypoxia/nutrient starvation conditions with a MTT

assay (Figure 2.11F). The authors stated that the loaded scaffolds were able to

significantly enhance cell growth at day 7 in both types of conditions. However, by

correlating these results with the release results observed from day 7 to day 21,

where the IGF-1 release is almost inexistent, it may have been expected that cell

survival would decrease after day 7. The authors also performed mechanical studies

on the scaffolds and observed that the incorporation of PLGA microspheres did not

significantly alter tensile strength, modulus and elongation break at the perpendicular

direction, while it did in the alignment direction, which may be a potential concern.

However mechanical properties at the perpendicular direction were very weak

compared to those at the alignment direction, before and after incorporation of

microspheres, which may be why incorporation did not alter significantly the

mechanical properties at the perpendicular direction [29].


- 83 -

Figure 2.11. (A) One-step fabrication of protein loaded microcapsules and nanofibre scaffolds by

simultaneous coaxial electrospraying/electrospinning techniques. (B-D) Structure of the fabricated

microcapsules. FITC-labelled SA was added into the protein solution, and rhodamine-B was loaded

into the PLGA solution before fabrication. The resulting microcapsule showed: (D) a core-shell

structure with (B) protein solution as the core and (C) PLGA as the shell. Scale bars are 2 µm. (E)

IGF-1 release kinetics from scaffolds fabricated with different PLGA concentration and viscosity and

IGF-1 loading at 37°C. (F) Effect of IGF-1 loading on MSC survival under hypoxia/nutrient starvation

conditions. MSCs were cultured for 1 day under normal culture conditions (21% O2, 5% O2, and 20%

fetal calf serum (FCS)) followed by 6 days under hypoxia/nutrient starvation conditions (5% O2, 5%

CO2, and 1% FCS). (G) Surface morphologies of scaffolds embedded with loaded microcapsules.

Abbreviations: HV: 40-75 kDa PLGA, LV: 5-15 kDA PLGA, HV0: scaffolds with no microspheres,

HV5-50: 5% PLGA – 50 µg/mL IGF-1, HV10-50: 10% PLGA – 50 µg/mL IGF-1, HV10-150: 10%

PLGA – 150µg/mL IGF-1, LV10-50: 10% PLGA – 50µg/mL IGF-1. Adapted from [29] with

permission. 2009 American Chemical Society.

2.7.2.2.3.2 Other Applications

Coating

A clear advantage in simultaneous electrospraying and electrospinning can be found

when the application is coating of nanofibres. This is achieved by electrospraying of

hydrogels as well as non-polymeric particles, such as metal oxide nanoparticles,

ceramics, or even cells. Using a simultaneous device for coating was first of all

shown to be more effective as compared to electrospinning and electrospraying in

sequence. In a comparative study, Jaworek et al. assessed the merits of simultaneous

electrospraying during the electrospinning process against electrospraying onto the

same rotating drum after electrospinning was completed and electrospraying onto the

electrospun mat removed from the drum and placed onto a heated table [159]. Metal

0

2

4

6

8

10

12

0 7 14 21 28

HV5-50HV10-50HV10-150LV10-50

14

Cu

mu

lati

ve IG

F-1

co

nce

ntr

atio

n

(ng

/mL)

Release time (days)

E

Gelatin/BSA/IGF-1 solution

PLGA solution

PU solution

Syringe pump A

Syringe pump B

Rotation

Collecting mandrel

Syringe pump C

A

0

20

40

60

80

100

HV10-150 Perc

enta

ge o

f sur

vive

d M

SCs

(% o

f da

y 1v

alue

)

1 day

7 day

HV10-0 (no IGF-1)

F

B C D

G

HV0 HV5-50 HV10-50 LV10-50


- 84 -

oxide nanoparticles such as TiO2, MgO and Al2O3 (20-100 nm size) were deposited

on poly(vinyl chloride) (PVC), polysulfone (PSU) or nylon nanofibres of a

maximum diameter of 500 nm. The authors observed that the simultaneous process

produced particle coating with lower density, but particles were distributed more

uniformly between fibre layers, an advantage for homogenous coating. Post-spinning

deposition allowed production of denser layers, but the particles were mainly

deposited on the mat surface, with only minor penetration into the mat while the

post-spraying as a separate process gave denser coating. However, in this latter case,

the coated surface was limited to the base of the spray plume that required scanning

deposition onto the mat in order to cover larger areas [159].

Bone Tissue Engineering

The simultaneous process was used for several studies requiring coating of fibres,

due to a need for homogeneous coating. For instance in bone tissue engineering,

electrospun nanofibres can mimic the composite nature of bone but lack the

osteoconductive property, which may be counterbalanced by using a blend of

hydroxyapatite, the mineral component of bone. However, blending HA with the

nanofibre material may mask the osteoinductive property of HA since the particles

are completely embedded inside the polymer fibres. Therefore, an electrosprayed

coating of HA on nanofibres was proposed to create a better environment for growth

and mineralisation of bone cells. A poly(L-lactic acid)-co-polycaprolactone

(PLACL)/gelatin blend was spun along with a HA methanol solution on a rotating

mandrel and the resulting properties were compared with direct blending of HA in

the polymer solution [160]. The electrospun fibres presenting electrosprayed HA

particles showed better cell proliferation, enhanced mineralisation and alkaline

phosphatase activity (ALP). This was due to the exposure of HA to the cells which

gave them the necessary cues to start to lay down bone matrix, but it also

enhanced/roughened surface topography which is preferential for cell adhesion.

Mechanical properties were also superior to the blend, collectively proving that

electrospraying of HA in combination with electrospinning of nanofibres produced

suitable osteoconductive scaffolds for bone tissue regeneration [160]. The same

authors also used this process to coat electrospun gelatin only with HA, followed by

cross-linking with 50% glutaraldehyde solutions, whose cytotoxic effect was negated

by washing and drying of the scaffolds. Results were compared with electrospun

HA/Gelatin nanofibres of different HA/Gelatin ratios. Electrospray-coated


- 85 -

nanofibres had a higher pore size and porosity than blended nanofibres, as well as

larger fibre diameters. Similarly to the previous study, proliferation and ALP activity

were significantly higher for electrospray-coated nanofibres at 5, 10 and 15 days of

culture, again due to the complete exposure of HA on the surface of nanofibres.

Cross-linking was found to confer better stability and mechanical properties than for

non-cross-linked scaffolds with a tensile strength of 2.7 MPa and a strain at break of

41.5% which are close to suitable values for guided bone tissue regeneration [161].

Cell Infiltration and Vascularisation

Co-spinning has also been employed for coating electrospun PCL/collagen

microfibres with electrosprayed Heprasil™, a synthetic hydrogel comprising

chemically modified hyaluronic acid (HyA) and heparin as an attractive template for

cells [155]. By comparing only microfibres with nanofibres, better cell infiltration

was shown for microfibres. Technical considerations included the size of the mandrel

used during the co-spinning process with 0.8, 1.4 and 1.7 cm diameter leading to a

20, 30 and 70% Heprasil collection efficiency respectively. As expected, larger

mandrels were able to capture the hydrogel droplets more efficiently and 1.7 cm was

further used to limit losses. Heprasil was loaded with AlexaFluor488-labeled SA,

allowing visualisation of the random dispersion of Heprasil regions within the

composite. Cell infiltration in Heprasil-coated PCL/collagen microfibres was

significantly higher than uncoated fibres, reaching more than 200 µm compared to 50

µm respectively, after 10 day culture with human foetal osteoblasts. The authors

stated that the inclusion of Heprasil regions within the mesh created a reduction in

the volume density of fibres and created compartments of hydrogel for cells to

further infiltrate [155].

A prospective advantage of co-deposition of hydrogel is also the loading of

bioactive molecules into the composite. Indeed the same authors further used their

device to load angiogenic factors (VEGF and PDGF), in order to recapitulate the

vascular system essential in all tissue-engineered constructs [30], which is often hard

to achieve (Figure 2.12B-F). They loaded the growth factors in the Heprasil hydrogel

mix which was further electrosprayed simultaneously with electrospinning of the

PCL/collagen blend microfibres, obtaining 200 ng/cm² of growth factors for a 32 cm²

area of PCL/collagen-Heprasil co-deposition (Figure 2.12B).


- 86 -

Figure 2.12. (A) Schematic of the microintegration process of SMCs into PEUU fibres for small-

diameter blood vessel construct fabrication. A perpendicular nozzle configuration was utilised for

electrospinning PEUU (6% wt in hexafluoro-2-propanol at 1.5 mL/h) and electrospraying SMCs (7.5

× 10-6

SMCs/mL in medium at 0.1 mL/min) onto a rotating small-diameter mandrel (4.7 mm, 250

rpm) transversing on a linear stage (1.6 mm/s). The macroscale appearance of SMC microintegrated

PEUU tubular constructs are illustrated after removal from the mandrel (bottom right hand corner) [162]. (B-F) 3D vascularisation of GF-releasing hybrid scaffold of PCL/collagen fibres and HyA

hydrogel [30]. (B) Schematic of the co-electrospraying/electrospinning setup used for production of

meshes, allowing simultaneous deposition of PCL/collagen fibres and HyA hydrogel. (C) Cellular

infiltration measured by von Willebrand factor (vWF) section staining (co-culture of HUVEC cells

and LF) and endothelial network formation in meshes cultured in media supplemented with

VEGF/PDGF (+ve control), meshed loaded with VEGF/PDGF in the Heprasil component during

fabrication (200 ng/cm2 each) (GF load), meshes loaded shortly prior to seeding (200 ng/cm

2 each)

(Pre-load) and meshes lacking VEGF/PDGF (-ve control). Scale bars are 50 µm. (D) Release of

VEGF and PDGF from meshes studied in vitro via ELISA. (E) Bioactivity of the incorporated VEGF

and PDGF by cell proliferation assessment. Percentage bioactivity was obtained through comparison

with equal amount of fresh VEGF and PDGF as 100% values. Adapted from [30, 162] with

permission. 2007, 2010 Elsevier Science Ltd.

The use of a co-culture assay of human umbilical vein endothelial cells and lung

fibroblasts with sequential seeding of LF followed by HUVECs permitted the

infiltration of cells in the mesh with a HUVEC:LF ratio 1:5 being the most

satisfactory, while seeding alone of HUVECs or higher HUVEC:LF ratios did not

yield favourable results. Cells also exhibited more physiological morphologies as

compared to conventional tissue culture plastic, reflecting a more physiological

cellular state that ultimately influence cellular function and behaviour. In terms of

release profiles, a burst release followed by sustained release was observed for both

GFs with approximately 48% and 30% of the total loaded VEGF and PDGF,

respectively, being released after 5 weeks (Figure 2.12D), while bioactivity of both

GFs constantly decreased from around 80-90% bioactivity after two days in vitro to

1-20% after 21 days, which authors explained by VEGF in particular being

Syringe pump 1 (SMC feed)

Syringe pump 2 (PEUU feed)

Electrospraying SMCs

Electrospinning PEUU

SMC microintegratedconduit

A B

Electrospinning of PCL/collagen

Electrospraying of hydrogel

Rotating mandrel

Power supply

D E F

Infi

ltra

tio

n d

ep

th (

µm

)

0

50

100

150

200

250 +ve controlGF loadPre-load-ve control

+ve control GF load

Pre-load -ve control

CC

um

ula

tive

re

leas

e (%

)

VEGFPDGF-BB

0

40

80

5 15 25Time (days)

35Time (days)

2 7 21

Bio

acti

vity

(%

)

0

80

40

VEGFPDGF-BB


- 87 -

susceptible to pH dependent deamidation and oxidation in vitro (Figure 2.12E).

Importantly cell penetration after 14 days of co-culture was shown to be similar for

the GF-loaded group and the positive control (constructs were only cultured in

endothelial media (EBM-20) supplemented with VEGF and PDGF) and was

significantly higher than the pre-load group (direct GF incorporation prior to cell

seeding, equivalent to bolus injection), reaching approximately 190, 210 and 85 µm

of infiltration depth in average, respectively (Figure 2.12C,F). In conclusion, the

PCL/collagen-Heprasil loaded hybrid scaffolds were shown to be able to recapitulate

the primitive capillary network required for vascularised TECs, by initiating a

capillary network not only on the surface but also throughout the scaffolds. However,

the previous release profiles and bioactivity results suggest that this system may be

effective only in the first days of cell culture, rather than providing a continuous

effectiveness over several weeks of culture. The morphogenic and chemotactic

actions provided by this initial kick-start may be responsible for initial migration of

cells in the constructs, triggering subsequent formation of endothelial network [30].

Electrospraying of Cells

Electrospraying has also been employed to produce cellularised constructs by

simultaneous electrospraying of smooth muscle cells and electrospinning of PEUU

nanofibres [153]. Such co-processing allowed the integration of cells into the

smallest pores of the electrospun scaffold as it was constructed, providing a large

numbers of cells which infiltrated throughout the bulk after a few days of perfusion

culture, which had spread within the scaffold. Importantly, there was no significant

decrease in cell viability and electrosprayed SMCs spread and proliferated at a

similar rate than the control unprocessed SMCs while cells sprayed from a bottle

without voltage did not. The sprayed cell suspensions were supplemented with 3%

wt bovine skin gelatin for increasing viscosity and maximising viability by protecting

cells from mechanical and chemical stresses, since the physical forces of the

pressurised spray in combination with the exposure of cells to processing solvents

initially caused a significant reduction in SMC viability. Mechanical integrity was

disrupted because of gelation within the fibre network. Because viability and

proliferation of electrosprayed cells were not affected, they were electrosprayed with

media alone, maintaining the mechanical properties of the construct. These results

underline the advantage of electrospraying over simple spraying and are consistent


- 88 -

with literature stating that cells can survive exposure to high voltage [153].

Importantly, the SMC-integrated PEUU composites presented lower tensile strengths

and higher breaking strains, which were explained by the cells disrupting the PEUU

fibre network and replacing elastic PEUU volume with cellular volume. The authors

still concluded that the measured properties were still more than sufficient for the

SMC-integrated PEUU composites to serve as a support structure for soft tissue

growth and mechanical training.

Following these encouraging results, the same authors extended their process to

the fabrication of small-diameter tubular conduits that possess mechanical properties

similar to native blood vessels, after only a few days in culture (Figure 2.12A) [162].

A 4.7 mm diameter mandrel was used in place of the previously employed 19 mm

for sheets [153]. Interestingly they decreased the TTC distance from 5.0 to 4.5 cm

and lowered the mandrel negative charge from -10 to -3 kV to obtain reproducible

and defect free small-diameter tubular constructs. SMC integration was uniform

radially and circumferentially within the conduits after initial static culture, while

conduits were strong and flexible with mechanical properties that mimicked those of

native arteries. Cultures of such cell-based scaffolds are recommended to be

performed in spinner flasks or perfusion rather than static, since in both cases they

led to much higher viable cells and enhanced spreading within the electrospun fibres

[153, 162].

In 2008, mention of simultaneous electrospraying of chondrocytes and

electrospinning of PCL was made in a review by Wu et al. [94]. They stated that

confocal microscopy was used to visualise the living cells embedded in the fibres

after being cultured in the cell media for a set time not mentioned. They also stated

that the experimental results revealed that 80% of the cells were still viable after the

electrohydrodynamic process, while no more information than this was provided (no

description of materials or methods).

Although the futility of the last study, electrospraying of cells with simultaneous

electrospinning of a polymer matrix remains an efficient and rapid method for the

production of tissue-engineered constructs. However, this is by no means a trivial

and straight forward procedure and issues of sterility and time required to produce

thicker scaffolds may potentially limit this application [152].

Section 2.8 Conclusions

- 89 -

2.8 CONCLUSIONS

The controlled and targeted delivery of therapeutic molecules is tantamount to the

success of many medical treatments. With the development of superior treatment

options for cancer, asthma and hormonal therapies there is a concomitant demand to

encapsulate and release the active molecules in a safe, reproducible and effective

manner.

The technique of electrospraying has emerged as a promising technology to

produce particles with entrapped therapeutic molecules which may be released as

the particle degrades. The size and morphology of the particles produced are of

paramount importance to enable batch-to-batch reproducibility and appropriate

efficacy of the system. We have reviewed the many variables and interplays of the

processing parameters which affect the production of microparticles and have

highlighted the shortfalls associated with many current technologies. Importantly we

have also highlighted the need to thoroughly assess and publish the encapsulation

efficiencies, bioactivity and denaturation of the encapsulated biomolecules, both in

vitro and in vivo. Only when all of these considerations are properly tackled can

a delivery system for the use in targeted biomolecule delivery - for example in tissue

engineering, be properly realised and translated to the clinic.

2.9 ACKNOWLEDGEMENTS

Thanks to Dr. Tristan Croll from the Tissue Repair and Regeneration program for

molecule designing, Jaime Nakahara for proof-reading and to the Biomaterials and

Tissue Morphology group, Tissue Repair and Regeneration program, Regenerative

Medicine group, IHBI and the Australian Research Council (ARC) (Discovery grant

no. DP0989000) for financial support.

- 91 -

Chapter 3: Electrospraying, a Reproducible

Method for Production of Polymeric

Microspheres for Biomedical

Applications

Nathalie Bock1,2,3

, Maria A. Woodruff1, Dietmar W. Hutmacher

2, Tim R.

Dargaville3

Published in Polymers, Volume 3, Issue 1, 2010, Pages 131-149.

© 2010 by the authors; licensee MDPI, Basel, Switzerland.



Nathalie Bock Developed the research questions

Designed and performed the experiments

Analysed and interpreted the results

Conceived and wrote the manuscript



Dietmar W. Hutmacher* Involved in the conception of the project



Assisted with thermal characterisation







Chapter 3 Reproducible Polymeric Microspheres by Electrospraying

- 92 -













requirements.





- 93 -

3.1 ABSTRACT

The ability to reproducibly load bioactive molecules into polymeric microspheres is a

challenge. Traditional microsphere fabrication methods typically provide

inhomogeneous release profiles and suffer from lack of batch to batch

reproducibility, hindering their potential to up-scale and their translation to the clinic.

This deficit in homogeneity is in part attributed to broad size distributions and

variability in the morphology of particles. It is thus desirable to control morphology

and size of non-loaded particles in the first instance, in preparation for obtaining

desired release profiles of loaded particles in the later stage. This is achieved by

identifying the key parameters involved in particle production and understanding

how adapting these parameters affects the final characteristics of particles. In this

study, electrospraying was presented as a promising technique for generating

reproducible particles made of polycaprolactone, a biodegradable, FDA-approved

polymer. Narrow size distributions were obtained by the control of electrospraying

flow rate and polymer concentration, with average particle sizes ranging from 10 to

20 µm. Particles were shown to be spherical with a homogeneous embossed texture,

determined by the polymer entanglement regime taking place during electrospraying.

No toxic residue was detected by this process based on preliminary cell work using

DNA quantification assays, validating this method as suitable for further loading of

bioactive components.

Figure 3.1. Abstract figure. Electrosprayed PCL microspheres.

10 µm


- 94 -

3.2 KEYWORDS

Electrospraying, drug delivery, microspheres, polycaprolactone.

3.3 INTRODUCTION

The use of polymeric particles has been of great interest in the biomedical field for

the last 50 years, with a particular niche present in the field of drug delivery [18,

163]. Current technologies allow the production of biodegradable nano- and micro-

sized materials able to encapsulate therapeutic molecules to be gradually released

through diffusion and degradation in vivo. The motivation behind this process is the

possibility to overcome the limitations faced in bolus delivery of molecules,

especially proteins where the harsh in vivo environment can cause denaturation,

shortening their half-life after delivery, and thus reducing action efficacy [35].

Polymeric particles are therefore presented as reservoir systems able to protect the

proteins from their environment, enhancing their long-term biological activity.

Ideally, these systems are also able to provide tailored release rates, required by

certain therapies, by the control of particle morphology, size and polymeric matrix

[17]. Importantly, such particles have the potential to minimise the propagation of

drug payloads to non-targeted areas, limiting unwanted effects and allowing a site-

specific delivery [18, 35, 40, 163].

Several techniques have been presented in the literature for fabrication of

polymeric nano- and microparticles, the most popular being based on emulsion

techniques [40, 84, 120, 126, 164, 165]. Molecules are dispersed or dissolved into a

polymer solution and emulsified to form micro-droplets that are further dried after

solvent removal [26]. However, the use of organic solvents, unless carefully

controlled, is a drawback in many of these techniques since it can lead to

denaturation of protein-based drugs during processing, increasing the variability in

encapsulation efficiencies and loading capacities [126]. Secondly, the size

distributions of particles fabricated by emulsion-based techniques tend to be

inhomogeneous and broad, contributing to their lack of reproducibility, which in turn

hinders their clinical use. Size distribution is a crucial parameter and it was shown

that monodisperse size would enable a better control of release profiles and

bioavailability of the loaded drug in the body [59, 64]. Particle morphology is also

important since it affects the internalisation by non-phagocytic cells and the


- 95 -

degradation of the polymer matrix which, in turn, determines the release kinetics of

the loaded component. Therefore, homogeneity of morphology is also an important

consideration to ensure particles with reproducible characteristics are obtained.

Fabrication of polymer microparticles by electrospraying has the potential to

overcome the limitations of emulsion-based techniques and to provide reproducibly

loaded nano- and microparticles [51]. Electrospraying is a one-step technique which

has potential to generate narrow size distributions of submicrometric particles, with

limited agglomeration of particles and high yields [97]. The principles of

electrospraying are based on the ability of an electric field to deform the interface of

a liquid drop, established by Lord Rayleigh in 1882 [75], further developed by

Zeleny in 1917 [166] and Sir Taylor in 1964 [167]. The theory of charged droplets

states that if an electrified field is applied to any droplet, the electric charge generates

an electrostatic force inside the droplet, known as the Coulomb force, which

competes with the cohesive force intrinsic to the droplet. When the applied Coulomb

force is able to overcome the cohesive force of the droplet manifested in the surface

tension, the droplet will undergo breakup into smaller droplets in the micro- to nano-

scales. This phenomenon begins at the Taylor Cone, referring to the progressive

shrinkage of the unstable, charged macro-droplet into a cone from which the smaller

charged droplets will be ejected as soon as the surface tension is overcome by the

Coulomb force. Once the charged droplet is in flight towards the collector, Rayleigh

predicted a limit where subsequent break-up of the droplet may occur, called the

Rayleigh limit, LR, expressed in Equation 3.1:

( ) (3.1)

Droplet break-up is also known as Coulomb fission and is an unwanted

phenomenon in monodisperse electrospraying, since it generates bimodal size

distributions [49, 52, 110, 168, 169]. LR is a function of q the surface charge of the

droplet, ε the permittivity of the surrounding medium, γ the surface tension of the

liquid and R the radius of the droplet. The maximum surface charge of the droplet is

given by Equation 3.2:

√ (3.2)


- 96 -

Based on these equations, monodisperse electrosprayed particles can be fabricated

by using appropriate parameters that allow enough chain entanglements in charged

droplets before the Rayleigh limit is reached.

Electrospinning is based on the same principles of charged droplets involving the

ejection of a nano-jet instead of droplets from the Taylor Cone. The difference

between the electrospinning and electrospraying techniques lies in the chain

entanglement density of the polymer solution [102]. Previous studies have

demonstrated that a critical polymer concentration called Cov can dictate the

behaviour of electrospraying/electrospinning [103]. This critical concentration can be

found for each type of polymer solution and represents the critical chain overlap

concentration where entanglement begins to occur. In order to produce fibres, the

polymer concentration, C, must be chosen such that a threshold ratio C/Cov is

overcome. Therefore, at low chain entanglement density, electrospraying of droplets

instead of electrospinning of fibres will occur at the Taylor Cone. The C/Cov ratio

must be determined experimentally for each type of polymer, for example C/Cov for

poly(methyl methacrylate) is between 3 and 10 depending on the molecular weight

distribution of the polymer chains in solution [103].

The electrospraying process is conceptually simple: a polymer solution is loaded

into a syringe and infused at a constant rate using a syringe pump through a small but

highly charged capillary (e.g. a 16-26 gauge needle). The applied voltage used is

typically up to + or - 30 kV and the collector might be placed at a 7 to 30 cm distance

from the capillary. Once the droplets have detached from the Taylor cone, the solvent

evaporates, generating dense and solid particles, propelled towards the collector. In

the context of drug loading, the bioactive molecule is mixed to the polymer solution

before electrospraying and can further be emulsified [73]. Some studies which have

been undertaken include encapsulation of hydrophilic and hydrophobic model drugs

[64], model proteins [47, 56, 72-74], antibiotics [51, 52] and anti-cancer drugs [59]

in polylactide (PLA) [64, 73], poly(lactic-co-glycolic acid) (PLGA) [52, 56],

polycaprolactone (PCL) [56, 59] and chitosan [51].

During the electrospraying process, there are several parameters which all have an

inter-dependent influence on viscosity, electrical conductivity, particle size,

distribution, encapsulation efficiencies, loading capacities and in vitro release

profiles [47, 51, 56, 64, 73, 97, 110]. These parameters include voltage, distance to

collector, needle gauge, flow rate, polymer, drug, solvent, surfactant,


- 97 -

protein/polymer ratio and organic/aqueous ratio. As a consequence, although

electrospraying is a promising technique, the number of parameters to be used can

render its optimisation highly complex. The characteristics of electrosprayed

particles are still not completely understood and it is important to proceed in a step-

wise manner intended to understand the relationship between processing parameters

and characteristics of electrosprayed microparticles before one progresses to the

inclusion of highly fragile and expensive bioactive molecules. This study details the

reproducibility of the process and identifies the key parameters responsible for

particle size, distribution and morphology as a prelude to using the validated

methodologies for the loading of a bioactive molecule.

Previous studies have correlated the effects of key variables of electrospraying

[49, 52, 110, 168, 169]. Most of these studies are PLGA-based, which are well-

known as the most common biodegradable and FDA-approved polymers in tissue

engineering literature. However, in the context of drug delivery for orthopaedic

applications, a slower degrading polymer like polycaprolactone should be

considered. PCL is also FDA-approved and various drugs have been encapsulated in

PCL microspheres and nanospheres [88, 90]. PCL is highly permeable to small drug

molecules and degrades through its ester linkages. As compared to PLGA-based

polymers, it also presents the advantage of a lesser acidic environment being

generated during degradation [90], however, only very few studies have investigated

the production of electrosprayed PCL particles [71, 80, 98]. As aforementioned, the

type of polymer will give different characteristics of electrosprayed particles and due

to the complexity and inter-dependence of variables involved in the process, PLGA

production parameters might not be translatable to PCL. The objective of this study

was therefore to study the morphology and particle size obtained for non-loaded

electrosprayed PCL microspheres and to ensure their reproducibility. Importantly,

the toxicity of so-produced microspheres was assessed in order to ensure that no

toxic residue remained after electrospraying. These are essential steps to be validated

and understood before progressing to protein loading.


- 98 -

3.4 EXPERIMENTAL SECTION

3.4.1 Materials

Polycaprolactone, Mn = 84 kDa (Perstorp Ltd, UK - Capa® 6500C) was used to

produce the microspheres. Ultra-pure chloroform (99.0-99.4%) from Merck,

Germany was used to dissolve PCL. Different concentrations were prepared: 5, 7.5, 9

and 10% wt/v (i.e. for 10% wt/v, 10 g of PCL were dissolved in chloroform and

made up to 100 mL total volume). The polymer solutions were magnetically stirred

for 3 hours at room temperature to allow complete dissolution before

electrospraying.

3.4.2 Microsphere Production

Figure 3.2A shows a typical schematic of the electrospraying setup used to produce

the microspheres. Initially, the electrospraying parameters chosen in this work were

based on previous studies on optimisation of electrosprayed particles [49, 52, 58, 97,

98, 100, 110, 168, 169]. Temperature and relative humidity ranged from 22 to 24°C,

and 44 to 49% respectively. Collectors were made of standard aluminium foils (20 ×

20 cm2). PCL solutions were loaded in a 2.5 mL glass syringe (Hamilton, USA)

fitted with a 21 or 26-gauge stainless steel nozzle (Terumo, Japan and Becton

Dickinson, USA). PCL solutions were extruded through the nozzle at a constant rate

of either 0.2 or 0.5 mL/h using a syringe pump (WPI, USA). The tip-to-collector

(TTC) distance was set to 15, 20 or 25 cm respectively. High-voltage was applied

between the needle and collector ranging from 10 to 18 kV. After electrospraying,

the collectors were placed under vacuum for a further 72 hours, to remove any

chloroform residue from microspheres. The microspheres were then transferred into

glass vials and further evacuated for storage. The different experimental parameters

employed are summarised in Table 3.1. The microspheres produced under each set of

parameters are abbreviated as M, with a number referring to the condition type (a set

of parameters). For example: for condition 1, the produced microspheres are M1-

type.

Section 3.4 Experimental Section

- 99 -

Figure 3.2. (A) Schematic of a typical electrospraying setup. (B) Picture of the aluminium foil

collected after the electrospraying process. (C) Dried microspheres collected from the aluminium foil

and further placed in a glass vial for storage. (D) Microspheres taken from the glass vial and analysed

on a microscope slide. (E) Picture of the electrospraying process, inside the safety box, comprising the

syringe pump, syringe loaded with polymer solution, collector and electrodes. (F). Picture of the

power supply located outside the safety box.

Table 3.1. Set of parameters tested for each condition.

polymer solution

aluminium foil

power supply

syringe

stainless steel capillary

aluminium plate

A B C

D

E F

Gauge

PCL

Concentration

(% wt/v)

Voltage

(kV)

TTC

Distance

(cm)

Flow

Rate

(mL/h)

Condition 1 (M1) 26 5 10 20 0.5

Condition 2 (M2) 26 7.5 10 20 0.5

Condition 3 (M3) 26 9 10 25 0.5

Condition 4 (M4) 26 10 10 15 0.5

Condition 5 (M5) 26 10 10 20 0.5

Condition 6 (M6) 26 10 10 20 0.2

Condition 7 (M7) 26 10 10 25 0.5

Condition 8 (M8) 21 10 10 25 0.5

Condition 9 (M9) 26 10 10 25 0.2

Condition 10 (M10) 26 10 16 25 0.5

Condition 11 (M11) 26 10 16 25 0.2

Condition 12 (M12) 26 10 18 25 0.5

Condition 13 (M13) 26 10 18 25 0.2


- 100 -

3.4.3 Physical Characterisation

The morphology and microstructure of electrosprayed microspheres were

characterised with a scanning electron microscope (FEI Quanta 200 SEM) operating

at 10 kV. Micrographs were taken from low and high magnifications, in order to

have overviews of batches and detailed morphology of microspheres respectively.

Microspheres were imaged directly on the microscope slides used for particle size

determination, or after collection from the aluminium foil. In the latter case, particles

were carefully deposited on carbon sticky tape, previously mounted on aluminium

stubs. Both microscope slides and stubs were gold sputtered (BIORAD SC-500

Sputter coater) for 75 s at 30 mA before imaging.

In order to determine particle size of electrosprayed microspheres, a microscope

glass slide was introduced in the electrospraying box and held in contact with the

collector, in the centre of the spraying zone for 5 minutes. The slide was then

removed and analysed by light microscopy (AxoVision, Carl Zeiss MicroImaging

GmbH, Germany). In order to assess the reproducibility of electrospraying, 3

replicates of each condition were generated. Voltage was turned off between

replicates. Particle size was assessed with Image J analysis software (NIH) based on

the micrographs. The results were plotted as box plots and expressed in medians,

with n = 100-1,000 for each replicate, whereas size distribution was shown for all

values obtained per condition.

Electrosprayed particles (M3-type) and unprocessed pellets were characterised by

differential scanning calorimetry (TA Instruments Q100 DSC) by scanning from

0°C 110°C 0°C with a heating and cooling rate of 10°C/min. The initial run

was followed by a repeat run with the thermal history erased.

3.4.4 Biological Effect of Microspheres

The effect of electrosprayed microspheres on cells was assessed by two methods: the

extraction method and direct contact method as per ISO 10993-12. The extraction

method, also known as the elution method, required an extract from the material to

be tested. The extract was placed on a near-confluent monolayer of fibroblast cells

and toxicity was evaluated by observing cell numbers using DNA measures. M3-type

microspheres were used for this experiment, UV sterilised for 40 minutes

immediately before the assay. 0.1 and 1% wt/v of microspheres were placed in

completed Dulbecco's modified Eagle medium (DMEM) (10% foetal calf serum, 1%

Section 3.5 Results and Discussion

- 101 -

penicillin/streptomycin) for 1 and 24 h. The extract solutions were further removed,

filtered and seeded on cells. For the direct contact method, 0.01 and 0.1% wt/v M3-

type microspheres were rinsed in media for 15 minutes and 1 hour, before being

placed on the near-confluent monolayer of fibroblast cells, or left non-rinsed but

incubated for the same amount of time (37°C, 5% C02). In both methods, NIH3T3

cells were cultured for 24 h before exposure to the test solutions for another 48 h

(initial seeding density: 3 × 104 cells). DNA quantification was determined using

CyQUANT® (Invitrogen) (n = 4). Results were expressed in normalised averages ±

standard errors (SE).

3.5 RESULTS AND DISCUSSION


3.5.1.1 Morphology

Electrospraying resulted in either microspheres or flattened particles, both with

textured surfaces. The spherical morphology was obtained only for high polymer

concentrations (9 and 10% wt/v), while flattened morphology and coalescence

between particles was observed for decreased polymer concentrations (5 and 7.5%

wt/v) (Figure 3.3). These results are in accordance with previous studies, where

polymer concentration was often shown to be the most critical parameter in the

morphology of electrosprayed particles [98].

The generation of electrosprayed particles is widely accepted to be controlled by

two main mechanisms: solvent evaporation from droplets en route from the tip to the

collector, and contemporaneous polymer diffusion during evaporation [49]. Rapid

polymer diffusion does not necessarily lead to spherical particles but will ensure

solid, dense particles. Both these mechanisms are dictated by the characteristics of

the electrosprayed polymer solution itself, dependent on molecular weight and

polymer concentration. For conditions 1 and 2 (Table 3.1) for instance, the polymer

solutions are only 5 and 7.5% wt/v, respectively and lead to the flattened morphology

shown in Figure 3.3A-B, rather than a spherical morphology, observed at higher

polymer concentrations (9 and 10% wt/v) as shown in Figure 3.3C-D. The flattened

morphology is even more pronounced for the 5% wt/v particles rather than the 7.5%

wt/v particles, indicating that lower polymer concentration favours the formation of

flat particles instead of spheres. These two cases are a direct consequence of


- 102 -

incomplete solvent evaporation, since the solvent contents are higher for decreased

polymer contents. The electrosprayed particles produced for these concentrations are

still partially dissolved when they hit the collector and therefore not fully dried,

leading to inhomogeneous semi-solid, flat particles that would further solidify after

deposition.

Figure 3.3. Influence of polymer concentration on microsphere morphology: (A) M1-type

microspheres (5% wt/v), (B) M2-type microspheres (7.5% wt/v), (C) M3-type microspheres (9%

wt/v), (D) M4-type microspheres (10% wt/v). Scale bar is 20 µm.

Chain entanglements also are important to the physical properties of the particles

produced. At low polymer concentrations, there are less entanglement possibilities

for the polymer chains where the operating regime is known as the semidilute

unentangled regime. In this state, the concentration is large enough for chains to

overlap, but not sufficient to generate a significant degree of entanglement [103]. At

higher concentrations, the same available hydrodynamic volume is occupied by more

polymer chains, introducing chain entanglements. Gupta et al. defined the crossover

of concentration from the semidilute unentangled to semidilute entangled regime as

the critical entanglement concentration, Cent, which marks the distinct onset of

significant chain entanglements in solution. Therefore in the semidilute unentangled

regime: Cov < C < Cent, where C is the polymer concentration and Cov the critical

chain overlap concentration [97, 103]. It is thus essential to use a polymer

concentration > Cent to have the entangled regime.

A B

C D


- 103 -

Reproducibility of electrosprayed particles is a problem when working in the

semidilute unentangled regime where entanglements are less frequent. In order to

obtain reproducible, homogeneous, and solid particles, it is necessary to ensure

complete solvent removal and to use polymer concentrations above Cent. This is

equally important for producing spherical particles as it was shown that if the

evaporating droplets present a sufficiently entangled network before they reach the

Rayleigh limit, the resulting particles will remain monodisperse and spherical, as the

entangled network stabilises the droplet against rupture, reducing the frequency of

smaller offspring particles being emitted [49]. There are some scaling laws to

determine Cent for each type of polymer solution, however it is most likely to be

determined experimentally. In the case of PCL 84 kDa dissolved in chloroform, it

can be deducted from this study that Cent is comprised between 7.5 and 10% wt/v,

where 10% wt/v was sufficient to produce solid particles on the collector, ensuring

homogeneity and sphericity of particles.

The texture observed for spherical particles was previously described as an

‘embossing golf-ball structure of the colloidal surface’ [80]. Another study on the

effect of the solvent properties on electrosprayed polymer particles described how

the morphology can be changed according to the type of solvent used and its

concentration [97]. It was shown that solvents with boiling points > 140°C, like N,N-

dimethylformamide 146°C, or benzaldehyde 178°C, would be able to generate

smooth surfaces during electrospraying. Chloroform has a much lower boiling point

of 61.2°C, explaining the textured surface that was observed. Such texture was also

seen in an even more pronounced way when electrospraying PCL with

dichloromethane [98], which boils at 40°C, corroborating this theory. In the case of

in vivo implantation, it is noted that the topographical complexity of a biomaterial is

preferred for cell attachment since it generates an increased number of anchoring

sites for cells [11]. As a consequence, from this point of view, chloroform can be

considered as an adequate solvent to be used in electrospraying of particles to be

used in vivo.

Figure 3.4 illustrates that some of the electrosprayed microspheres presented a

certain degree of concomitant fibre formation between the particles. This was shown

to occur for the highest polymer concentration (10% wt/v) and was favoured by

lower rates (0.2 versus 0.5 mL/h), indicating that a higher concentration favoured the

formation of fibres. No differences in fibre formation were seen in the 10 to 18 kV


- 104 -

voltage range while flow rate was clearly a determinant in that respect as seen in

column 1 (0.2 mL/h) compared to column 3 (0.5 mL/h) of Figure 3.4. Fibres were

extremely thin, on the nanometre scale, acting as a discrete coating on top of the

spheres. Such nanofibres were almost non-existent for the 0.5 mL/h generated

microspheres, while they were more numerous on the 0.2 mL/h generated

microspheres. This is explained by the high polymer concentration selected here, 10

% wt/v, which ensures a strong entanglement network in the evaporating droplets.

This regime is at the onset of beaded fibre formation and full jet break-up needs to be

achieved by increasing two other variables; current or flow rate. Here, a 0.2 mL/h

flow rate was not sufficient to ensure full break-up on the full 10-18 kV voltage

range, while 0.5 mL/h was sufficient, indicating that flow rate was more determinant

than current. Thus it can be concluded that for the specific combination selected here,

a high polymer concentration needs to be coupled with a higher flow rate for

maintaining fibre-free and spherical particles.

Figure 3.4. Influence of flow rate and voltage on microsphere morphology with a high and low

magnification for each type of microsphere: (A-B) M9-type (C-D) M7-type, (E-F) M11-type, (G-H)

M10-type, (I-J) M13-type, (K-L) M12-type. In the first and third columns, scale bar is 10 µm. In the

second and fourth columns, scale bar is 100 µm.

3.5.1.2 Particle Size

The microspheres produced for conditions 3 and 5-13 had narrow quasi-

monodisperse size distributions with average diameters ranging from 10.64 to 17.80

0.2 mL/h 0.5 mL/h

10 kV

16 kV

18 kV

A B C D

E F G H

I J K L


- 105 -

µm (standard deviations (SD) ranging from 2.05 to 4.93) (Figure 3.5).

Electrospraying was shown to be reproducible for most conditions, with average

particle sizes always comprised within a 3 µm difference. A short tip-to-collector

(TTC) distance, however, did not ensure reproducibility with 15 cm being too close

(condition 4). Distances of 20 cm and 25 cm led to better reproducibility, with a

trend showing a slightly narrower size distribution when using a slower flow rate

(0.2 mL/h instead of 0.5 mL/h), as shown by standard deviations from Figure 3.5D.

In accordance with previous studies, the flow rate and polymer concentration were

the main parameters to tune particle size, while gauge had non-significant effect on

particle size as shown by M7 and M8 sizes in Figure 3.5 (21 G versus 26 G).

However, it was observed that the size distribution was slightly broader for the

bigger gauge with SD = 2.40 for 26 G (M7) and SD = 3.42 for the 21 G (M8)

(internal diameter = 0.241 mm and 0.495 mm, respectively). It is inferred from this

result that a smaller gauge can produce a narrower size distribution.

Figure 3.5. (A) Table of parameters for each electrospraying condition. (B) Average particle size of

each replicate obtained per electrospraying condition (3 replicates per condition) expressed as box

plots (n = 100-1,000). (C) Histograms of size distributions. (D) Average particle size of each

condition expressed as means and standard deviations. (E) Inset showing the average particle size

(means ± standard errors) as a function of the voltage, for 2 different rates (0.2 and 0.5 mL/h).

0

5

10

15

20

25

30

35

40

45

50

R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3

M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 M13

Concentration (%w/v) 9 10 10 10 10 10 10 10 10 10 10Voltage (kV) 10 10 10 10 10 10 10 16 16 18 18TTC distance (cm) 25 15 20 20 25 25 25 25 25 25 25Feed rate (mL/h) 0.5 0.5 0.5 0.2 0.5 0.5 0.2 0.5 0.2 0.5 0.2Gauge 26 26 26 26 26 21 26 26 26 26 26

Number of

particles

(%)

Particle size (µm)

Particle

size (µm)

A

B

C

E

Mean (µm) 17.80 19.82 17.57 12.85 16.82 17.02 17.04 13.09 10.64 13.43 10.95

SD 4.76 7.12 3.42 2.48 2.40 3.42 4.93 2.61 2.31 2.38 2.05

D


- 106 -

The polymer concentration impacts the surface tension of the solution, affecting

particle size, as reported by Hartman et al. in the cone-jet mode of electrospraying

[42] and shown in Equation 3.3 and 3.4:

(

)

(3.3)

( )

(3.4)

where d is the droplet diameter (m), α is a constant, Q is the liquid flow rate (m3/s), ρ

is the solution density, I is the current, ε0 is the permittivity of vacuum, γ is the

surface tension of solution in ambient air, and K is the liquid conductivity. These

equations indicate that particle size increases with decreasing surface tension and

increases for increasing flow rate, which is in accordance with the results presented

in Figure 3.5D. Yet, for the 9-10% wt/v range studied here, very little differences in

size were observed: for similar electrospraying conditions (M3 and M7), PCL

microspheres made from a 9% wt/v solution led to an average diameter of 17.8 µm

(SD = 4.76) versus 16.8 µm (SD = 2.40) for the 10% wt/v solution. For any drug

release application, such a small size difference would likely not lead to dramatic

differences in release profiles.

Interestingly, although voltage is known to have very little effect on particle size

in PLA-based polymers [100], a significant effect on size was observed from 10 kV

to voltages ≥ 16 kV for PCL microspheres. At 16 and 18 kV, no significant

differences in particle size were observed for each flow rate, as shown by the inset in

Figure 3.5, and as expected from the theory, particle size was only decreased at low

flow rates (0.2 mL/h versus 0.5 mL/h). However, at 10 kV, particle size was

significantly larger than at higher voltages, regardless of flow rates. It might be

inferred that voltage has an influence on particle size, as shown by Equation 3.3

where particle size decreases with increasing current. The reason for this decrease is

the presence of fibres as confirmed by the morphology images (Figure 3.4). In fact,

in Figure 3.4A-D, there is no significant increase in fibre formation for a decreased

flow rate at a fixed voltage of 10 kV, and so there is no significant difference in

particle size for that condition. However at 16 and 18 kV the number of fibres is

increased when decreasing the flow rate and the particle size is decreased

accordingly. Therefore it can be concluded that this decrease in particle size is not


- 107 -

only due to the increased voltage, but it is due to fibre formation occurring for a

combination of lower flow rates and higher voltages. As a consequence, the size of

particles was reduced since a fraction of smaller offspring droplets were ejected from

the initial droplet, drawing extruded fibres along. Using higher voltages are therefore

to be used with care, and a sufficient high flow rate should be chosen to compensate

the need of a high voltage.

From the size distribution plots shown in Figure 3.5C, a bimodal character could

be observed for some conditions, made up of a majority of primary droplets and a

small percentage of smaller particles (less than 5%). These smaller droplets are

offspring droplets caused by Coulomb fission [42, 100]. They are easily ejected from

the primary droplet during shrinkage of droplets occurring during evaporation. The

offspring phenomenon was emphasised for decreased polymer concentration

contents as shown by M3-type microspheres compared to M7-type microspheres (9

and 10% wt/v, respectively). This is likely due to decreased entanglement for

decreased polymer contents, which favours the occurrence of offspring droplets.

However, in this context, the frequency of offspring droplets is extremely low for

most electrospraying conditions and thus would have no significant effect on release

profiles in the case of drug loading.

3.5.1.3 Reproducibility of Electrospraying

To assess reproducibility of particle formation, condition 7 was used, 6 weeks apart.

During this six week break the electrospraying setup was used by other researchers

such that all the parameters had to be reset. The average particle size for the three

runs at week zero was 16.82 µm (n = 330, SD = 2.40) (Figure 3.5) while the repeat at

6 weeks had average particle size of 16.16 µm (n = 289, SD = 3.98) with unchanged

morphology. Apart from non-significant increase in size distribution, the two

conditions were identified as identical.

The same reproducibility was not, however, observed for all conditions. For

instance, conditions, where flat particles were observed (conditions 1 and 2),

intrinsically lacked reproducibility based on same-day repeats. These conditions

reflected the semidilute unentangled regime where smaller offspring droplets were

ejected and droplets were not fully dried.

For semidilute entangled regimes, shown for concentrations > 9% wt/v, more

parameters will influence the reproducibility of electrospraying. For instance, a high


- 108 -

TTC distance has to be ensured for complete solvent evaporation. If the TTC

distance is too short, as in condition 4 (TTC = 15 cm), entangled but not fully dry

particles are produced, leading to very broad size distributions as seen in Figure

3.5C. In this study, TTC distances of 20 and 25 cm were shown to be ideal in terms

of reproducibility. The flow rates used were shown to generate reproducible samples

either at 0.2 mL/h or 0.5 mL/h, with the formation of fibres observed for the lowest

flow rate, when working at high polymer concentration (conditions 9, 11 and 13).

Although not ideal in terms of final morphology of microspheres, these conditions

were shown to be reproducible from one replicate to the next.

An interesting observation was that although the change in voltage did not affect

reproducibility in particle size and distribution, several collection points of particles

appeared at high voltage, variable for each run, whereas at low voltage the pattern

reflected one collection point, circular and centred. This may be attributed to the

stable cone-jet mode at low voltage versus the multi cone-jet mode at high voltage as

observed previously [27, 97]. This may be an issue when the collector is a secondary

scaffold, for example when making composite scaffolds.

To conclude, it must be understood that reproducibility can be achieved with

electrospraying, but only for certain parameters intrinsic to each polymer solution.

These parameters have to be determined and optimised first, so that the

reproducibility of the process is ensured. The entanglement regime is the most

important to start with where sufficient entanglements in charged droplets are

necessary to meet reproducibility, while sufficiently high TTC distance is also

important, to ensure full solvent evaporation. At high polymer concentration and

high flow rate are equally important to ensure fibre-free particles.

3.5.1.4 Thermal Characterisation

Differential scanning calorimetry (DSC) was used to ensure that chloroform and

electrospraying process would not lower the crystallisation of PCL, impacting on the

characteristics of electrosprayed particles. The initial DSC run was followed by a

repeat run with the thermal history erased in order to check the polymer itself.

However, the study of the first run of each sample allowed assessing any eventual

polymer discrepancies caused by the electrospraying process or contact with

chloroform.


- 109 -

DSC results, presented in Figure 3.6, show single melting peaks with a maximum

melting temperature (Tm) of unprocessed PCL of 61.2°C and a heat of melting (Hm)

of 63.6 J/g, while Tm = 58.9°C and Hm = 71.9 J/g for electrosprayed microspheres,

translating respectively to 45.6% and 51.5% crystallinity (based on Hm for 100%

crystalline PCL of 139.5 J/g [170]). Tm is shown to be slightly decreased, inferring

that smaller crystallites are formed during electrospraying. The degree of crystallinity

of PCL after the second run, with the thermal history removed was shown to be

35.0% and 38.3% for unprocessed PCL and electrosprayed beads, respectively.

These non-significant values confirmed that no intrinsic changes were made to the

polymer that went into contact with chloroform, and further electrosprayed. The use

of chloroform to dissolve PCL can therefore be validated as an appropriate solvent

for electrospraying PCL solutions.

Figure 3.6. DSC traces of electrosprayed M3-type microspheres and unprocessed PCL pellet, first

runs, exothermic is up.

3.5.2 Biological Effect of Microspheres

Electrospraying remains a process that employs organic solvents. It is therefore

important to ensure that these organic solvents are fully removed after process;

otherwise the electrosprayed particles might be toxic to cells regardless of whether

they comprise FDA-approved polymers. This is an important step which is often

overlooked in many studies, where results are shown for loaded particles directly.

However, when loading expensive growth factors, the risk of induced toxicity by

other components is of concern. For this reason we tested the electrosprayed


- 110 -

microspheres by a standard compatibility assay from the ISO 10993-12, called the

elution or extract dilution assay, to ensure that no toxic residue is released. The

microspheres were eluted into media for different times, which was then removed

and seeded on a near-confluent monolayer of cells. If any chloroform residue, which

is known to be highly toxic to cells, was extracted, the DNA quantification by the

CyQUANT® assay was expected to be lower than the control. For the design of our

experiment, we used 0.1 and 1% wt/v of extract and incubated it for 1 and 24 hours,

to probe for entrapped chloroform. The results are presented in Figure 3.7, which

show no statistical difference between cells cultured with or without addition of

extract solution in terms of DNA content, as stated by a two-way analysis of variance

(ANOVA) analysis (p > 0.05). These preliminary results are a good indication that

no chloroform is entrapped in electrosprayed beads, hence no adverse affect on the

cells, and are in accordance with the DSC results above. However this study should

be repeated for longer incubation times matching the whole course of PCL

degradation.

Figure 3.7. Results of CyQUANT® assay, expressing the DNA quantification after 48 h of exposure

to cells (normalised to control), for different test solutions being incubated for different times before

exposure to cells. 0.1% EX and 1% EX refer to 0.1 and 1% wt/v extracts respectively, obtained from

M3-type microspheres immersed in media, for different times. Results are expressed in normalised

averages ± SE (n = 4).

A direct contact assay from the same ISO 10993-12 was also performed to ensure

that microspheres were physically compatible with cells. The microspheres were

rinsed or left non-rinsed in media for different incubation times so that the effect of

different conditions and times on cell viability could be assessed. Different

90

95

100

105

110

115

120

125

130

Media 0.1% EX 1% EX

DN

A q

uan

tifi

cati

on (

% t

o c

on

tro

l)

Test solutions

Non-Incubated

Incubated 1 h

Incubated 24 h


- 111 -

concentrations of microspheres were used: 0.01% and 0.1% wt/v. It must be noted

that such densities were extremely high relative to cell numbers and the study of cell

morphology after contact with microspheres was not possible with microscopy since

immersed microspheres would settle down on cells and completely cover them.

However, after removal of microspheres, cell DNA contents were not lower than the

controls for any of the condition tested and combinations (rinsed microspheres, non-

rinsed microspheres, 1 h incubation, 24 h incubation), showing that microspheres did

not have any physical adverse effects on cells (Figure 3.8), as determined by a two-

way ANOVA analysis (p > 0.05).

Figure 3.8. Results of CyQUANT® assay, expressing the DNA quantification after 48 h of exposure

to cells (normalised to control), for different test solutions being incubated for different times before

exposure to cells. 0.01% NR and 0.1% NR refer to the 0.01 and 0.1% wt/v M3-type microspheres

rinsed in media before exposure. 0.01% R and 0.1% R wt/v refer to the 0.01 and 0.1% wt/v M3-type

microspheres non-rinsed in media before exposure, but incubated for the same amount of time as

rinsed microspheres. Results are expressed in normalised averages ± SE (n = 4).

3.6 CONCLUSIONS

Electrospraying was shown to be a reproducible method for generating spherical

PCL particles with narrow quasi-monodispere size distributions, with average sizes

ranging from 10 to 20 µm, which could be tuned with electrospraying flow rate and

polymer concentration. Control of particle morphology was shown to be tailored

with these same variables by determining the polymer entanglement regime taking

place in the course of electrospraying. In order to avoid fibre formation and offspring

85

90

95

100

105

110

115

120

125

DN

A q

uan

tifi

cati

on

(%

to

co

ntr

ol)

Test solutions

Control

Non-Incubated

Incubated 15 min

Incubated 1 h

Incubation time of

test solutions before

exposure to cells

(37 C)


- 112 -

droplets, an increased polymer concentration must be coupled with an increased flow

rate, ensuring electrospraying in the semidilute entangled regime, which leads to the

formation of spherical, homogeneous and reproducible particles. Chloroform was

shown to be an appropriate solvent for PCL particles, conferring a reproducible

embossed texture to the electrosprayed microspheres, potentially beneficial for cell

adherence. Chloroform did not act as a plasticiser in contact with PCL and was

inferred to be fully removed after drying. Furthermore, electrosprayed microspheres

showed no adverse effects on cell viability after 48 h exposure. In conclusion, this

study has demonstrated precise control over polymer microsphere characteristics

which may be used as a template for future microsphere-growth factor delivery

systems.


T. R. D. acknowledges the Queensland Smart State Fellowship Scheme and Tissue

Therapies Ltd for financial support. M. A. W. is supported by the QUT Vice

Chancellor’s Fellowship Scheme and an ARC Linkage Grant (LP100200084).

Thanks to the ARC (Discovery grant no. DP0989000) for financial support.

3.8 REFERENCES AND NOTES

© 2010 by the authors; licensee MDPI, Basel, Switzerland. This article is an open-

access article distributed under the terms and conditions of the Creative Commons

Attribution license (http://creativecommons.org/licenses/by/3.0/).

- 113 -

Chapter 4: Controlling Microencapsulation and

Release of Micronised Proteins using

Poly(Ethylene Glycol) and

Electrospraying

Nathalie Bock1,2,3

, Tim R. Dargaville1, Maria A. Woodruff

2

Published in the European Journal of Pharmaceutics and Biopharmaceutics,

Volume 87, Issue 2, 2014, Pages 366-377.

© 2014 Elsevier B.V. All rights reserved.








Assisted in reviewing the manuscript


Assisted in reviewing the manuscript






Chapter 4 Microencapsulation of Proteins with PEG and Electrospraying

- 114 -













requirements.





- 115 -

4.1 ABSTRACT

The fabrication of tailored microparticles for delivery of therapeutics is a challenge

relying upon a complex interplay between processing parameters and materials

properties. The emerging use of electrospraying allows better tailoring of particle

morphologies and sizes than current techniques, critical to reproducible release

profiles. While dry encapsulation of proteins is essential for the release of active

therapeutics from microparticles, it is currently uncharacterised in electrospraying.

To this end, poly(ethylene glycol) (PEG) was assessed as a micronising and

solubilising agent for dry protein encapsulation and release from electrosprayed

particles made from polycaprolactone (PCL). The physical effect of PEG in protein-

loaded poly(lactic-co-glycolic acid) (PLGA) particles was also studied, for

comparison. The addition of 5-15% wt PEG 6 kDa or 35 kDa resulted in reduced

particle sizes and broadened distributions, which could be improved by tailoring the

electrospraying processing parameters, namely by reducing polymer concentration

and increasing flow rate. Upon micronisation, protein particle size was reduced to the

micrometer domain, resulting in homogenous encapsulation in electrosprayed PCL

microparticles. Microparticle size distributions were shown to be the most

determinant factor for protein release by diffusion, and allowed specific control of

release patterns.

4.2 KEYWORDS

Electrospraying, drug delivery, encapsulation, in vitro release, microparticles,

micronisation, polycaprolactone, poly(ethylene glycol), protein.

4.3 INTRODUCTION

Conventional intravenous delivery of therapeutics suffers from excessive dosage

being administered and poor bio-availability [11]. In response, the encapsulation of

drugs and proteins in carriers has drawn the attention of pharmaceutical research for

many decades, yet the systems developed are far from optimal. A resorbable carrier,

such as a polymer matrix, has the potential to efficiently protect therapeutic

molecules after administration, while providing sustained delivery upon matrix

degradation [125]. However, the processes of encapsulation involve organic solvents,


- 116 -

shear forces and hydrophobic polymers which may partly denature hydrophilic

protein molecules and affect release kinetics [171]. Dry encapsulation of proteins is a

strategy to minimise protein denaturation by avoiding water-in-oil interfaces used in

traditional techniques, well-acknowledged to denature proteins [35]. While the dry

protein is in contact with organic solvent during processing, the solvent provides

limited molecular mobility for the protein due to the anhydrous environment.

Although the native state of a protein is not favoured thermodynamically in such

solvents, this combination provides a kinetic trap for the protein, which maintains

protein activity [34].

Size reduction of protein aggregates, or micronisation, is critical to ensure high,

homogeneous, and dispersed protein encapsulation in polymeric carriers upon

solvent removal, and represents an intricate challenge in the current protein delivery

systems. Techniques such as spray-drying and ultrasonic atomization have been used

to obtain fine protein particles, but low yields make the techniques unpractical while

harsh stresses (mechanical, heat) may lead to protein denaturation issues [171]. An

easier process was developed by Morita et al., involving co-lyophilisation of the

protein with poly(ethylene glycol) (PEG) [172]. Lyophilisation in the pharmaceutical

field has been subjected to ongoing development and is well-known as an approach

to overcome the physical and chemical instabilities of protein molecules [173]. When

co-lyophilising PEG and a protein solution, PEG effectively raises the energy barrier

for protein molecules to extend their hydrophobic domains to each other, resulting in

reduced aggregation with a progressive shrinkage of protein particles [171, 174].

Extensive protein conformational changes are also prevented by PEG coating at the

surface of proteins, improving both protein stability and delivery capacity [175, 176].

Various proteins have been successfully micronised into spherical microparticles

with diameters less than 5 µm, including serum albumin (SA), superoxide dismutase

(SOD), horseradish peroxidase (HRP) and gelatin, according to different

PEG:protein ratios [127, 172, 177]. Protein aggregate size was found to decrease

linearly with the increase in PEG:protein weight ratio and a critical value was found,

upon which the size decrease was slower. Importantly, the micronisation process did

not alter the protein and full activity was recovered in the case of SOD [127].

Efficient size reduction, less than 1 µm in diameter, improved the encapsulation

efficiency (EE) of active HRP from 24% to 87% [127], and reduced burst release of

SA [178] using solid/emulsion-based microencapsulation techniques. It was also


- 117 -

shown that proteins and PEG could form stable nano-sized complexes in polar

organic solvents by non-covalent interactions, allowing for homogeneous and high

EE when dispersed in a PLGA solution and spray-dried [179].

Bioactivity and release profiles can also be efficiently tailored by the use of PEG

in formulations since the presence of a hydrophilic additive in a hydrophobic

polymer matrix increases diffusion of the encapsulated protein by increasing the

degree of pores in the matrix whilst increasing transport of acidic degradation

products away from the matrix [40, 171, 180]. In a study by Jiang et al., different

concentrations of PEG were indeed shown to affect the release of SA from

polylactide (PLA) microspheres, with increased release rates for 20% of PEG present

in the PLA matrix, compared to 0-10%, but similar profiles were obtained when

comparing the inclusions of PEG 10 kDa and 20 kDa [180]. Protein particle size is

also critical in directing release profiles with larger protein particles leading to burst

release profiles, due to a reduced diffusion of larger protein particles inside a

polymer droplet, resulting in an increased protein concentration near the surface of

polymeric particles [52]. For micronised proteins, a homogeneous and fine

distribution within the particle allows thorough water intrusion, leading to a dense

pore network upon release. Such a feature is essential in enabling sustained,

reproducible and complete release, although it is currently under-assessed [119].

While several techniques allow solid encapsulation, the emerging technique of

electrospraying, in particular, may be highly suited for the efficient encapsulation of

therapeutics in polymeric particles [181]. In electrospraying, the protein may directly

be dispersed in the polymer solution, which, following subjection to high voltage,

results in the extrusion of loaded droplets from a syringe. Droplets undergo solvent

evaporation and can be collected, dry, from a conductive substrate. This simple

process does not require heat or a sophisticated setup, but involves a complex

interplay between processing parameters and polymer solution properties.

Nevertheless, our previous reports have shown that a tight control of particle size and

morphology can be obtained [83, 182], which in combination with dry encapsulation

may be suitable in ensuring reproducible release profiles and active proteins being

released. However, no reports to date, have mentioned or addressed protein

micronisation prior to electrospraying, which is critical for homogeneous

encapsulation of proteins in polymeric particles [181]. Hence, it is hypothesised in

this study that PEG may be used as a micronising agent and a means of tailoring


- 118 -

release from electrosprayed particles. A model protein, SA, will be micronised by co-

lyophilisation with PEG prior to electrospraying and the effect of PEG in the final

particle formulation comprising polycaprolactone (PCL) and PLGA 85:15, both

FDA-approved polyesters suitable for sustained delivery systems, will be assessed in

terms of miscibility of polymers, particle microstructure, protein encapsulation

efficiency and protein release.


4.4.1 Materials

Polycaprolactone (Mn = 84 kDa, PDI 1.53) was obtained from Perstorp Ltd, UK.

Poly(lactic-co-glycolic acid) with a lactide:glycolide (L:G) ratio of 85:15 (Mn = 41.3

kDa, PDI 1.6) was purchased from Evonik Industries, USA. Poly(ethylene glycol)

with Mn = 6 kDa and Mn = 35 kDa, referred hereafter as PEG 6k and PEG 35k,

respectively, dichloromethane (DCM), sodium dodecyl sulphate (SDS), serum

albumin (SA) and fluorescein isothiocyanate (FITC)-conjugated SA were purchased

from Sigma-Aldrich, Australia. Chloroform was purchased from Merck, Germany.

4.4.2 Particle Fabrication

4.4.2.1 Solid Dispersion of Dry Protein

First, the protein was micronised [172]. Briefly, a series of solutions containing the

protein (SA or FITC-SA) mixed with PEG (6k or 35k) were freeze-dried. Various

polymer solutions made of PCL or PLGA 85:15 were prepared in chloroform or

DCM and subsequently added to the protein:PEG lyophilisate under magnetic

stirring (see Table 4.1 for details of constituents and ratios). The resultant dispersions

were vortexed for 10 s (after addition of 1 mL of polymer solution and ultimately

probe sonicated for 1 min at 0.5 W (continuous regime, Misonix 3,000, USA) to

ensure protein dispersion in the organic solvent.

4.4.2.2 Electrospraying

Electrospraying was used to produce dried microparticles encapsulating the proteins.

Ambient temperature and relative humidity ranged from 23 to 24°C, and 34 to 49%,

respectively. The polymer dispersions were loaded in a 1 mL glass syringe

(Hamilton, USA) and extruded through stainless steel nozzles ranging from 26 to 21

G (Terumo, Japan and Becton Dickinson, USA) at constant rates ranging from 0.5


- 119 -

mL/h to 3 mL/h (see Table 4.1 for details of parameters) using a syringe pump

(World Precision Instruments, USA). A voltage of 10 kV was applied to the needle

tip. The tip-to-collector (TTC) distance was either 15 or 25 cm. Collectors consisted

of standard aluminium foils (20 × 20 cm2) (General purpose, Bulls Eye Food

Services) washed with 70% ethanol. After electrospraying, the collectors were placed

under vacuum for a further 72 hours, to remove any solvent residue. The dry

microparticles were then transferred into glass vials and stored at -18°C until further

analysis.

Table 4.1. Summary of formulations. Applied voltage was 10 kV.

Polyester PEG Polyester:

PEG ratio Solvent

Polymer

% wt/v Protein

Protein

% w/w

Protein:

PEG ratio

Flow

rate

(mL/h)

TTC

(cm)

PCL - - Chloroform 10 - - - 0.5 25

PCL PEG 6k 90:10 Chloroform 5 - - 1:10 0.5 25

PCL PEG 6k 90:10 Chloroform 6 SA 5 1:2 1 25

PCL PEG 6k 90:10 Chloroform 6 SA 1 1:10 0.5 25

PCL PEG 6k 90:10 Chloroform 6 FITC-SA 1 1:10 0.5 25

PCL PEG 35k 90:10 Chloroform 9 - - 1:10 3 25


PCL PEG 35k 90:10 Chloroform 9 SA 1 1:10 3 25


PCL PEG 35k 90:10 Chloroform 9 FITC-SA 1 1:10 3 25

PCL PEG 35k 90:10 Chloroform 9 FITC-SA 3 1:3 3 25

PLGA 85:15 - - Dichloromethane 10 FITC-SA 1 - 0.8 15

PLGA 85:15 PEG 35k 90:10 Dichloromethane 10 FITC-SA 1 1:10 0.5 15

PLGA 85:15 PEG 35k 90:10 Chloroform 11 FITC-SA 1 1:10 0.5 15


- 121 -


4.4.3.1 Proteins

The dispersion of proteins within electrosprayed particles was assessed using

albumin labelled with a florescent dye (FITC). The distribution of FITC-SA within

microparticles was visualised using confocal laser scanning microscopy (CLSM). To

randomly collect FITC-SA-loaded particles, a microscope glass slide was introduced

in the electrospraying apparatus housing and held in contact with the collector, in the

centre of the spraying zone for 5 minutes, while electrospraying. The slide was then

removed and fluorescence images were captured using a Leica TCS SP5 confocal

laser scanning microscope (Leica Microsystems, Wetzlar, Germany), 63× objective

with a 9.7× zoom. Excitation was 488 nm and emission was captured between 495

nm and 633 nm.

4.4.3.2 Electrosprayed Microparticles

Particle morphology was characterised with a FEI Quanta 200 scanning electron

microscope (SEM) operating at 10 kV in high vacuum mode. Microparticles were

gently taped on aluminium stubs and gold coated for 225 s at 30 mA (SC500 sputter

coater, Bio-Rad, Australia). Particle size was assessed with ImageJ analysis software

(National Health Institutes (NIH)) based on light micrographs (AxoVision, Carl

Zeiss MicroImaging GmbH, Germany). Results were plotted as box plots and

expressed in medians with n = 330-570 particles per formulation.

4.4.4 In Vitro Characterisation

4.4.4.1 Encapsulation Efficiency

Protein content in the microparticles was determined using two extraction

procedures. For the first extraction (referred as EX1), particles (8 mg) were dissolved

in DCM (2 mL), n = 3. Phosphate buffer saline (PBS) (3 mL) was then added to the

dispersions and tubes were vortexed for 2 min to extract SA (Vortex Mixer SA3,

Stuart Scientific). The resultant emulsions were centrifuged at 5,000 rpm for 15 min

to separate the aqueous phase containing SA from the organic phase containing

dissolved polymer. The aqueous phase was collected and another extraction cycle

was performed to maximize SA recovery. The collected aqueous phase was analysed

by the micro-bicinchoninic acid (µBCA) assay (Thermo Fisher Scientific, Australia)

using a standard curve prepared by serial dilutions of the supplied SA from 40 to 0

Chapter 4 Encapsulation of Proteins in Electrosprayed Microparticles

- 122 -

µg/mL. A polynomial fit was deducted from the corresponding absorbance readings

at 562 nm (R2 = 0.9991) (Microplate Manager V5.2, Benchmark Plus

spectrophotometer, Bio-Rad, USA). When determining the EE of FITC-SA loaded

microparticles, a separate calibration curve was similarly prepared with FITC-SA to

ensure no interferences with the µBCA assay (absorbance read at λ = 562 nm). The

calibration curve was the same as normal SA, within the linear range up to 20

µg/mL, above which slightly reduced absorbance values were detected.

Encapsulation efficiency (EE) (%) was measured according to EE = Measured SA

content (µg) / Theoretical SA content (µg) × 100. The second extraction procedure

(referred as EX2) was similar than the first one, except that particles (10 mg) were

extracted with PBS (4 mL) supplemented with 5 mM of SDS, n = 5. Only one

extraction cycle was performed with no centrifugation. Protein was quantified using

the µBCA assay.

4.4.4.2 In Vitro Release

Particles (10 mg) were placed in 2 mL screw-capped microtubes (Sarstedt, Germany)

and filled with PBS (1.5 mL) containing 0.02% wt/v sodium azide. Tubes were

agitated at a speed of 8 rpm at 37°C for 81 days. At specific time points, tubes were

centrifuged at 28,000g for 2 min to settle particles before 1 mL of supernatant was

collected and replaced by the same amount of fresh PBS. The supernatant of both

RS1 and RS2 was analysed by the µBCA assay using the same technique described

in the encapsulation efficiency section.



4.5.1.1 Miscibility Considerations with Hansen Solubility Parameter

Any given mixture is governed by a combination of thermodynamics, kinetics and

evaporation processes. Miscibility, in particular, is very important in electrospraying

since immiscible substances may lead to electrospraying jet instabilities, resulting in

poor chain entanglements and non-spherical/irreproducible morphologies. Hence, in

this section, thermodynamics insights will be considered in regards to the miscibility

of PLGA and PCL with PEG, and with possible solvents used in electrospraying, in

order to determine a suitable combination for optimal protein encapsulation in

electrosprayed particles.


- 123 -

4.5.1.1.1 Definitions

In drug-polymer systems, local interdiffusion is possible and can be expressed by

Flory-Huggins theory [183, 184], and the Hansen solubility parameter (HSP) of a

substance [185, 186]. HSP can be divided into δD, δP and δH, representative of the

non-polar or dispersion (D) forces (such as van der Waals interactions), the polar (P)

forces and the hydrogen (H) bonding nature of species [185] (Equation 4.1):

(4.1)

The closer the parameters of two substances are, the better the miscibility of one

substance in the other [187], and hence solubility parameters are useful in

determining the miscibility of drugs in polymers in drug-polymer binary systems.

For instance Mastumoto et al. used a variant of HSP to predict the presence of

cisplatin and other drugs in the PLGA or PLA phase of blended PLGA:PLA

microspheres with various solvents [188].

The limitation of the solubility parameter and Flory-Huggins theory is that they

cannot be applied to complex structures such as proteins. While proteins can be

characterised by their chain conformation like polymers, they adopt specific native

conformations under different conditions, hence resulting in different atomic and

molecular forces [189]. When mixed in an organic solvent, hydrogen bonding forces

would be significantly affected, and results in even more complex calculations,

although extended Hansen regression models can be used [190]. Gander et al.

showed that the solubility parameters were indeed insufficient for predicting

microsphere properties of spray-dried PLA encapsulating serum albumin, but that

extending the solubility parameters to electrostatic and covalent considerations was

more powerful [191]. Due to the complexities of protein tertiary structure, we have

thus chosen here to use HSP to determine an ideal polymer-solvent combination that

excludes the protein.

The Hansen partial solubilities of a substance can be identified as 3D coordinates

in the ‘Hansen solubility space’, representing the centre of a solubility sphere which

includes the solvents and excludes the non-solvents. While the radius of the sphere

R0 needs to be determined experimentally, R0 of many polymers have been

characterised and can be found in the literature and software databases [185, 186].

Hence, in order to define the miscibility of a polymer (p) in any solvent (s), the


- 124 -

‘distance’, Ra, between two materials can be plotted in the Hansen solubility space,

and is expressed in Equation 4.2 [185]:

( )

( )

( )

(4.2)

Materials with similar solubilities will have a small Ra and hence will be more

miscible. Although Ra can be thought as a physical distance in the Hansen solubility

space, it represents energy and is expressed in (J/cm3)1/2

. By dividing Ra by Ro, the

relative energy difference (RED) number can be defined as in Equation 4.3:

(4.3)

A RED number less than 1 indicates miscibility while a number higher than 1

indicate immiscibility. In general, RED numbers progressively higher indicate lower

affinities.

4.5.1.1.2 Applications of Hansen Solubility Parameter

The Hansen partial solubility parameters can be calculated by the group contribution

(GC) method, however this calculation does not take into account the molecular

weight of a polymer, which has been shown to affect to a certain extent the

parameters [192]. As a result, partial solubility parameters can also be assessed

experimentally, by swelling tests, turbidimetric titrations, viscosity measurements

and inverse gas chromatography [192]. While results from swelling are generally in

good agreement with the GC method, δP values obtained by titration methods can be

low and unrealistic [193].

Here, PLGA and PCL were assessed in association with PEG and the possible

solvents used in electrospraying. Table 4.2 summarises the partial solubility

parameters of these materials found in the literature with the closest molecular

weights and L:G ratio (for PLGA) of the polymers used in this study.


- 125 -

Table 4.2. Solubility parameters of selected polymers and solvents obtained by different methods,

expressed in √ .

δD δP δH R0 Method Ref

PCL 17 4.8 8.3 - GC [194]

PCL 14 kDa 17.8 6.1 7.8 7.1 Swelling [194]

PCL 65 kDa 17.8 6.2 7.7 5.5 Swelling [194]

PLGA 85:15 16 9.3 11.4 - GC [193]

PLGA 85:15 75 kDa 17.4 8.3 9.9 8 Swelling [193]

PEO, PEG 17 10 5 8 HSP

database [186]

PEG 17.8 11.1 9.1 - GC [195]

Chloroform 17.8 3.1 5.7 - GC [185]

Dichloromethane 18.2 6.3 6.1 - GC [185]

Ethanol 15.8 8.8 19.4 - GC [185]

Acetone 15.5 10.4 7 - GC [185]

N,N-dimethylformamide 17.4 13.7 11.3 - GC [185]

When plotting a 2D projection of the solubility sphere of the three polymers of

interest and possible solvents used in electrospraying (Figure 4.1), both chloroform

and dichloromethane (DCM) are within all the spheres of PLGA, PCL and PEG.

None of the polymers are miscible with ethanol, but while PLGA and PEG are

miscible with acetone and N,N-dimethylformamide (DMF), it is not the case of PCL.

RED values are presented in Table 4.3, with RED values larger than 1 indicating

immiscibility. From the solvents studied here, RED was shown to be the smallest for

chloroform and DCM.

The association of PEG to PLGA provided similar compatibility to PLGA alone,

but an increase of Ra was observed with the addition of PEG to PCL compared to

PCL alone (Figure 4.2). In all cases these values are low, indicating that the blending

of PEG to PCL or PLGA with both chloroform and dichloromethane did not

dramatically, or not at all, decrease miscibility, and as such they can be considered

good solvents for those blends and will be used hereafter.


- 126 -

Figure 4.1. 2D projections of the 3D solubility coordinates of various polymers and their radius of

interaction R0 (R0 = 5.5, 8, 8 √ for PCL, PLGA and PEG, respectively [186, 193, 194]) and

various solvents. To account for the invisible third axis (δD), open/closed symbols have been used;

open symbols indicate that the solvent is outside the sphere (immiscible), closed symbols indicate that

the solvent is inside the sphere (miscible).

Table 4.3. RED values measured from Equations 4.2 and 4.3. R0 = 5.5, 8, 8 √ for PCL, PLGA

and PEG, respectively [186, 193, 194]. These values are from polymers which molecular weights are

closest to the polymers selected in this study.

PCL PLGA 85:15 PEG

Chloroform 0.67 0.84 0.89

Dichloromethane 0.33 0.57 0.57

Ethanol 2.30 1.25 1.83

Acetone 1.14 0.65 0.45

N,N-Dimethylformamide 1.52 0.70 0.92

Figure 4.2. Distance Ra (√ ) of various polymers and blends with 10% wt PEG with chloroform

and dichloromethane, obtained from Equation 4.2.

It must be noted that the solubility values used here for PEG were calculated by

the GC method, which does not account for molecular weight. However in the

present study we are assessing two molecular weights of PEG, namely 6 kDa and 35

kDa. While no experimental values could be found for these polymers at room

0

5

10

15

20

0 5 10 15 20

δH

δP

ChloroformDCM

Acetone

DMF

Ethanol

0

5

10

15

20

0 5 10 15 20

δH

δP

Acetone

DMF

Ethanol

0

5

10

15

20

0 5 10 15 20

δH

δP

Acetone

DMF

Ethanol

PCL

PLGA

85:15

PEGChloroformDCM

ChloroformDCM

0

2

4

6

8

PCL PCL:PEG

PLGA PLGA:PEG

Chloroform

Dichloromethane


- 127 -

temperature, Adamska et al. measured several MW PEGs (from 2 to 35 kDa) by

inverse gas chromatography at high temperatures above melting points, and while

differences in solubility parameters were observed [196], no clear trends was

observed on this range, making it safe to assume that no dramatic changes in

miscibility would be observed between PEG 6k and PEG 35k at room temperature.

4.5.1.2 The Effect of PEG on Electrosprayed Particles

While the previous section addressed the thermodynamics of a polymer-solvent

system, electrospraying is a technique where polymer droplets in solution undergo

full evaporation in milliseconds [75]. Hence in this context, evaporation processes

are an important contribution in directing the final characteristics of dry

electrosprayed microparticles. Briefly, as the solvent evaporates, two competing

effects occur: polymer concentration increases and entanglements commence, which

stabilise the droplet from further subdivision while surface charge increases at the

same time, driving droplet subdivision when droplets are not sufficiently entangled

[102]. Chain entanglements are thus a critical factor in electrospraying and are in part

responsible for the final morphology of electrosprayed particles [103, 181]. Solution

properties, including polymer concentration and molecular weight, significantly

affect the particle/fibre formation in respect to other important parameters, such as

surface tension and conductivity [102].

Here the effect of blending PCL with two different molecular weights of PEG was

studied. A 10% wt/v PCL solution in chloroform (Mn of 84 kDa) (Figure 4.3A) was

compared with a PCL:PEG solution with identical final polymer concentration but

containing 10% wt of PEG 35k (Figure 4.3B) or PEG 6k (Figure 4.3C). In both

cases, the size distribution was strongly influenced with reduced sizes and

polydispersity, and a beaded-fibre morphology was obtained with PEG 6K (Figure

4.3D). This is explained by the PCL:PEG blend generating an increased degree of

instabilities during the electrospraying process, in turn increasing the driving force of

droplet subdivision, leading to the reduction in particle sizes. While a large

molecular weight distribution (MWD) has been shown to lower chain entanglement

density [103], the presence of a bimodal MWD was particularly unfavourable here.

Since a minimum of 10% wt PEG was necessary for studying changes in physical

and in vitro properties, the change of overall concentration remained the most

judicious choice for ensuring reproducible and spherical electrosprayed particles.


- 128 -

Hence for PEG 6k, the overall concentration was reduced in half (5% wt) and

optimal spherical morphology was recovered (Figure 4.3E). As expected when

reducing concentration, smaller beads were also obtained [181].

In the case of PEG 35k, the initial 10% wt/v polymer concentration was shown to

be efficient in producing fibre-free particles, and although size was reduced, no

similar adverse effects as for PEG 6k were observed on morphology (Figure 4.3B).

This result was due to the larger MWD of the PCL:PEG 35k blend compared to PCL

alone, but smaller than the PCL:PEG 6k blend, thus less detrimental. In this case, in

order to increase particle size, which facilitates the encapsulation of large molecules

such as proteins, the increase of the electrospraying flow rate was the best option.

Hence, by increasing the flow rate 6-fold, larger and reproducible microparticles

were obtained (Figure 4.3D).

Figure 4.3. The effect of PEG incorporation to a PCL solution in chloroform. Transmitted light

microscopy images of particles produced with electrospraying parameters: applied voltage (AV) = 10

kV, TTC = 25 cm, and flow rate (FR) = 0.5 mL/h for a, b, c, e, and 3 mL/h for d.

+ 10% wt PEG 35k

1-fold decrease in

overall concentration

6-fold increase in

flow rate

40 µm

A

40 µm

C

40 µm

D

40 µm

E

+ 10% wt PEG 6k

10% wt/v PCL

40 µm

B


- 129 -

In summary, the addition of PEG to a PCL polymer solution led to instabilities in

electrospraying, in turn generating non-ideal electrosprayed particles, to a greater

extent for bimodal and larger MWD. Nevertheless, the detrimental effect generated

by the addition of a dissimilar polymer in the electrospraying blend could be

counterbalanced by appropriate changes in polymer concentration and

electrospraying flow rate, critical parameters in tailoring particle size and

morphology in electrospraying [83, 181].

4.5.1.3 The Effect of PEG on Protein Encapsulation

The encapsulation of fluorescent FITC-SA enabled the visualisation of the protein

inside the particles upon electrospraying. Figure 4.4 and Figure 4.5 show the results

of encapsulation of 1% wt FITC-SA in either PLGA or PCL-based particles. When

PEG was used as a micronising agent for FITC-SA, a protein:PEG ratio of 1:10 was

selected to ensure size reduction of the protein particle within the micrometer order

[172].

Figure 4.4. 1% wt FITC-SA loaded electrosprayed PLGA particles upon; (A) no micronisation, (B-C)

protein micronisation with PEG 35k. Polymer concentration ranged from 10 to 11% wt/v,

electrospraying parameters were: FR = 0.5-0.8 mL/h, AV = 10 kV, TTC = 15 cm.

When the protein particles were not micronised, they presented large sizes,

resulting in non-encapsulation in most of the produced microparticles (Figure 4.4A).

Following protein micronisation with PEG and addition of the FITC-SA:PEG

lyophilisate to the PLGA solution, the resulting electrosprayed particles presented

shapeless and irregular particle morphologies (Figure 4.4B). While it was difficult to

obtain spherical particles by tuning the processing parameters, it was eventually

obtained by changing the solvent, dichloromethane, to chloroform (Figure 4.4C).

The physical properties of an electrospraying solution include surface tension,

vapour pressure, electrical conductivity and dielectric constant, which all play a role

in the final morphology and structure of electrosprayed particles. Table 4.4

20 µm

A) PLGA DCM B) PLGA + PEG 35k DCM C) PLGA + PEG 35k Chloroform

20 µm 20 µm

FITC-SA


- 130 -

summarises some of these properties for dichloromethane and chloroform. Here, the

more spherical morphologies of PLGA particles in chloroform were attributed to

chloroform having a lower vapour pressure compared to dichloromethane, (21 kPa vs

47 kPa, respectively), which enhances polymer diffusion in electrosprayed droplets,

and thus sphericity [181]. A high dielectric constant and electrical conductivity also

correlate with high droplet surface charge, resulting in increased Coulombic

repulsion, undesirable in electrospraying [100]. Considering those parameters from

Table 4.4 and in agreement with the experimental results, it can be concluded that

chloroform represented here a better solvent, since it has a lower vapour pressure

than dichloromethane, more favourable for polymer diffusion, and lower dielectric

constant and electrical conductivity, limiting Coulombic repulsion.

Table 4.4. Physical properties of solvents, relevant in the context of electrospraying [100, 108].

Chloroform Dichloromethane

Formula CHCl3 CH2Cl2

Density (g/cm3) 1.48 1.33

Boiling point (°C) 61.2 39.6

Vapour pressure (kPa) 21 47

Dielectric constant 4.8 9.1

Electrical conductivity (S/m) < 1×10−10

2.75×10−8

In terms of protein particle size, protein reduction was evident upon micronisation

with PEG, as shown in both Figure 4.4 and Figure 4.5 with smaller domains of

fluorescent protein within either PLGA or PCL particles within the micrometer scale.

Proteins were encapsulated in all polymeric particles, assuming microparticle sizes

above 8 µm, while less homogenous encapsulation was observed for smaller particles

(Figure 4.4C and Figure 4.5B). Here, reproducible spherical morphologies for the

PLGA formulation (electrosprayed with 10 wt% PEG 35k and 1 wt% protein in

chloroform) were obtained only for sizes below 8 µm. Indeed, the average size of the

PLGA microparticles was 5.6 ± 0.8 µm for flow rate (FR) = 0.5 mL/h, and 7.1 ± 1.7

µm for FR = 1 mL/h, but for higher FR, poor reproducibility and irregular

morphology was observed, which impaired further size increase. Conversely, the size

increase of PCL particles was easily achieved, while maintaining reproducible

spherical morphologies, by using higher molecular weight PEG and increasing

polymer concentration and flow rate. Particle size increased from 4.9 ± 0.5 µm to

12.0 ± 4.0 µm, allowing homogenous encapsulation of micronised protein in all

polymeric particles (Figure 4.4D). Most importantly, CLSM allowed optical slicing


- 131 -

of the microparticles, confirming that the protein was actually entrapped within the

particle and not sitting on the particle surface (Figure 4.5F). The inability to increase

the sizes of PLGA microparticles was attributed to the lower glass transition

temperature (Tg) of PLGA compared to PCL (40-50 °C for PLGA 85:15 and -60 °C

for PCL). Hence when electrospraying, which was undertaken at room temperature,

PLGA was in its vitreous state (below Tg) while PCL was in its rubbery state (above

Tg), allowing for reorganisation of chains during the electrospraying process, which

happens even in the presence of minute amounts of solvents for a polymer in his

rubbery state. This characteristic enabled a larger range of particles to be produced in

the case of PCL particles, paving the way for more particle tuning compared to a

polymer like PLGA that is in its vitreous state at room temperature. Such state

allowed less chain reorganisation during electrospraying, hence hindering PLGA

particle size increase, in turn providing non-homogeneous encapsulation which limits

reproducibility of release profiles [181].


- 132 -

Figure 4.5. Microscopy images of 1% wt FITC-SA loaded electrosprayed PCL particles upon; (A-B)

protein micronisation with PEG 6k (6% wt/v in chloroform, FR = 0.5 mL/h, AV = 10 kV, TTC = 25

cm) (C-D) protein micronisation with PEG 35k at low magnification and (E-F) high magnification

(9% wt/v in chloroform, FR = 3 mL/h, AV = 10 kV, TTC = 25 cm). (F) Maximum projection of

CLSM z-stack of 57 images (step size = 0.21 µm).

Protein particle size reduction was similar for both molecular weights of PEG.

This is in agreement with Morita’s work, where PEG 6, 20 and 40 kDa provided

similar size reductions of SA [172]. This property confers a higher degree of freedom

when tailoring the size of loaded electrosprayed particles; smaller microparticles can

be obtained by decreasing the molecular weight of PEG used during the

micronisation step, without incidence on protein size after micronisation. However,

microparticle size must be addressed so that size is sufficiently large to allow

encapsulation of the micronised protein.

4.5.1.4 The Effect of PEG and Protein on Particle Microstructure

Electrospraying with PLGA:PEG was possible to a certain extent, however it was

clearly less tailorable compared to the PCL:PEG combination, mostly due to the

lower PLGA Tg, hence reducing opportunities for loading and release. Since PCL

Transmitted Light Microscopy Fluorescence Light Microscopy

PC

L:P

EG

6k

PC

L:P

EG

35

k

30 µm

PC

L:P

EG

35k

30 µm

30 µm 30 µm

14.6 µm

A B

C D

E F


- 133 -

presented enhanced microparticle characteristics compared to PLGA, we focused the

rest of the study on the effect of PEG on encapsulation of SA and SA release from

PCL-based microparticles. Figure 4.6 shows the average particle sizes, size

distributions and morphologies of electrosprayed particles upon the addition of 10%

wt PEG 6k/35k and with/without 1% wt SA. Table 4.5 shows the corresponding

electrospraying parameters.

Figure 4.6. (A) Electrosprayed particle sizes (PCL:PEG ratio is 90:10), expressed as box plots

showing the medians and 50% of the population. Extremities represent the minimum and the

maximum values. (B-E) SEM images of loaded and non-loaded PCL:PEG microparticles. (F-I) Size

distributions.

Table 4.5. Electrospraying properties used to ensure reproducible and spherical particles, and

resulting particle size characteristics upon addition of PEG and SA.

PCL:PEG

proportions PEG type Protein

Polymer

concentration

(% wt/v)

Flow

rate

(mL/h)

Voltage

(kV)

TTC

distance

(cm)

Average

size

(median)

(µm)

Size

distribution

100:0 - - 10 0.5 10 25 17.9 Monomodal

90:10 PEG 6k - 5 0.5 10 25 3.4 Monomodal

90:10 PEG 35k - 9 3 10 25 13.0 Bimodal

90:10 PEG 6k 1% wt SA 6 0.5 10 25 4.5 Monomodal

90:10 PEG 35k 1% wt SA 9 3 10 25 10.9 Bimodal

In terms of morphology, the surface of particles was smooth upon addition of

PEG (Figure 4.6B,D), as opposed to raw PCL particles, which had a textured surface

for similar electrospraying conditions [83]. This is explained by a reduced

evaporation rate with the addition of PEG, generated by the two different types of

polymers present in solution. As a result, polymer diffusion was enhanced compared

to PCL alone, before full evaporation of chloroform, resulting in smooth surfaces and

0

20

40

60

80

1 3 5 6 8 10

Nu

mb

er

of

Part

icle

s (

%)

Particle Diameter (µm)

1 3 5 6 8 10

0

5

10

15

20

25

30

PCL PCLPEG6k

PCLPEG 35k

PCLPEG6kSA

PCLPEG35kSA

Part

icle

Siz

e (

µm

)

PCL:PEG 6k PCL:PEG 35k

20 µm20 µm

A

0

10

20

30

40

50

60

1 6 11 16 21 26

Nu

mb

er

of

Part

icle

s (

%)


1 6 11 16 21 26

20 µm 20 µm 20 µm 20 µm

Blank Blank1% wt SA 1% wt SAB C D E

F G H I


- 134 -

also reduced particle sizes [60]. The loading of 1% wt SA had no significant effect

on morphology, as would be expected for low loadings (Figure 4.6C,E) [52].

With regard to particle sizes, the addition of smaller PEG chains to the PCL

matrix decreased microparticle size to a greater degree with PEG 6k than PEG 35k

(Figure 4.6A), as expected. This is indirectly due to the electrospraying parameters

required to obtain reproducible spherical particles, listed in Table 4.5. For instance,

to ensure reproducible morphology, polymer concentration was reduced from 10 to

6% wt/v when using PEG 6k, generating significantly smaller particles. Conversely,

when using PEG 35k, a higher flow rate was necessary to maintain reproducible

particles which increased particle size but also affected size distribution, resulting in

a bimodal character. This is typical in electrospraying where increasing flow rates

encourages the droplet subdivision force over chain entanglements [181]. This can

lead to the formation of secondary/satellite droplets being generated from primary

droplets [42], and is illustrated in Figure 4.7. The electrospraying jet break-up

mechanism can be divided in two main modes: the varicose jet break-up mode and

the whipping jet break-up mode. For low current/flow rate, monodisperse particles

can be achieved in the varicose mode (Figure 4.7A). As the current or flow rate

increases, an increase in the surface charge leads to an increase of the ratio of normal

electric stress over surface tension stress, leading to the ejection of secondary

droplets from the main filament, known as varicose instabilities (Figure 4.7B). At

even higher flow rates, the repulsion force of the charge is so strong that it leads to

jet whipping and kink lateral instabilities (Figure 4.7C). Here, the PCL:PEG polymer

blend was shown to have a very small window of monodisperse varicose jet break-up

due to more jet instabilities generated by the blend. As the flow rate increased above

0.6 mL/h, the whipping jet break-up mode was rapidly observed, hence generating

bimodal size distributions and reported in Figure 4.8.


- 135 -

Figure 4.7. Evolution of jet-break up modes in electrospraying.

Figure 4.8. Average particle size of primary droplets (closed symbols) and secondary droplets (open

symbols) obtained for increased flow rates for PCL:PEG 35k electrosprayed microparticles. Errors

bars represent SD.

The loading of 1% wt SA to either PCL:PEG 6k and PCL:PEG 35k resulted in

similar average particle sizes compared to particles without SA (Figure 4.6A),

however size distributions were broader and the occurrence of secondary droplets

was more pronounced in the case of PEG 35k (Figure 4.6I). This is due to the

Varicose

Jet Break-UpWhipping

Jet Break-Up

Electrospraying Flow Rate / Current Increase

Secondary Droplet

Primary

Droplet

A B C


- 136 -

presence of SA within the polymer network which partly hinders local polymer chain

entanglements, acting as a weak link, resulting in a greater frequency of secondary

droplets being generated for higher flow rates [103].

Finally, the effect of PEG content on the final microstructure of the loaded

electrosprayed particles was studied with PCL:PEG 35k blends of 95:5, 90:10 and

85:15 ratios (Figure 4.9). While 5% of PEG had little effect on microparticle sizes

(Figure 4.9A), bimodal size distributions were obtained for 10 and 15% of PEG

(49% secondary droplets in both cases), (Figure 4.9B-C) which were attributed to an

increased presence of PEG in the PCL solution, with smaller PEG chains in the

PCL:PEG blend generating more jet instabilities for PEG content > 5% wt. It is

therefore essential to lower PEG content and PEG MW to ensure monodispersity

while simultaneously produce particle sizes superior to 8 µm to ensure homogeneous

encapsulation, which is obtained for high MW PEG, higher flow rates and higher

concentrations.

Figure 4.9. Increasing PEG contents from 5 to 15% wt and its effects on size distributions of PCL

electrosprayed particles containing 1% wt SA.


4.5.2.1 Encapsulation Efficiencies

When encapsulating hydrophilic proteins, the use of dry (as opposed to aqueous)

protein is an effective means of avoiding molecules being washed away in aqueous

phases or aggregating in solution. This enhances the encapsulation efficiency (EE) of

the system, which is a measure of the amount of therapeutic molecules efficiently

loaded compared to the initial amount. In electrospraying, while up to 100% EE can

be achieved in theory, a wide range of values have been obtained, according to the

processing parameters, leading to differences in the resulting particle size, but also

0

10

20

30

40

1 5 9 13 17 21

Nu

mb

er o

f P

art

icle

s (

%)


0

10

20

30

40

1 5 9 13 17 21


0

10

20

30

40

1 5 9 13 17 21


PCL:PEG 95:5 PCL:PEG 90:10 PCL:PEG 85:15

49 % Secondary Droplets 49 % Secondary DropletsNo Secondary Droplets

A B C


- 137 -

according to different loadings, with increasing EE obtained for higher sizes and

lower loadings [181].

Here, EE results of SA in PCL:PEG microparticles are presented in Figure 4.10.

No significant differences were observed when increasing the amount of PEG from 5

to 10% wt ratio (Figure 4.10A). However, when assessing the encapsulation of

FITC-SA against SA, EE was significantly lower for FITC-SA (Figure 4.10B). The

presence of the FITC may thus have reduced encapsulation, although the reactive

group makes up a very small contribution of the SA protein. Lu et al. have reported

similar observation with lower encapsulation efficiencies of FITC-SA compared to

TGF-ß1, which they attributed to aggregation in solution [124]. When comparing the

EE of SA in particles made of PEG 6k and PEG 35k, EE was always higher for

larger particles, made of PEG 35k, either loaded with 1% wt SA (Figure 4.10C) or

1% wt FITC-SA (Figure 4.10F). This trend is expected, as explained in the previous

section, where size is critical to allow homogeneous encapsulation and it is known

that smaller microparticle sizes lower EE values [64]. An increase in loading also

reduces EE [55, 57], which was indeed observed here when loading 3% wt FITC-SA,

compared to 1% wt (Figure 4.10E).

Figure 4.10. Encapsulation efficiencies obtained for specific formulations of electrosprayed particles

made of PCL:PEG encapsulating 1% wt of protein, upon EX1 (n = 2) where only PBS was used to

extract the protein (closed symbols), and upon EX2 (n = 5) where PBS supplemented with SDS was

used (open symbols). Means ± standard errors (SE) are presented.

0

10

20

30

40

50

1% FITC-SA 3% FITC-SA

0

10

20

30

40

50

FITC-SA SA

0

10

20

30

40

50

95:5 PCL:PEG

90:10 PCL:PEG

EE

(%

)

0

10

20

30

40

50

PEG 6k (~ 5 µm)

PEG 35k(~ 10 µm)

0

10

20

30

40

50

PEG 6k(~ 5 µm)

PEG 35k(~ 10 µm)

0

10

20

30

40

50

EX1 (PBS only)

EX2(PBS + SDS)

EE

(%

)

PCL:PEG 90:10

1% wt SA

PCL:PEG 35k, 90:10

1% wt SA

PCL:PEG 35k, 95:5

1% wt SA

PCL:PEG 90:10

1% wt FITC-SA

PEG 35 k – 1% wt SA

PCL:PEG 35k, 90:10

FITC-SA

A B C

D E F


- 138 -

As can be seen from these results, protein extraction was incomplete. This was

due to the tremendous difficulty in extracting the molecule from the polymer carrier

without generating protein instabilities. For instance, protein aggregation and non-

specific adsorption via the hydrophobic domains of proteins are common issues,

especially during extraction [124, 171, 197], where the concentration of protein at the

water-in-oil interface can reach concentrations as high as 100 mg/mL [34]. This is

because in the absence of an exogenous surfactant, the protein itself assumes this role

and consequently interacts with other protein molecules and polymer

macromolecules. This was evidenced here with release assays releasing larger

protein amounts than the corresponding EE values measured by extraction (using

separate release assays over time), showing evident discrepancies between what the

relatively harsh extraction techniques were indicating, compared to the slower

release assays into PBS. Encapsulation efficiencies were consequently

underestimated due to the harsh nature of the extraction procedure, resulting in

incomplete extraction [198]. Pean et al. observed similar results with SA and PLA in

a DCM/water system, where the protein penetrated irreversibly the interfacial layer

[197]. For this reason, an alternative extraction procedure (EX2) was performed,

here, where the aqueous phase contained 5 mM of an anionic surfactant, sodium

dodecyl sulphate. Unfortunately, SDS was not sufficient to recover all the protein,

since similar EE values were measured (Figure 4.10D). This result is proof of the

often overlooked challenges of measuring EE of proteins in polymer particles, since

proteins act as strong surfactant-like compounds that compete for the water-in-oil

interfacial layer. This phenomenon was strong here during SA extraction, where even

the presence of an anionic surfactant did not suffice to fully extract the proteins.

Hence it can be argued that extraction is not an appropriate mean of measuring the

encapsulation efficiency of particulate systems that contain surfactant-like

compounds such as proteins. We would suggest that in all such cases, the values

obtained reflect more of the extraction efficiency rather than of the encapsulation

efficiency and we would proceed with caution when interpreting any publication in

this area, where encapsulation data should be supported by the release data.

Importantly, here, it should not be lost sight of the fact that the encapsulation is still

occurring even if the extraction does not give representative values since the results

from the release assays imply that encapsulation is indeed high, although

quantification by extraction is inaccurate. Hence, strong protein interactions are not


- 139 -

ideal for characterisation purposes, but paradoxically they actually favor

encapsulation [199]. A more difficult, but efficient way to measure EE would be to

perform a bioactivity cell assay with a calibration curve from exogenously delivered

proteins [29, 30].

4.5.2.2 The Effect of PEG on Protein Release

In this section, the release kinetics of various polymeric formulations involving PEG

as a solubilising agent able to tailor the release profiles of 1% and 5% wt SA-loaded

electrosprayed microparticles was studied. Two molecular weights (PEG 6k and PEG

35k) and two contents (5 and 10% wt), within the PCL matrix, were assessed.

Loading is often a critical variable in release profiles from particles made by

traditional techniques, where higher loading correlates with increased burst release.

In Figure 4.11A-B, the release profiles for SA loadings of 1 and 5% wt were

compared. The higher loading generated a maximum cumulative release of 71%,

with 66% burst within 24 hours. This pattern is attributed to the protein molecules

being localised close to the surface of the particles and therefore being rapidly

solubilised out of the microparticles. Indeed, the diffusion coefficient of solutes

inside an electrosprayed droplet decreases upon increasing the concentration of

solutes, which was high in this instance (5% wt) [52]. As a result, solutes were

unable to diffuse properly towards the centre of the electrosprayed droplet during

solvent evaporation and thus concentrated at the surface of particles (Figure 4.11d),

providing a quick release by diffusion over the first few hours of water penetration

[181]. Conversely, the 1% loading did not generate a similar burst release and

provided sustained release over the whole period of the study (3 months), due to a

better protein distribution within the particles as seen by CLSM in the protein

encapsulation section (Section 4.5.1.3 and Figure 4.11c). The overall cumulative

release was however low after 3 months (maximum of 19%) suggesting incomplete

release over the period of study. This could be due to protein-polymer interactions in

solution as mentioned earlier or due to the protein still being released after that time.

This may be valid considering the polymer being used: PCL which is highly

hydrophobic and can take more than a year to degrade when used in the form of

microparticles [88].


- 140 -

Figure 4.11. The Effect of Loading. (A-B) Release profiles of SA from PCL electrosprayed particles

upon 1% wt (closed symbols) and 5% wt (open symbols) loading. (A) Cumulative amount of SA

released compared to the initial amount of SA used for loading. (B) Cumulative amount of SA

released normalised to the total amount of SA released. PCL:PEG 6k ratio is 90:10. Means ± SE are

presented (n = 3). (C-D) Schematic of protein dispersion inside an electrosprayed microparticle for

(C) 1 and (D) 5% wt loading.

The effect of PEG content and PEG molecular weight are shown in Figure 4.12.

In Figure 4.12a-b, the addition of 5% wt PEG provided higher burst release than 10%

and was due to the different size distributions of the formulations generated by the

increase in flow rate. The 10% wt PEG formulation had indeed a bimodal size

distribution, with 49% of secondary droplets (3-4 µm) (Figure 4.12e), which led to

stronger agglomeration in solution compared to the 5% wt PEG formulation, which

presented a monodisperse size distribution. Release profiles can be described by two

phases; the first phase, dominated by diffusion and the second phase by polymer

degradation. In a recent study by Almería et al., it was shown that the first stage is

highly affected by agglomeration properties and substantially affected by particle

size in both amounts released and rate of release. However, the second phase of

release was much less dependent on size [50]. Here electrosprayed microparticles

were mostly made of PCL, and after SEM assessment (not shown), very little

degradation occurred over the period of study (3 months). Particles maintained their

integrity and spherical morphology, and whereas surface morphology presented more

wrinkles, particle size was not significantly decreased. It is thus safe to assume that

the release profiles presented here are showing only the first stage of release, driven

0

20

40

60

80

100

0 14 28 42 56 70 84

Norm

aliz

ed C

um

ula

tive S

A

rele

ased

(%

)

Time (days)

1% wt SA

5% wt SA

0

20

40

60

80

0 14 28 42 56 70 84

Cum

ula

tive S

A r

ele

ased (

%)

Time (days)

1% wt SA

5% wt SA

High concentration of

protein at the surface

Microparticle

Protein

1% wt Loading

5% wt Loading

A

B

C

D


- 141 -

by diffusion, which is significantly affected by size and agglomeration [49, 50].

While there was no means of measuring particle agglomeration, aggregation in

solution was visually observed here for the 10 % wt PEG formulation. As a result,

less burst and less overall amounts released were obtained (Figure 4.12a-b), in a

same fashion as observed by Almería et al. for rhodamine-loaded PLGA

microparticles [50].

Figure 4.12. The Effect of PEG Content and PEG MW. Release profiles of SA from electrosprayed

particles for: (A-B) increasing contents of PEG 35k, c-d) different molecular weight of PEG (10% wt

PEG). (A-C) Cumulative amount of SA released compared to the initial amount of SA used for

loading. (B-D) Cumulative amount of SA released compared to the total amount of SA released.

Loading is 1% wt. Means ± SE are presented (n = 3). (E) Average particle size of primary droplets

(closed symbols) and secondary droplets (open symbols) obtained for different PEG content and PEG

MW. Errors bars represent SD.

When comparing the effect of PEG MW on release kinetics, again, size

distribution is an important consideration. The PEG 6k formulation had a smaller

particle size with a monomodal distribution compared to the bimodal distribution

from PEG 35k (Figure 4.12e). While it is acknowledged that smaller particles

degrade faster due to increased surface area to volume ratio [200], the blend particles

were not at an established degradation stage over the period of study and are

considered to release SA under diffusion mechanisms, only. Here, particles

comprising PEG 6k had smaller sizes than those made with the PEG 35k, thus it is

expected that proteins were encapsulated within the polymer matrix with a reduced

presence close to the surface of the particle, generating less release by diffusion [52].

Simultaneously, the presence of smaller PEG generated a less porous network within

0

5

10

15

20

PEG 35k5 % wt

PEG 35k10 % wt

PEG 6k10 % wt

Part

icle

Siz

e (

um

)

0

10

20

30

0 14 28 42 56 70 84Cum

ula

tive S

A rele

ased (

%)

Time (days)

10% wt PEG 6k

10% wt PEG 35k 0

20

40

60

80

100

0 14 28 42 56 70 84

Norm

aliz

ed C

um

ula

tive

SA

rele

ased (

%)

Time (days)

10% wt PEG 6k

10% wt PEG 35k

0

20

40

60

80

100

0 14 28 42 56 70 84

Cum

ula

tive S

A rele

ased (

%)

Time (days)

5% wt PEG 35k

10% wt PEG 35k

40

60

80

100

0 14 28 42 56 70 84

Norm

aliz

ed C

um

ula

tive

SA

rele

ased (

%)

Time (days)

5% wt PEG 35k

10% wt PEG 35k

A B

C D

E


- 142 -

the microparticles, thus reducing the initial burst release of larger protein molecules.

Hence, while overall release was low with PEG 6k, it reached the same amount as

with PEG 35k at 84 days and was sustained over the period of study, with no initial

burst release. Importantly, release had not reached a plateau after 84 days, suggesting

further potential release by degradation (Figure 4.12c-d).

These results suggest that the use of PEG in a microparticle system is not

sufficient to control release kinetics, but involves a complex combination of other

parameters. Compared to traditional fabrication techniques, the influence of PEG has

often shown controversial trends; the amount of PEG is more determinant than a

change in PEG MW, but still less significant than a change in matrix MW (PLGA)

[127], and while PEG is usually known to increase burst release, PEG has also been

shown to decrease cumulative release, due to a more acidic pH resulting in more

protein aggregation in solution [124]. Here, in electrosprayed microparticles, it could

be concluded that a lower MW PEG was efficient in reducing burst release, however

protein-polymer interactions impaired complete release, for any formulation type.

Importantly, release kinetics were a result of the particle size characteristics obtained

from different processing parameters required to electrospray microparticles with

PEG to maintain spherical and reproducible morphology. While a smaller particle led

to reduced burst release, a threshold size needs to be met to allow homogenous

encapsulation of protein, and bimodal size distributions, which can be generated for

too high PEG content or high electrospraying flow rate, can lead to particle

aggregation in solution and decreased cumulative release.

4.6 CONCLUSIONS

In conclusion, we demonstrated here that micronised proteins could be

homogeneously encapsulated in electrosprayed polymeric particles using a non-

aqueous route. The presence of PEG within the electrosprayed microparticle matrix

provided a tight control over the characteristics of particles, with PEG content and

MW but also with electrospraying flow rate. Low protein loading, micronisation with

PEG 6k, particle monodispersity and moderate microparticle sizes were efficient in

providing homogeneous encapsulation, and sustained and burst-free release of SA

from PCL particles up to 84 days. Conversely, PEG 35k allowed for burst release

within 3 days. The results presented here are of particular importance for the delivery

Section 4.7 Acknowledgements

- 143 -

of growth factors in tissue engineering applications, since growth factors are

sensitive molecules requiring different types of delivery, from burst to sustained

delivery, according to their function. For instance, a quick delivery of VEGF is

known to be essential at the early stages of bone repair while a more sustained

delivery of bone morphogenetic proteins is required throughout the process. The

possibility of dry encapsulation and the control of encapsulation and release profiles

obtained here by electrospraying and PEG as a micronising and solubilising agent

may thus be well-suited to address the requirements of growth factor delivery

therapies.


The authors would like to thank L.-J. Vandi (University of Queensland) for helpful

discussion on Hansen Solubility Parameter, Dr. Christina Theodoropoulos (QUT) for

help with SEM imaging, Dr. Leonore de Boer (QUT) for help with CLSM imaging.

Thanks to the ARC (Discovery grant no. DP0989000) for financial support.

- 145 -

Chapter 5: Growth Factors Loaded into

Electrosprayed Microparticles:

Detection and Bioactivity Discrepancies

with In Vitro Assays

Nathalie Bock1,2,3

, Tim R. Dargaville1, Giles T. S. Kirby

2, Dietmar W.

Hutmacher3, Maria A. Woodruff

2

Manuscript submitted

© 2014 Nathalie Bock, all rights reserved









Giles T. S. Kirby* Provided technical guidance with cell assays

Assisted in DNA quantification










Chapter 5 In Vitro Release, Detection and Bioactivity of Growth Factors

- 146 -













requirements.





- 147 -

5.1 ABSTRACT

Purpose To provide an efficient growth factor (GF) delivery system that

maintains GF activity and to assess current in vitro means for GF activity

quantification.

Methods Vascular endothelial growth factor (VEGF) and bone morphogenetic

protein 7 (BMP-7) were encapsulated in poly(lactic-co-glycolic acid) (PLGA)

electrosprayed microparticles with poly(ethylene glycol) (PEG) and trehalose, to

assist GF bioactivity. Typical quantification procedures, such as extraction and

release assays using saline buffer were compared with cell bioactivity assays.

Results Saline assays showed that quantification procedures generated a

significant degree of GF interactions, impairing accurate assessment by ELISA

assays, although this shortfall was partially addressed by the use of surfactants in

solution. When both dry BMP-7 and VEGF were vortexed with chloroform, as is the

case during the electrospraying process, reduced concentrations were measured by

ELISA, but the biological effect on myoblast cells (C2C12) or endothelial cells

(HUVECs) was unaffected. When electrosprayed particles containing BMP-7 were

cultured with pre-osteoblasts (MC3T3-E1), significant cell differentiation, assessed

with alkaline phosphatase activity, was observed up to three weeks, contrary to that

predicted by assays in PBS.

Conclusions Electrosprayed particles ensured efficient delivery of fully active GFs

and major discrepancies in quantifying GFs in microparticle systems were

highlighted, when comparing ELISA with cell-based assays.

5.2 KEYWORDS

Bioactivity, bone morphogenetic protein 7, enzyme-linked immunosorbent assay,

electrospraying, microparticles.

5.3 INTRODUCTION

Formulations for the controlled release of therapeutics have been in the spotlight of

the biotechnology industry for many years, and while robust, low molecular weight

drugs have extensively been addressed, developing similar systems for proteins has

proven more challenging [34]. Protein therapeutics are increasingly being explored


- 148 -

and utilised within the field of tissue engineering (TE), since many mechanisms

involved in tissue regeneration are driven by proteins, which need to be delivered

efficiently for maximum benefit [6, 147]. However, development of controlled

release formulations for proteins faces many challenges, due to the relatively large

and complex architecture of most proteins, which incorporate hydrophilic and

hydrophobic domains with numerous reactive groups [171, 176]. The unfolding of

polypeptide chains can expose hydrophobic groups which can interact with other

molecules (aggregation) and with hydrophobic matrices (non-specific adsorption)

(Figure 5.1A) [124, 197, 201]. Hence, proteins are challenging molecules to

encapsulate and deliver from polymeric carriers, but also difficult to quantify. It is in

fact paramount to understand that, while protein encapsulation and in vivo delivery

may be affected by protein denaturation, in vitro experimental conditions and

characterisation methods also account for a significant part of denaturation, leading

to underestimated protein quantification, a vast issue in the field [176]. With this in

mind, two objectives can be set: One is to provide a sustained release formulation

with full preservation of the protein’s native state and the other is to use

representative assays that accurately assess the formulation, both ambitious goals.

In order to address the first objective, protein-carrier formulations require

strategies to preserve the native state and low immunogenicity of proteins [34].

When considering biodegradable polymeric particles, solid encapsulation has

become a superior option compared to aqueous incorporation, which requires a

water-in-oil (w/o) emulsion, and is accepted as a potent cause of protein denaturation

[35] due to the w/o interfacial tension causing protein molecules to unfold [171]. By

using a solid encapsulation process, such as spray-drying or solid emulsion, the

bioactivity of several proteins was improved or maintained [121, 179]. Stabilisers

can also be used to protect proteins [127], by either providing a microenvironment

that reduces the free energy of protein molecules or increases the energy barrier

between the native and denatured states [171]. Micronisation of the protein with

poly(ethylene glycol) (PEG) upon co-lyophilisation [172] and prior to encapsulation,

can effectively lead to a more favorable state of the protein [174-176], as seen for

nerve growth factor [123] and albumin [175], and reduces protein adsorption to the

polymer matrix [176, 197]. Alternatively, the use of saccharides is another approach

to protect proteins [171, 201, 202]. Sugars are small osmolyte molecules which

stabilise proteins via preferential hydration of the native form by hydrogen bonding


- 149 -

during lyophilisation [34, 201] and were shown to protect growth factors (GF) [203]

and proteins [31]. Since low molecular weight sugars dissolve rapidly and are not

retained within the polymer carrier, the presence of PEG would be paramount in

protecting proteins from adsorption onto the polymer matrix once released [171].

Protein denaturation is a key limitation of traditional encapsulation techniques and

has limited the clinical translation. Electrospraying of polymers with therapeutic

molecules is an emerging technique which has been shown to maintain the

bioactivity of some proteins and GFs, including insulin-like GF 1 (IGF-1) [29],

platelet-derived GF (PDGF) [30], transforming GF ß-3 (TGFß-3) and bone

morphogenetic protein 6 (BMP-6) [31]. Although PEG and sugars blended into the

matrix have been widely used for improving protein bioactivity in traditional

techniques [123, 171, 174-176, 199], no studies have investigated their use in

electrospraying.

Further to providing an optimised protein-carrier formulation, accurate

quantification of active proteins is the next challenge. Several techniques can assess

protein denaturation and inform on structural and conformational changes, size and

shape distributions [204]. When encapsulating a protein in a polymeric device,

however, it becomes difficult to assess the protein conformation after extraction or

release in buffer, since irreversible changes can be induced by the processes

themselves (Figure 5.1B) [205]. Buffer saline release conditions, for instance, often

presents an acidic pH and destabilising factors that impair accurate evaluation of

release kinetics. This has been long demonstrated in the literature where polyester-

based microparticles almost always present incomplete release profiles [176] when

assessed with enzyme linked immunosorbent assays (ELISA) and simple protein

assays. The study of GF activity is, hence, more relevant when cell-based assays are

used, which are the best indicator of GF bioactivity. Changes in cell proliferation or

differentiation upon exposure to ‘active’ GFs can be evaluated [181].


- 150 -

Figure 5.1. Schematic description of: (A) growth factor interactions in solution, via hydrophobic

domains (adapted from [171]), and (B) concentration of GF complexes at the water-in-oil interface

upon an extraction procedure.

In order to address the two important challenges of encapsulation and assessment,

we have evaluated two complementary GFs used in bone tissue engineering; bone

morphogenetic protein 7 (BMP-7) and vascular endothelial GF (VEGF) for

encapsulation using an electrospraying technique and assessed the potential of PEG

and trehalose as protective additives within the formulation. We have indeed shown

for the first time, in our previous work, homogeneous loading of micronised proteins

with PEG in electrosprayed polyester particles, and obtained tight control over

particles characteristics [206], making PEG a legitimate choice for further

assessment as a protective agent for GF bioactivity. Here, we have assessed GF

bioactivity at various stages of the encapsulation process with typical buffer assays

using ELISA and compared the results with actual bioactivity results from cell

assays. A direct contact assay of GF-loaded particles with cells was also used to

evaluate the efficiency of the delivery system, and provided key findings regarding

the quantification of GFs in microparticle systems using fundamentally different

assays.

Hydrophilic domains

Hydrophobic domains

Growth Factor (GF)

Hydrophobic

polymer

Non-specific adsorption

Organic droplet

Aqueous droplet

Aqueous phase

Organic phasePolymer chain

GF

Aqueous phase

Organic phase

High concentra-

tion of adsorbed/

aggregated GF

Adsorbed GF-

polymer complex

Covalent/non-covalent aggregation

Aggregated

GF complex

A

B


- 151 -


5.4.1 Materials

Poly(lactic-co-glycolic acid) (PLGA) 85:15 (Mn 41.3 kDa, PDI 1.6) was purchased

from Evonik Industries. Poly(ethylene glycol) (PEG) with Mn = 35 kDa, trehalose,

chloroform, dichloromethane (DCM), polysorbate 20 (PS20), sodium dodecyl

sulphate (SDS), and human serum albumin (SA) were purchased from Sigma-

Aldrich. Recombinant vascular endothelial growth factor (VEGF) was purchased

from ProsSpec-Tany TechnoGene Ltd. via BioNovus Life Sciences. Recombinant

human bone morphogenetic protein-7 (BMP-7) was generously donated by Stryker.

5.4.2 Particle Fabrication

Either SA alone or GF:SA were first micronised to ensure protein particle size

reduction, prior to electrospraying [127, 172, 177]. Aqueous solutions were prepared,

containing SA and PEG, with and without GF, and with and without trehalose. The

GF:SA and (GF:SA):PEG part ratios were maintained at 1:9 and 1:10, respectively

(Table 5.1) [172]. Samples were dissolved in 0.2 µm filtered doubly distilled water

(1 mL) and frozen by immersion in liquid nitrogen. After freeze-drying, PLGA was

dissolved in chloroform and added to the lyophilised GF under magnetic stirring. The

resultant dispersions were probe sonicated for 1 min at 0.5 W (Misonix 3,000). The

final polymer (PLGA:PEG) content was 11% wt/v for a PLGA:PEG part ratio of 9:1.

The dispersions were immediately loaded into 1 mL glass syringes, fitted with a 21 G

stainless steel nozzle and electrosprayed. The dispersions were extruded at a rate of

0.8 mL/h using a syringe pump (World Precision Instruments) and a voltage of 10

kV was applied to the needle tip. The tip-to-collector (TTC) distance was 15 cm and

collectors consisted of aluminium foils (15 × 15 cm2) sterilised with 70% ethanol.

After electrospraying, collectors were placed under vacuum for a further 72 hours.

The microparticles were transferred into glass vials and stored at -18°C until further

analysis.


- 152 -

Table 5.1. Summary of microparticle formulations.

Formulation

Name

GF GF

(µg)

SA

(mg)

Trehalose

(mg)

PEG

(mg)

PLGA

(mg)

Loading (µg GF/mg

microparticles)

Blank - - 1.8 - 33.3 299.7 -

V VEGF 10 0.09 - 16.7 150.3 0.06 (Low)

B1 BMP-7 30 0.27 - 50 450 0.06 (Low)

B2 BMP-7 200 1.8 - 33.3 299.7 0.6 (High)

BlankT - - 1.8 3.3 33.3 299.7 -

VT VEGF 10 0.09 1.67 16.7 150.3 0.06 (Low)

BT1 BMP-7 30 0.27 5 50 450 0.06 (Low)

BT2 BMP-7 200 1.8 3.3 33.3 299.7 0.6 (High)

5.4.3 Particle Characterisation

Particle morphology was characterised with a FEI Quanta 200 scanning electron

microscope (SEM) operating at 5 kV in high vacuum mode. Microparticles were

taped on aluminium stubs and gold coated at 30 mA (SC500 sputter coater, Bio-

Rad). Particle size was assessed with ImageJ analysis software by automated

measurements of particle diameter (National Institutes of Health) based on light

micrographs (AxoVision, Carl Zeiss MicroImaging GmbH).


5.4.4.1 Encapsulation Efficiency

5.4.4.1.1 Extraction Method

Particles (5 mg) were loaded into 15 mL Falcon tubes (Fisher Scientific) and

dissolved in DCM (1 mL), n = 4, and vortexed for 30 s. PBS (1 mL) was added and

tubes were vortexed for 30 s to extract GFs into the aqueous phase. The aqueous

phase was analysed using a human BMP-7 enzyme-linked immunosorbent assay

(ELISA) from R&D systems, according to the manufacturer’s protocol, upon sample

dilution to fit the detection range of the assay.

5.4.4.1.2 Direct Dissolution Method

Particles (5 or 15 mg) were loaded into 15 mL Falcon tubes (Fisher Scientific) and

dissolved in DCM (1 mL), n = 4, and vortexed for 30 s. Tubes were centrifuged at

10,000 rpm for 5 min then the DCM was left to evaporate overnight. PBS (1 mL),

with or without PS20 (0.05%), was added to the tubes and the aqueous phase was

analysed using either a human BMP-7 or human VEGF ELISA assay (R&D

systems), upon sample dilution.


- 153 -

5.4.4.2 Growth Factor Recovery through In Vitro Processing

Using the extraction and direct dissolution procedures stated above, the recovery of

VEGF with and without the presence of unloaded PLGA microparticles was

investigated. A 250 ng/mL VEGF solution (1 mL) (containing 0.1% wt SA) was

lyophilised with and without unloaded microparticles (5 mg), n = 3. The mixtures

were then subjected to the EX or DD procedures and the aqueous phases were

analysed using ELISA, after sample dilution. For BMP-7, recovery upon freeze-

drying and EX procedure that contained surfactants was assessed. Three aqueous

solutions were prepared: PBS, PBS + PS20 (0.05%) and PBS + SDS (0.05%). A 600

ng/mL BMP-7 solution (1 mL) (containing 0.1% wt SA) was lyophilised with and

without unloaded microparticles (5 mg), n = 3. The mixtures were then subjected to

the EX procedure with the various aqueous solutions and the aqueous phases were

analysed using ELISA, after sample dilution. Controls consisted of reconstituted GFs

following freeze-drying, in the same aqueous solutions.

5.4.4.3 In Vitro Release

Particles (10 mg) were placed in 2 mL screw-capped high purity polypropylene

microtubes (Sarstedt), supplemented with the release solution; PBS or PBS + PS20

(0.05%), (1.5 mL). Tubes were agitated at a speed of 8 rpm at 37°C. At specific time

points, microtubes were removed from the incubator and agitation was stopped. After

allowing for natural particle settlement at the bottom of microtubes, the supernatant

(1.3 mL) was collected and replaced by the same amount of fresh release solution.

Supernatants were immediately stored at -20°C for further analysis using ELISA

assays.

5.4.4.4 Growth Factor Bioactivity

5.4.4.4.1 HUVEC Proliferation Assay

The bioactivity of VEGF after micronisation with PEG, SA, trehalose and contact

with organic solvent was studied in vitro with an optimised human umbilical vein

endothelial cell (HUVEC) proliferation assay (section 5.9.1.1, supporting

information). The growth medium consisted of Dulbecco's modified Eagle medium

(DMEM) F12-K nutrient mixture (Invitrogen), 0.1 mg/mL heparin (Sigma), 0.05

mg/mL endothelial cell growth supplement (ECGS) (Millipore), 10% foetal calf

serum (FCS) and 1% penicillin/streptomycin (P/S) (both from Invitrogen). Each

processed VEGF was reconstituted in PBS before assessment with HUVECs, except


- 154 -

for one treatment that included vortexing the VEGF lyophilisate with chloroform, in

order to mimic the process of solid GF dispersion in organic solvent that is

undertaken prior to electrospraying. The solvent was left to evaporate overnight,

before re-dissolution in PBS. A 96-well plate was gelatin-coated and 8,000 cells/well

were seeded. The culture medium contained no ECGS and contained only 5% FCS.

On the first treatment day (24 hours after initial cell seeding and culture in growth

medium), 150 µL of culture medium and 50 µL of VEGF treatments or controls were

added to each well, for a VEGF concentration of 12 ng/mL per well. Cells were then

cultured for 3 days without media change followed by medium removal and freezing

for 48 h, n = 6. After treatment with Proteinase K (Sigma), DNA content was

analysed with the PicoGreen® assay (Invitrogen), according to the manufacturer’s

protocol. In parallel, VEGF concentrations were also measured by ELISA (n = 4).

5.4.4.4.2 C2C12 Differentiation Assay

The bioactivity of BMP-7 was evaluated in vitro using a mouse myoblast cell

(C2C12) differentiation assay. Cells were cultured in DMEM medium (Invitrogen)

supplemented with 10% FCS and 1% P/S. Concentrations of 5,000 cells/well were

seeded in a 48-well plate and incubated for 24 h. First, the effect of BMP-7 dose on

cell proliferation and cell differentiation was studied using doses ranging from 0.7 to

3.3 µg/mL of BMP-7 (section 5.9.1.2, supporting information). Subsequently, the

effect of freeze-drying of BMP-7 (1.4 µg in 1 mL) alone and in the presence of PEG,

SA and trehalose was assessed. In addition, the effect of chloroform was studied by

vortexing the BMP-7 lyophilisate with chloroform and letting the solvent evaporate

overnight, before re-dissolution in PBS (1 mL) and addition to C2C12 cells. The

various treatments were given on day 1, which was replaced by fresh medium

containing identical treatments on day 3. All treatments were removed on day 5 and

cells frozen for 48 h. The number of replicates was 12 per condition, 6 replicates

were used for DNA content quantification by a PicoGreen® assay (Invitrogen) after

Proteinase K treatment (Sigma), and 6 replicates were used for quantification of

alkaline phosphatase (ALP) expression. This latter was measured by adding para-

nitrophenyl phosphate (pNPP) (Sigma) to cells, which is converted to p-nitrophenol

(pNP). The absorbance of pNP was measured at 405 nm with a spectrophotometer

(Bio-Rad). In parallel, the concentrations of BMP-7 were also measured by ELISA

(n = 3).


- 155 -

5.4.4.5 In Vitro Microparticle 2D Culture

The ability of PLGA microparticle formulations to deliver active BMP-7 was

assessed via an in vitro direct contact culture with murine calvaria pre-ostoblast

(MC3T3-E1) cells over 3 weeks. High BMP-7 loading formulations were selected

(B2 and B2T, Table 5.1) to probe for an effect from the trehalose. Microparticles that

did not contain BMP-7 were used as a negative control. Cells were cultured in α-

minimum essential medium (α-MEM) supplemented with 10% FCS and 1% P/S as

the standard growth media for all experimental conditions, except the positive

control 1 (PC1), which contained osteogenic media and included 10 mM β-

glycerophosphate, 0.1 mM ascorbate-2-phosphate and 100 nM dexamethasone in

standard growth media. Concentrations of 20,000 cells/well were seeded in 24-well

plates and incubated with growth media for 24h. On the experimental start day,

media was aspirated and cells were treated with different conditions (1 mL, n = 36, N

= 288) summarised in Table 5.2. Briefly, 2.5 mg of microparticles/well were

selected, representing a maximum of 1.5 µg BMP-7, assuming a 100% loading in

microparticles. Two positive controls were used which comprised the same amount

of BMP-7 as in microparticle formulations, but were delivered in two ways;

representing either a bolus delivery or a sustained delivery. The bolus delivery was

represented by 1.5 µg BMP-7 administered once at the start of the experiment,

directly added to the cells in group PC2, which represented a bolus/burst delivery. In

the sustained delivery group, PC3, 1.5 µg BMP-7 was divided into 7 doses, enabling

BMP-7 to be added, fresh, at each media change (7 in total), hence representative of

a sustained delivery of 214 ng BMP-7 per change, done every 3 days (half volume

removed).


- 156 -

Table 5.2. Experimental conditions for in vitro microparticle culture (n = 36 per group, i.e. n = 12 per

time point).

NC Negative Control Standard media

PC1 Positive Control 1 Osteogenic media

PC2 Positive Control 2 Bolus delivery of fresh BMP-7 (1.5 µg/well at start)

PC3 Positive Control 3 Sustained delivery of fresh BMP-7 (214 ng/well at each media

change)

C1 Condition 1 2.5 mg/well BMP-7 loaded microparticles (B2)

C2 Condition 2 2.5 mg/well BMP-7 loaded microparticles (containing 1%

trehalose) (B2T)

C3 Condition 3 2.5 mg/well unloaded particles (Blank)

C4 Condition 4 2.5 mg/well unloaded particles (containing 1% trehalose)

(BlankT)

Cells were assayed at days 7, 14, 21. Twelve replicates per group were selected

for each time point, 6 replicates were used for DNA content quantification using a

PicoGreen® assay after Proteinase K treatment, and 6 replicates were used for

quantification of alkaline phosphatase (ALP) expression as described previously.

Optical microscopy (Nikon Eclipse TS100-PixeLINK) was used to assess cell

morphology and interactions with microparticles. At each time point, a fraction of

the microparticle-cell sheet was recovered and centrifuged at 1500 rpm for 10 min.

Media was removed and the microparticles were rinsed twice with doubly distilled

water. After final centrifugation, microparticles were freeze-dried and imaged with

SEM according to section 5.4.3.

5.4.5 Statistical Analysis

Statistical analysis was performed with PASW Statistics 18 (IBM Corp). For particle

size, analysis was done on medians using a Mann-Whitney non-parametric test, after

Levene’s test confirmed inequality of variances. Elsewhere, analysis was performed

with a two-way analysis of variance (ANOVA), fitting the interactions as well as the

main effects and post-hoc tests were performed using Games-Howell, assuming

unequal variances. The significance level was determined for p < 0.05.

5.5 RESULTS

5.5.1 Particle Microstructure

Polymeric microparticles encapsulating BMP-7 or VEGF were prepared by

electrospraying. The co-lyophilisation of growth factors with PEG prior to

electrospraying was used to form micron-sized GF-particles [127, 172, 177, 206].

Section 5.5 Results

- 157 -

Upon dispersion of the lyophilised protein mixture in PLGA 85:15 solution and

further electrospraying with optimised parameters, spherical and narrowly dispersed

microparticles were obtained with an average size of 5.0 ± 1.3 µm. After

micronisation with 1% wt trehalose and further electrospraying, a slight, but

significant (p < 0.001) increase in average size and size distribution was observed

(5.7 ± 1.6 µm) due to the slight increase of overall concentration of solids and

presence of the additive [181]. Morphologies were similar; spherical and smooth, and

were not affected by incorporation of trehalose (section 5.9.2, supporting

information).

5.5.2 GF Encapsulation Efficiency

The results of GF encapsulation efficiency (EE) into PLGA:PEG microparticles are

presented in Figure 5.2. Details of microparticle formulations can be found in Table

5.1. During the process of electrospraying, solid proteins are dispersed in a polymer

solvent, thus there is no protein dissolution into an aqueous phase and high EE values

are expected. However, all results were below 50%, with high dispersity and with no

clear trend with the addition of trehalose (Figure 5.2). As the extraction procedure

(EX) itself was thought to be the reason for such low readings, the direct dissolution

(DD) procedure was used to avoid the interface by allowing the organic solvent

(DCM) to evaporate, before re-dissolution in PBS. However, EE results were not

significantly different to the EX procedure (p = 0.31), as shown in Figure 5.2A for

BMP-7 loaded microparticles, without trehalose (B1) and with trehalose (BT1)

within the matrix, and presented an even bigger variance than EX results, indicating

a lower reproducibility of the technique. When more particles, 15 mg instead of 5

mg, were used to undergo the DD procedure again, results were significantly lower

(p = 0.007), showing a clear impact of the polymer matrix in solution (Figure 5.2B).

Finally, when a surfactant (PS20) was used as a means to potentially

dissociate/displace aggregated/adsorbed GF (Figure 5.2C), results were even lower

for all formulations (p = 0.006), showing no improvement, in fact detrimental effects

were seen for GF recovery in this case.


- 158 -

Figure 5.2. Encapsulation efficiencies of various formulations determined; (A) for different

procedures, (B) for different amounts of particles, and (C) for different re-dissolution solutions,

according to the direct dissolution procedure (B,C). Means ± standard errors (SE).

5.5.3 GF Recovery through In Vitro Processing

A VEGF solution (250 ng/mL) was lyophilised and subjected to the extraction and

direct dissolution procedures, in order to study the impact of the encapsulation

efficiency processes–involving contact with DCM and a w/o interface in the case of

EX–on GF quantification by ELISA. Figure 5.3A shows the recovery of VEGF when

the solution was processed without, and in the presence of, unloaded microparticles.

Both procedures provided similar results (p = 0.534), with DD showing a larger

variance. Only around 30% of VEGF was recovered in the absence of particles,

suggesting GF aggregation. Since the DD procedure did not involve a w/o interface,

it could be concluded that the lyophilisation step was more determinant in GF

aggregation than the w/o interface. In the presence of unloaded particles, only 8%

recovery was obtained, significantly lower than without particles (p = 0.002),

showing again the negative impact of polymer matrix in solution in recovering GFs,

which had non-specifically adsorbed to the matrix.

The experiment was repeated for BMP-7 and was expanded to include two

possible surfactants, PS20 and SDS in order to attempt to dissociate aggregated GFs

and separate adsorbed GFs from the polymer, should those interactions be reversible.

The results are presented in Figure 5.3B-C. The first striking result was the low

0

10

20

30

40

50

B1 BT1

EE

(%

)

Direct Dissolution

Extraction

A B

C

0

10

20

30

40

50

B1 BT1

EE

(%

)

5 mg

15 mg

0

10

20

30

40

50

60

V VT B1 BT1 B2 BT2

EE

(%

)

PBS

PBS+Polysorbate 20

Section 5.5 Results

- 159 -

recovery of BMP-7 (less than 20%), as measured by ELISA, after only freeze-drying

and re-dissolution (Figure 5.3B), identifying freeze-drying as a significant issue for

GF detection too. Next, similar to the results attained for VEGF, BMP-7 recovery

was lower after the EX procedure and after contact with unloaded particles (Figure

5.3C). While this lower recovery was statistically significant compared to lyophilised

and re-suspended BMP-7, values were similar for the extracted BMP-7 with and

without particles (p = 0.159), suggesting that aggregation phenomena were more

critical here than non-specific adsorption to the matrix. PS20 was able to recover

some BMP-7, however only SDS was able to fully recover BMP-7 (Figure 5.3C).

Figure 5.3. GF recovery after in vitro processing, as measured by ELISA. (A) VEGF recovery after

the direct dissolution (DD) and extraction (EX) procedures in the presence and absence of unloaded

microparticles in PBS. (B) BMP-7 recovery after different treatments and mediums. Means ± SE, n =

3.

5.5.4 In Vitro GF Release

In order to obtain accurate values of GF released in solution from microparticles, it is

necessary to minimise GF interactions with other GF molecules, Eppendorf vials and

PLGA microparticles. When placing GF loaded-formulations in PBS, all release

profiles provided less than 1% cumulative release (not shown). When PS20 (0.05%)

was added to the release medium, to minimise GF interactions with their

environment, release profiles were higher (section 5.9.3, supporting information).

However, cumulative release reached only 16% after 3 weeks, still suggesting

incomplete release and/or underestimation of amounts released due to GF-

environment interactions. According to the previous section, PS20 may indeed not be

able to fully displace adsorbed GF and dissociate aggregated GF.

0

20

40

60

80

100

120

140

Solution FD FD+EX

FD+EX

+Particles

BM

P-7

Recovery

(%

)

PBS

PBS+PS20

PBS+SDS

0.0

5.0

10.0

15.0

20.0

25.0

FD FD+EX

FD+EX

+ParticlesB

MP

-7 R

eco

ve

ry (

%)

PBS

PBS+PS20

PBS+SDS

B C

0

10

20

30

40

50

60

DD DD +Particles

EX EX +Particles

VE

GF

Re

co

ve

ry (

%)

A


- 160 -

Figure 5.4. The Effect of Trehalose in Particle Formulation. Release profiles of BMP-7 loaded

PLGA:PEG microparticles for; (A) low BMP-7 loading (0.06% wt) and (B) high BMP-7 loading

(0.6% wt). Means ± SE, n = 3.

The presence of trehalose in the microparticle formulation had a significant effect

on released/measured GF with lower overall cumulative release for both low and

high loadings of BMP-7 (p < 0.001) (Figure 5.4A and Figure 5.4B, respectively).

Cumulative release values were similar for low and high loadings of BMP-7 loaded

microparticles that did not contain trehalose, but release was reduced for high

loading when trehalose was present in the formulation, thus showing a correlated

effect of trehalose and loading on release (p < 0.001).

5.5.5 In Vitro GF Bioactivity

5.5.5.1 VEGF

HUVEC cells are known to proliferate in a dose-dependent manner upon exposure to

VEGF, but in vitro culture conditions, involving cell seeding density, medium

content and VEGF dose are critical in HUVECs’ response [207]. Hence, the optimal

culture conditions to measure the effect of VEGF on HUVECs were established by

testing several culture parameters. Results are presented in supporting information,

section 5.9.1.1. It was concluded that VEGF stimulated the linear proliferation of

HUVECs up to 20 ng/mL. The optimal culture conditions were observed for at least

3,000 cells/well seeding density in a medium that contained only 5% FCS and no

endothelial growth supplement.

The next step was to assess VEGF bioactivity after various stages of potential

denaturation during processing, involving micronisation with PEG, SA and trehalose

and further vortexing with chloroform (See Table 5.3 for details), which are the most

critical steps during microparticle processing where VEGF may get denatured. A

starting VEGF concentration of 48 ng/mL was used for testing the processing steps.

A B

0

10

20

30

40

50

0 3 6 9 12

Cum

ula

tive B

MP

-7 r

ele

ase (

%)

Time (days)

B1

BT1

0

10

20

30

40

50

0 3 6 9 12

Cu

mu

lative

BM

P-7

re

lea

se

(%

)

Time (days)

B2

BT2

Low BMP-7 Loading High BMP-7 Loading

Section 5.5 Results

- 161 -

After reconstitution in PBS, VEGF samples were dispensed on cells with a final

concentration in wells of 12 ng/mL. The quantification of VEGF concentration and

HUVEC proliferation are presented in Figure 5.5.

Table 5.3. Summary of treatments on HUVECs.

Name Description

NC1 Culture medium

NC2 Culture medium with 25% PBS + SA

NC3 Culture medium with 25% PBS + SA + PEG

NC4 Culture medium with 25% PBS + SA + PEG + trehalose

NC5 Culture medium with 25% PBS + SA + trehalose after contact with

chloroform

PC Fresh VEGF

T1 Freeze-dried VEGF

T2 Micronised VEGF with PEG and SA and trehalose

T3 Micronised VEGF with PEG, SA and trehalose, subjected to chloroform

First, upon micronisation and unlike BMP-7, it appeared that VEGF was fully

detectable according to ELISA analysis (Figure 5.5a, T1-T2). After contact with

chloroform (T3), however, only 69% VEGF was detected, suggesting GF

aggregation. Strikingly, when compared with the proliferation results, although

proliferation was slightly lower compared to the control (Figure 5.5b), there was no

statistical difference (p = 0.15) between PC and T3 groups. Importantly, the

proliferation result was identical for micronised VEGF before (T2) and after (T3)

contact with chloroform, with around 83% bioactivity (Figure 5.5C) in both cases.

Furthermore, it can be seen from the decreasing histograms in Figure 5.5B that each

subsequent processing step (freeze-drying, micronisation with additives) had an

increasingly negative impact on cell proliferation, although there were no statistical

differences with the control. This suggested that the processing steps themselves

were more critical for the bioactivity of VEGF than the contact of VEGF with

chloroform. Lastly, in order to confirm that it was not the presence of the various

additives (SA, PEG, trehalose) that impaired cell proliferation, HUVECs were

treated with negative controls that did not contain VEGF, but the various additives

(NC2 to NC5, Table 5.3) and results are shown in Figure 5.5D. No statistical

differences were observed between any groups.


- 162 -

Figure 5.5. The Effect of VEGF Micronisation and Vortexing with Chloroform. (A) VEGF

quantification by ELISA (n = 4). (B,D) DNA content measured by a PicoGreen® assay (n = 6), (C)

VEGF bioactivity, as compared to freeze-dried BMP-7 re-suspended in PBS (n = 6). Means ± SE. (*

indicates statistical significance compared to PC). Refer to Table 5.3 for description of treatments.

5.5.5.2 BMP-7

BMP-7 has been shown to be an effective inducer of C2C12 myoblast cell

differentiation into osteoblast cells at more than 200 ng/mL [208]. The extent of

differentiation can be correlated to the expression of alkaline phosphatase, an early

marker of osteogenic differentiation, assuming that an effective BMP-7

concentration range is used, so that the bioactivity of BMP-7 may be determined.

Hence we first confirmed that PEG and SA were non-cytotoxic to C2C12 cells

(section 5.9.1.2, supporting information). Next we looked at proliferation and

differentiation and established that the most effective concentration range of BMP-7

to be used with C2C12 was between 0.2 and 1.4 µg/mL. An excessively higher dose

of BMP-7, 3.3 µg/mL, was shown to reduce both proliferation (40%) and ALP

expression (section 5.9.1.2, supporting information).

While this was not the case for VEGF (previous section), it appeared from section

5.5.3, that the freeze-drying process of a BMP-7 solution and further re-dissolution in

PBS, itself lowered the detection of reconstituted BMP-7 by ELISA analysis in the

presence and absence of surfactant, suggesting GF aggregation. The experiment was

thus repeated so that the bioactivity of reconstituted BMP-7 could be assessed in

parallel (section 5.9.4, supporting information). ELISA analysis detected

0

10

20

30

40

50

60

PC T1 T2 T3

VE

GF

concentr

ation (

ng

/mL)

0

200

400

600

PC T1 T2 T3

DN

A c

oncentr

ation (

ng

/mL)

A BVEGF Concentration Cell Proliferation with VEGF

70

80

90

100

110

T1 T2 T3

Bio

activity

(%)

C DBioactivity Cell Proliferation - Negative Controls

0

50

100

150

200

NC1 NC2 NC3 NC4 NC5DN

A c

oncentr

ation (

ng/m

L)

*

Section 5.5 Results

- 163 -

significantly less BMP-7 after freeze-drying (64%, p = 0.048), but no differences in

bioactivity were observed (p = 0.08). This indicated that once in solution with C2C12

cells, freeze-dried BMP-7 performed in a similar way as unprocessed BMP-7.

For the final experiment, the effects of various stages of potential denaturation

involved during the electrospraying process on the bioactivity of BMP-7 were

assessed (see Table 5.4 for details of the groups). The freeze-dried and reconstituted

BMP-7 was referred to as the control, since there were ELISA discrepancies between

fresh and freeze-dried BMP-7, as explained above. BMP-7 concentration, C2C12

proliferation and ALP expression of C2C12 cells were determined and are presented

in Figure 5.6. In a similar fashion to VEGF, but even more pronounced with BMP-7,

the micronisation of BMP-7 and subjection to chloroform was detrimental to BMP-7

detection by ELISA (Figure 5.6A). Cell proliferation was, however, not impaired for

any condition (Figure 5.6B) and bioactivity was not altered for micronised BMP-7.

Upon treatment with chloroform, BMP-7 retained 98% of bioactivity (Figure 5.6D),

hence demonstrating the non-denaturing effect of organic solvent on BMP-7, which

was similar to those attained for VEGF.

Table 5.4. Summary of BMP-7 treatments on C2C12 cells.

Name BMP-7 treatment before re-dissolution in PBS

T1 Freeze-dried BMP-7

T2 Micronised BMP-7 with PEG, SA and trehalose

T3 Micronised BMP-7 with PEG, SA and trehalose, subjected to

chloroform


- 164 -

Figure 5.6. The Effect of BMP-7 Micronisation and Vortexing with Chloroform. a) BMP-7

quantification by ELISA (n = 3). b) DNA content measured by a PicoGreen® assay (n = 6), c) ALP

expression (n = 6), d) BMP-7 bioactivity, as compared to freeze-dried BMP-7 re-suspended in PBS.

Means ± SE. (* indicates statistical significance compared to control (T1 here)).

5.5.6 In Vitro Microparticle 2D Culture

In this experiment, known amounts of microparticles were placed in direct contact

with MC3T3-E1 pre-osteoblast cells, which are known to differentiate into

osteoblasts upon stimulation with BMP-7. First, a preliminary experiment was

performed to find an appropriate amount of particles which would not impair cell

proliferation due to physical stresses on the cells, while still providing an ALP

reading. Based on the results presented in section 5.9.5.1, supporting information

(0.5 to 5 mg/well were tested), 2.5 mg/well was selected. Since the B2 and BT2

PLGA formulations (with the high BMP-7 content) were evaluated here, the

maximum amount available per well would have been 1.5 µg BMP-7. Hence two

positive controls with fresh BMP-7 were used to mimic this amount, but one (PC2)

was delivered as bolus (one-time injection at the beginning of experiment),

analogous to the surgical procedure routinely employed in the clinic for OP-1 for

example. The second control was reasoned as a sustained delivery, by dividing the

initial dose so that a small dose was given at each media change, representative of a

sustained delivery (PC3).

During the entire 3-week culture, MC3T3s proliferated extensively and DNA

content was similar across all groups (within each time point), except for the positive

0

0.4

0.8

1.2

1.6

2

T1 T2 T3

Co

nce

ntr

atio

n (

µg/m

L)

BMP-7 Concentration

0

25

50

75

100

125

150

T1 T2 T3

Cell

Pro

lifera

tion (

%)

Cell Proliferation

0.E+00

1.E-03

2.E-03

3.E-03

4.E-03

T1 T2 T3

pN

P A

bsorb

ance/D

NA

C

on

tent

ALP activity

0

25

50

75

100

125

150

T1 T2 T3

Bio

activity (

%)

Bioactivity

BA

C D

**

Section 5.5 Results

- 165 -

control PC1 (osteogenic media) which triggered more proliferation than all the other

groups at day 7 (data not shown), due to the fundamentally different composition of

media which boosted proliferation at this early time point. Neither formulation of

BMP-7 loaded- or unloaded microparticles had any negative impact of proliferation

compared to all the other controls, indicative of the cyto-compatibility of

electrosprayed particles in contact with MC3T3s. All wells showed confluent

monolayers of cells at all time points, and particles fully covered the monolayers

(section 5.9.5.2, supporting information).

Figure 5.7. SEM images of BMP-7 loaded PLGA microparticles after in vitro culture with MC3T3-

E1 cells.

Control cells had a similar morphology in all groups. In the wells that contained

particles, all cells had a similar morphology (independently of BMP-7 or presence of

trehalose) and showed positive interactions with the microparticles by attaching to

them. In fact, at the first analysis point (day 7), it was already impossible to recover

any particles without detaching the entire cell monolayer which had established

strong bonds with the microparticles. After culture, microparticles were recovered

for SEM imaging. Results are presented in Figure 5.7. At all time points and for all

formulations, cells spread around particles, covering several particles simultaneously

and spanning to adjacent particles until a micro-patterned cell monolayer

incorporating all particles was established.

The results of ALP expression of MC3T3-E1 over time are presented in Figure

5.8. As expected, all positive controls were superior to the negative control (NC) (p <

0.001) and there were no significant differences with unloaded particles (p = 0.795

Dense cell sheet covering PLGA microparticles

B2 – Day 21

B2 Formulation BT2 Formulation

Da

y 7

Da

y 1

4D

ay 2

1

10 µm 10 µm

10 µm 10 µm

10 µm 10 µm 50 µm


- 166 -

and 0.877 for Blank (no trehalose in formulation) and BlankT (trehalose in

formulation), respectively – See Table 1 for details). Interestingly, while the bolus

delivery (PC2) consisted of only one BMP-7 injection at the beginning of the

experiment, ALP expression was still observed after 3 weeks, although decreasing

over time. Except at day 14, where the sustained delivery (PC3) was higher than the

bolus delivery (p = 0.04), there were no differences at day 7 and day 21, implying

here that there were no long-term benefits in the sustained delivery compared to the

bolus delivery although the effect at 14 days was significant and so proved that a

constant stimulus with fresh BMP up to 2 weeks provided more differentiated cells

than with the bolus delivery.

As can be seen from Figure 5.8, both formulations of BMP-7-loaded particles

(B2/B2T, +/- trehalose) had enhanced readings compared to the negative controls

(NC, C3, C4) at all time points, proving that both formulations could still have a

positive effect on the differentiation of MC3T3-E1 cells up to three weeks, while the

release data (Figure 5.4B) suggested no more release of BMP-7 was detected after

three days. Compared to the sustained delivery control (PC3), readings were however

lower at all time points, suggesting that the dose delivered by microparticles was

inferior to its fresh counterpart. Importantly, contrary to what was suggested by the

release profiles in Figure 5.4, there were no statistical differences between the

trehalose free BMP-7 formulation (B2) and the trehalose BMP-7 formulation (BT2),

which performed in a similar way at all time points.

Section 5.6 Discussion

- 167 -

Figure 5.8. ALP activity of MC3T3-E1 cells during a 3-week culture with various treatments. Results

were divided by the DNA content measured for each group separately. Means ± SE (n = 6 for ALP, n

= 6 for DNA content). (* indicates statistical significance compared to PC1 (osteogenic media) for

PC2, PC3, C1 and C2, # indicates statistical significance compared to PC3 (sustained fresh delivery)

for C1 and C2, ns and † = local non-significance and significance, respectively, between local

groups).

5.6 DISCUSSION

5.6.1 GF Quantification with In Vitro Assays

When assessing the encapsulation efficiency of a system, the protein of interest needs

to be protected. Serum albumin is relatively resistant against degradation and BMP-7

and VEGF possess a positive net charge and lower molecular weight in respect to

SA, and thus SA represents a good option to be used as a stabiliser in our formulation

[205]. Typically, SA competes with the therapeutic protein at w/o interfaces, which

makes it relevant when using an extraction procedure to quantify GF encapsulation

efficiency. However, here, the presence of SA was not sufficient to prevent

incomplete and irreproducible extraction for both GFs, with and without the presence

of a surfactant (Figure 5.2). As explained in our previous study [206], even in the

presence of stabilisers in the formulations or surfactants in solution, proteins can act

as very strong surfactant-like compounds that compete for the water-in-oil interfacial

layer, generating interactions via the hydrophobic domains of the GFs (represented in

Figure 5.1A). As a consequence, the resulting complexes are trapped at the

interfacial layer of the w/o interface during the EX process, providing

0.0E+00

4.0E-05

8.0E-05

1.2E-04

Day 7 Day 14 Day 21

pN

PA

bsorb

ance/D

NA

Co

nte

nt

NC: Growth media

PC1: Osteogenic media

PC2: 1.5 µg BMP-7/well at start

PC3: 214 ng BMP-7/well at each media change

C1: BMP-7 loaded PLGA microparticles (B2)

C2: BMP-7 loaded PLGA microparticles (+ trehalose) (BT2)

C3: unloaded PLGA microparticles (Blank)

C4: unloaded PLGA microparticles (+ trehalose) (BlankT)

ns

ns

ns

ns

ns

†

* *

*

*

*

*

*

*#

# #

# #


- 168 -

underestimated EE values (Figure 5.1B). This was further proven in section 5.5.3,

when ‘non-encapsulated’ VEGF and BMP-7 were submitted to the extraction

process, showing first GF aggregation, and then non-specific adsorption to polymer

once unloaded particles were added to both GF solutions (Figure 5.3).

Importantly, BMP-7 showed extensive aggregation during the freeze-drying step

(only 20% recovery), even before any extraction treatment (Figure 5.3), as measured

by ELISA. A similar result was observed by Lochmann et al. where one single

freeze/thaw cycle caused a loss of about one third in ELISA detection of BMP-2,

clearly suggesting that the freezing step was the issue [199]. Remarkably, when

testing here freeze-dried BMP-7 on cells, no bioactivity differences were observed in

vitro in contact with C2C12 cells, compared to control BMP-7 (Figure 5.6). Hence,

while the ELISA assay could have suggested covalent GF aggregation upon freeze-

drying, the bioactivity assay confirmed that the interactions were non-covalent in

nature, since BMP-7 was not denatured and performed equally as well as before

freeze-drying. This result really emphasises the limitations of the ELISA assay,

especially for assessing the ‘bioactivity’ of a GF. For instance Wang et al. showed

that after lyophilisation, they had lost 75% of the BMP-2 detectable by ELISA and

thus increased their in vivo dose for compensation, stating that 75% of the

‘bioactivity’ was lost [209]. This approach is erroneous considering that the ELISA

assay is proven here to be an insufficient and misleading means of ‘bioactivity’

detection, and that cell assays remain a superior method for assessing the bioactivity

of a processed GF.

5.6.2 The Use of Surfactants in In Vitro Assays

Surfactants are amphiphilic molecules commonly used to lower the surface tension

between dissimilar phases. In the presence of proteins, they act by effectively raising

the energy barrier for intermolecular interactions between proteins by surrounding

them with their hydrophilic ends. The alignment of surfactant molecules around the

proteins and the thickness of the surfactant layer are determinant in the surfactant’s

performance [175]. Here, two types of surfactants were investigated in GF recovery,

EE and release assays; a non-ionic surfactant, polysorbate 20 (PS20), and a less

commonly used anionic surfactant, sodium dodecyl sulphate (SDS).

When a BMP-7 solution was subjected to the extraction procedure with and

without the presence of unloaded particles, both surfactants were beneficial in


- 169 -

dissociating non-covalent BMP-7 aggregates and displacing adsorbed BMP-7 from

the polymer matrix (Figure 5.3). SDS was more effective than PS20 for BMP-7, with

fully recovered BMP-7, which is attributed to the ionic and steric differences of the

surfactant’s heads. Importantly, these results proved that the extraction process,

involving GF dispersion in organic solvent and a w/o interface generated only

reversible conformational changes. However, when using PS20 for extracting both

BMP-7 and VEGF from loaded microparticles, lower encapsulation efficiencies than

with PBS only were measured by ELISA (Figure 5.2), when the direct dissolution

(DD) procedure was used. This is explained by the absence of w/o interface in the

DD procedure. In the extraction case, surfactants compete with the GFs for the

interface, hence efficiently shielding the proteins from the interface. In the DD

procedure, upon evaporation of the organic solvent, the dried polymer becomes the

preferential site for GF interactions, hence favouring aggregation and adsorption.

Once these phenomena have happened, the presence of surfactant in the re-

dissolution medium actually favours those interactions, hence inhibiting dissociation

of aggregated and adsorbed BMP-7 [205]. In contrast, when PS20 was added to the

release medium, up to 16% cumulative release was observed for BT1 particles,

compared to less than 1% for PBS only. Thus, contrary to the EE results, PS20

clearly promoted release. This inferred that encapsulated BMP-7 within the

microparticles was less aggregated and less adsorbed to the polymer matrix than it

was upon the DD process, leading to a beneficial effect of PS20 in release media but

not in the EE procedure. Incomplete release and low amounts recovered still

suggested that the protection by PS20 was poor. An ionic surfactant may thus be

more relevant in the context of protein recovery [204, 210].

5.6.3 The Use of Stabilisers in Microparticle Formulations

The use of sugars as stabilising agents for encapsulation of therapeutic molecules

into polymer particles has been incredibly well covered over the last twenty years

[34, 125, 205, 211]. Like other stabilisers, sugars protect proteins by preferential

interactions [201, 204, 205]. However, it is clear from the literature that there is not

always a beneficial effect in using any sugar with any protein, and sugars rather work

on a protein-to-protein basis, possibly drawing ambivalent pictures for different

proteins. Bilati et al. have summarised some recent studies, involving, amongst

others, lysozyme, NGF, IGF-I, with mono- and polysaccharides [125]. They


- 170 -

concluded that trehalose was statistically more effective than other sugars, as

opposed to mannitol, which was largely inefficient. While many studies report

protein stability data, less information is found upon the effect of trehalose on the

actual encapsulation efficiency and release of proteins from polymeric particles. As

mentioned by Wu et al., the influence on release characteristics needs to be

addressed since some stabilisers are compromised with burst release or aggregation

upon release [171].

Here, when 1% wt trehalose was present in the final particle formulations that

contained high amounts of BMP-7, reduced encapsulation efficiency (Figure 5.2) and

burst release followed by incomplete release were observed (Figure 5.4). However,

there were no statistical differences between the trehalose-loaded and trehalose-free

formulations when cultured with MC3T3-E1 cells, which presented similar ALP

values over time and triggered differentiation up to three weeks (Figure 5.8),

contrary to both what the encapsulation efficiency and release assay in solution (no

cells) had predicted. Clearly, the presence of trehalose had no negative impact on the

actual encapsulation and delivery of BMP-7 and its presence in the microparticle

formulation was only detrimental to the detection of BMP-7 in solution. Hence,

trehalose is shown here to promote GF non-covalent aggregation during

lyophilisation, which impaired accurate quantification, but not effectiveness of the

delivery system. This is an important result considering that many studies only

address additives in solution and do not compare results with actual cell assays,

which, as demonstrated here, may show different results. Here, because there was no

actual benefit to use trehalose in the formulation, but rather it led to inaccuracies in

quantification, trehalose may not necessarily need to be used within BMP-7 loaded

electrosprayed PLGA microparticles, since PEG only may have been sufficient to

ensure GF bioactivity. Interestingly, 10% wt PEG in the microparticle formulation

may not have been sufficient to prevent GF aggregation for ELISA readings. In

general, PLGA/PEG blends in traditional microparticles are efficient in reducing

aggregation [176]. However, Jiang and Schwendeman showed that less than 20% wt

PEG content in PLA/PEG microspheres resulted in incomplete and insoluble non-

covalent SA aggregates [180]. A similar issue may have happened here with the

PLGA/PEG formulations. However, increasing the amount of PEG in the

electrospraying solution leads to less reproducible morphologies due to

electrospraying jet instabilities [206].


- 171 -

As stated by Walle et al., while the use of trehalose and PEG is very popular, a

clear need for these additives has not yet emerged and case-by-case basis studies are

required [211]. In electrospraying, we showed that PEG was efficiently used to

homogeneously encapsulate dry proteins and gave tailored protein release profiles,

but PEG was not sufficient to address in vitro quantification assays where non-

covalent aggregation hindered accurate measurements by ELISA.

5.6.4 Bioactivity of GF through In Vitro Processing

In the literature, the bioactivity of proteins (SA) and growth factors (IGF-1, PDGF,

VEGF, BMP-7, TGF-β3) immediately released from electrosprayed particles and

assessed with cell proliferation assays was high (80-90%) [29, 30, 71]. Interestingly,

over time, the bioactivity of released VEGF and PDGF decreased (less than 21%

after 21 days), which the authors attributed to the assay conditions which denatured

the GFs, prone to oxidation and pH-dependent deamidation [30]. This result, in

particular, underlines the unsuitability of PBS assays, where cells are not present, for

assessing released GFs.

Here, when encapsulating GFs into electrosprayed particles, there are two steps

which may potentially denature GFs; first the micronisation with additives, then the

mixing with organic solvent prior to electrospraying. It is thus important to assess the

bioactivity of GFs after those two steps, rather than after release in buffer which in

itself involves denaturing factors. Hence, those conditions were assessed separately

and in combination, for BMP-7 and VEGF. The micronisation step and contact with

solvent were proven non-harmful to BMP-7 bioactivity (98% bioactive), although the

ELISA quantification measured a lower reading for those conditions (only 43% of

the control), (Figure 5.5). This confirmed that the micronisation and contact with

solvent generated a large degree of BMP-7 aggregation, but which was non-covalent

since bioactivity was not impaired. With VEGF, the ELISA analysis detected no

aggregation during the micronisation step, but like BMP-7, the reading for VEGF

was lower after contact with organic solvent (58% of the control), although, as with

the BMP-7, no significant differences in bioactivity were noted (83% bioactive),

(Figure 5.6). These results have thus shown that for both growth factors studied here,

a large degree of aggregation occurred after vortexing with organic solvent, which

biased quantification with an ELISA assay. However aggregated GFs had no

significant impact on both GFs’ bioactivities, and performed equally to controls.


- 172 -

5.6.5 GF Delivery in an In Vitro 2D Culture

The previous sections validated the pre-electrospraying processing steps as safe for

maintaining GF bioactivity but really serve to emphasise the difficulties of GF

quantification. Next, the GF bioactivity post-processing was assessed in combination

with a direct contact assay with cells. Cells actively consume growth factors, either

released in the medium or by direct contact with the microparticles. An in vitro

context with cells has some similarities with the in vivo context, and is fundamentally

better than simple PBS assays where no cells are present to take up released growth

factors, which may fatally get adsorbed or aggregated in solution before analysis.

Therefore it was paramount for assessing the efficiency of the electrosprayed

microparticles to place the microparticles directly in contact with cells in a 2D in

vitro culture, which was done here with BMP-7 loaded formulations in contact with

MC3T3-E1 cells.

Microscopy analysis revealed the positive interactions of particles with cells,

which firmly bonded together to form particle-integrated monolayers (Figure 5.7),

and did not impair cell proliferation. The size of electrosprayed particles was within

the order of size of MC3T3-E1 cells, thus was suitable for cellular stimulation by

topographical cues [212]. Positive interactions were expected considering that

unloaded and protein-loaded electrosprayed microparticles on the 5 to 10 µm size

range had shown positive effects on several cell types including fibroblasts [83] and

pre-osteoblasts [182] but such positive effects were shown here for the first time with

growth-factor loaded electrosprayed particles.

When BMP-7 was freshly delivered in a sustained fashion (PC3), ALP readings

were higher than the readings from electrosprayed particles at all time points (Figure

5.8). This is explained by the particles not being degraded after 21 days as evidenced

by SEM analysis (Figure 5.7) and we hypothesise that more BMP-7 may be released

afterwards. This can only be addressed by a complementary assay done over a longer

time frame, such as alizarin red staining, which highlights mineralised bone-like

tissue, since ALP is only an early marker of osteogenic differentiation. Here, the

dose selection for PC3 was quite high; it was the maximum possible dose

encapsulated in particles, i.e. 1.5 µg, which was evenly dispensed during media

changes, i.e. 214 ng per change. This strategy inferred that the full dose was

dispensed after 3 weeks which could only be rigorously compared to the

microparticles if they had fully degraded, which was not the case. Slow degradation


- 173 -

was due to the polymer used here, PLGA 85:15, which contained a high lactide

fraction. The 10% PEG in the formulation did not increase degradation significantly

over the 3-week study. To conclude, it is more relevant to compare the ALP readings

from particles with the negative controls (NC), which showed that released BMP-7

was still active and triggering cell differentiation after 3 weeks.

Finally, an interesting result from this study is the similar effect of bolus (PC2)

and sustained delivery (PC3) of fresh BMP-7, which, except at day 14, did not show

statistical differences during the 3-week period (Figure 5.8), implying here that there

were no long-term benefits in the sustained delivery compared to the bolus delivery

in terms of stimulating osteogenic differentiation of MC3T3-E1 cells. However, the

effect of sustained delivery at 14 days was highly significant, proving that a constant

stimulus with fresh BMP-7 up to 2 weeks provided differentiated cells sooner than

with the bolus delivery. This phenomenon is explained by BMP-7 molecules getting

rapidly degraded with the bolus delivery, due to the short half-life of BMPs delivered

in vitro in solution [30]. Ideally, by using release systems such as the electrosprayed

particles presented here, that can lower the daily dose while ensuring bioactivity of

released GFs, we will avoid unnecessary overcrowding within the defect site, and

dispense lower doses of growth factors over longer timeframes than exhibited by

bolus deliveries. This may in turn lead to safer and more efficacious GF treatments,

which are also less expensive owing to containment of lower doses of GF.

5.7 CONCLUSIONS

In this study, bone growth factors; VEGF and BMP-7, were encapsulated into PLGA

85:15 microparticles by electrospraying and assessed by PBS-based assays and

cellular assays. Fundamental differences were observed, where quantification

procedures that did not involve cells led to GF interactions, which impaired accurate

ELISA detection and biased the actual results. When processed GFs were tested with

cells, GF interactions were indicated to be non-covalent since GF bioactivity was

verified at all steps of microparticle processing (involving micronisation and contact

with organic solvent), although ELISA recovery was lower for both VEGF and

BMP-7 after contact with chloroform. Similarly, the presence of trehalose in the

microparticle formulations did not affect GF bioactivity, although it negatively

impacted GF detection via ELISA in encapsulation and release assays in PBS, hence


- 174 -

inferring that PEG was sufficient to protect GF bioactivity in electrosprayed

particles. The positive effect of BMP-7 loaded electrosprayed particles in

differentiating pre-osteoblasts up to 3 weeks provided further proof of the

discrepancies between release assays in PBS and measured by ELISA, and the actual

effectiveness seen with direct contact with cells. Taken together, these important

results highlight the importance of looking at appropriate analysis techniques for GF

delivery, which is a complex undertaking with multiple interactions. A major change

in the assessment of microparticle systems containing intricate molecules such as

growth factors may be needed, hopefully paving the way to further development and

use of cell-based assays to accurately evaluate protein-carrier formulations.


The authors wish to thank Prof. George Muscat from the University of Queensland

for providing C2C12 cells and Dr. Mary Wang for help with handling. Thanks to the

Australian Research Council (ARC), LP130100945, for financial support. N.B. also

acknowledges the financial support from QUT in the form of an Australian

Postgraduate Award scholarship, and top-up from the Deputy Vice Chancellor.

M.A.W. acknowledges support from the ARC LP100200084.

5.9 SUPPORTING INFORMATION

5.9.1 Culture Conditions for GF Bioactivity Assessment

5.9.1.1 VEGF

The effect of endothelial growth supplement (ECGS) and foetal calf serum (FCS) on

the proliferation of human umbilical vein endothelial cells (HUVECs) was assessed,

for a starting seeding density of 3,000 cells/well in a 96-well plate and after 3 days of

incubation. Results are presented in Figure S5.9a.

Section 5.9 Supporting Information

- 175 -

Figure S5.9. The Effect of FCS amount, SA and ECGS on Proliferation of HUVECs. Means ± SE, n =

5.

There was a significant impact of ECGS on HUVECs (p < 0.001), where the addition

of the supplement increased cell proliferation in a four-fold manner, for any amount

of FCS (5 or 10%). Without ECGS, the addition of 10% FCS instead of 5% had a

slight, but significant (p = 0.021), negative effect on cells. It could be concluded that

the ECGS supplement was the most critical factor to proliferation of HUVECs cells.

Since serum albumin (SA) was used as an excipient in electrosprayed microparticles,

its effect on cell proliferation was also verified (Figure S5.9b). It was observed that

the presence of SA slightly decreased cell proliferation, but not in significantly

manner (p = 0.4).

Figure S5.10 shows the proliferation results of HUVECs for increasing doses of

fresh VEGF. When ECGS was present in the medium, the addition of extra VEGF

led to a decrease in cell proliferation (Figure S5.10), representative of an excessive

amount of GFs in solution. In the absence of ECGS, the exogenous VEGF delivery

significantly increased cell numbers at all concentrations; 4, 8 and 20 ng/mL, in a

linear dose-dependent manner (Figure S5.10A).

Figure S5.10. The Effect of VEGF Dose on Proliferation of HUVECs. Means ± SE, n = 5.

0

200

400

600

800

1000

(-)ECGS (+)ECGS

DN

A c

on

ce

ntr

atio

n (

ng

/mL

)

5% FCS

10% FCS

0

200

400

600

800

1000

(-)ECGS (+)ECGS

DN

A c

oncentr

ation (

ng/m

L)

5% FCS

5% FCS with SA

A B

A B

100

200

300

400

500

0 4 8 12 16 20 24

DN

A c

oncentr

ation (

ng/m

L)

VEGF Concentration (ng/mL)

(-) ECGS

200

300

400

500

600

700

0 4 8 12 16 20 24

DN

A c

once

ntr

ation (

ng/m

L)

VEGF Concentration (ng/mL)

(+) ECGS


- 176 -

In conclusion, the best conditions for using a HUVEC proliferation assay to assess

VEGF bioactivity were:

to use VEGF concentrations between 4 and 20 ng/mL,

in a medium containing 5% FCS and no ECGS,

for a minimum of 3,000 cells/well seeding density, since the use of 1,000

cells/well showed to be initially insufficient to provide a notable cell response

(data no shown) and,

for 3 days of incubation before DNA content analysis.

5.9.1.2 BMP-7

First, the effect of medium on C2C12 proliferation was studied (Figure S5.11). A

C2C12 seeding density of 5,000 cells/well in a 48-well plate was initially used.

Figure S5.11. The Effect of Medium on C2C12 Proliferation. (A) DNA concentration (ng/ml) and (B)

cell proliferation % for different mediums. C2C12 cells were cultured for 5 days and treatments were

subjected to cells twice on day 1 and day 3. Means ± SE, n = 6 (p = 0.051).

Figure S5.11 shows that the presence of half volume of PBS in the culture medium

decreased cell proliferation to 78%, compared to full medium (non-significant), and

that the presence of PEG and SA in PBS provided similar proliferation results (80%).

This shows that there was no inhibition of cell proliferation compared to PBS alone,

and thus PEG and SA could be considered non-cytotoxic to C2C12 cells, as

expected.

Next, the effective concentration range of BMP-7 was determined in terms of cell

proliferation and ALP activity. Results are presented in Figure S5.12 and show that

proliferation of C2C12 was similar or superior to controls up to 1.4 µg/mL of BMP-

7. ALP activity was significantly upregulated for 0.7 µg/mL and increased further for

1.4 µg/mL. An excessively higher dose of BMP-7, 3.3 µg/mL, was shown to reduce

both proliferation (40%) and ALP expression. This is due to the down-regulation of

cell receptors on the surface of cells with higher amounts of GFs in solution, which

0

2000

4000

6000

Full Medium

Half PBS Half PBS + PEG and

SA

DN

A c

oncentr

ation (

ng/m

L)

0

50

100

150

Full Medium

Half PBS Half PBS + PEG and

SA

Cell

pro

lifera

tion (

%)

A B


- 177 -

lead cells to senescence or die. Here, the effective concentration range of BMP-7 to

be used with C2C12 is thus found to be between 0.2 [208] and 1.4 µg/mL.

Figure S5.12. The Effect of BMP-7 Concentration on; (A) proliferation and (B) ALP expression of

C2C12 cells cultured for 5 days. Treatments were subjected to cells twice on day 1 and day 3. Means

± SE, n = 6.

5.9.2 Particle Microstructure

Figure S5.13. SEM images and particle size distributions of PLGA:PEG microparticles loaded with

1% wt of SA without addition of trehalose (a, c) and with addition of 1% wt trehalose (b, d). Mean

sizes ± standard deviations (SD), n = 150-200.

0

50

100

150

200

0.0 0.7 1.4 2.1 2.8 3.5

Pro

lifera

tion (%

)

Concentration (µg/mL)

0.0E+00

5.0E-04

1.0E-03

1.5E-03

2.0E-03

2.5E-03

3.0E-03

0.0 0.7 1.4 2.1 2.8 3.5

ALP

activity

(pN

PA

BS

/DN

A c

onte

nt)

Concentration (µg/mL)

A B

B2 Formulation: No trehalose BT2 Formulation: 1% trehalose

0

20

40

60

80

0 2 5 7 10 12

Num

ber

of P

art

icle

s (

%)


0

20

40

60

80

0 2 5 7 10 12

Num

ber

of P

art

icle

s (

%)


A B

C DMean: 5.0 1.3 µm Mean: 5.7 µm 1.6 µm

10 µm 10 µm


- 178 -

5.9.3 In Vitro GF Release

Figure S5.14. The Effect of Surfactant in Release Medium. Release profiles of BMP-7-loaded PLGA

microparticles (BT1 formulation – low loading of BMP-7, trehalose in formulation). Means ± SE, n =

3.

5.9.4 The Effect of Freeze-Drying on BMP-7

Figure S5.15. The Effect of BMP-7 Freeze-Drying on BMP-7 detection by ELISA (n = 3) and ALP

expression of C2C12 cells (n = 6). Means ± SE.

5.9.5 In Vitro Microparticle 2D Culture

5.9.5.1 Preliminary Culture

The in vitro effect of microparticle amount on proliferation and differentiation of

MC3T3-E1 cells was assessed for the B2 formulation (BMP-7 highly-loaded

particles) and the Blank formulation (unloaded particles). Concentrations of 20,000

cells/well (500 µL/well) were seeded in 24-well plates and cells were let to adhere

for 24 h before treatments, which are presented in Table S5.5. Cells were grown for 7

days before analysis of DNA content and ALP expression. Half the volume of media

was changed twice before analysis. Results are presented in Figure S5.16.

0

5

10

15

20

0 3 6 9 12 15 18 21Cu

mu

lative

BM

P-7

re

lea

se

(%

)

Time (days)

BT1 in PBS + 0.05% PS20

BT1 in PBS

1.0E-03

1.5E-03

2.0E-03

2.5E-03

3.0E-03

0

100

200

300

400

500

600

700

Before FD After FD

AL

P a

ctivity

(pN

P A

bso

rba

nce

/ D

NA

co

nte

nt)

BM

P-7

Co

nce

ntr

aio

n (

ng

/mL

)

BMP-7 Concentration

ALP activity


- 179 -

Table S5.5. Experimental conditions.

Negative Control NC Normal Media

Positive Control PC Osteogenic media

Condition 1 0.5 mg (Blank) 0.5 mg/well unloaded particles

Condition 2 5 mg (Blank) 5 mg/well unloaded particles

Condition 3 0.5 mg (BMP7) 0.5 mg/well BMP-7 loaded particles

Condition 4 5mg (BMP7) 5 mg/well BMP-7 loaded particles

Figure S5.16. (A) Proliferation and (B) Differentiation Results of MC3T3-E1 Cells after Direct

Contact with Microparticles. Proliferation was assessed by PicoGreen® (n = 4) and differentiation

was assessed by ALP expression normalised to DNA content (n = 4). (* and # indicate p < 0.05

compared to NC and PC, respectively, and † indicates p < 0.05 compared to PC).

Proliferation was enhanced when the osteogenic media was used (p = 0.004). There

were no statistical differences on proliferation between the loaded and unloaded

particles groups (p = 0.142). The 0.5 mg/well groups performed equally than their

respective controls (p > 0.7), but 5 mg/well significantly lowered proliferation (p <

0.001). ALP expression, indicative of the effectiveness of BMP-7 released from

microparticles, was significantly higher for the 5 mg/well group, but no differences

were observed for 0.5 mg/well.

In conclusion, the use of 5 mg/well was sufficient to trigger significant ALP

expression from MC3T3-E1, however cell proliferation was significantly impaired.

In addition, with such a high amount of particles in wells, it was impossible to

observe cell monolayers with optical microscopy, due to heavy particle coverage.

Because 0.5 mg/well was not sufficient to induce significant ALP expression, an

intermediate value of 2.5 mg/well was recommended, to ensure significant ALP

expression, while minimising proliferation damage and allowing cell imaging.

0.0E+00

5.0E+03

1.0E+04

1.5E+04

2.0E+04

2.5E+04

NC 0.5 mg Blank

5 mg Blank

PC 0.5 mg BMP-7

5 mg BMP-7

DN

A c

on

ce

ntr

atio

n (

ng

/mL

)

*#

ns

ns

0.E+00

2.E-05

4.E-05

6.E-05

8.E-05

1.E-04

1.E-04

NC 0.5 mg Blank

5 mg Blank

PC 0.5 mg BMP-7

5 mg BMP-7

pN

P A

bso

rba

nce

/DN

A c

on

ten

t

†

ns

A B


- 180 -

5.9.5.2 Final Culture

Figure S5.17. Optical microscopy images of tissue culture wells after 14 days of MC3T3-E1 culture

for controls (left) and treatments (right) with 2.5 mg particles per well.

Normal media Osteogenic media

Bolus BMP-7 Delivery Sustained BMP-7 Delivery

BM

P-7

Lo

ad

ed

No trehalose

A) Controls B) Particles

Un

loa

de

d

1% trehalose

50 µm 50 µm

50 µm 50 µm

50 µm 50 µm

50 µm 50 µm

- 181 -

Chapter 6: Composites for Delivery of

Therapeutics: Combining Melt

Electrospun Scaffolds with Loaded

Electrosprayed Microparticles

Nathalie Bock1,3,4

, Maria A. Woodruff1, Roland Steck

2, Dietmar W.

Hutmacher3, Brooke L. Farrugia

4, Tim R. Dargaville

4

Published in Macromolecular Bioscience, Volume 14, Issue 2, 2014, Pages 202-

2014.

© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. All rights reserved.




Designed and performed most experiments





Roland Steck* Performed µCT experiments and analysis


2 Trauma Research Group





Chapter 6 Electrosprayed/Electrospun Composites for Delivery ofTherapeutics

- 182 -



Brooke L. Farrugia* Assisted in release sample collection

Provided some technical guidance



Conceived aspects of the experimental design

Assisted in GPC and DSC analysis














requirements.





- 183 -

6.1 ABSTRACT

Delivery of therapeutics from structural scaffolds is an emerging strategy for guiding

cells towards regeneration of tissues, however difficulties arise when encapsulating

therapeutics directly within scaffolds. Here a novel strategy is reported to produce

polycaprolactone microfibre-scaffolds independently layered with high densities of

poly(lactic-co-glycolic acid) microparticles encapsulating a model protein. The use

of melt electrospun scaffolds confers high porosity while direct electrospraying

provides reproducible scaffold coating throughout the entire architecture. The burst

release is significantly reduced when compared with release from microparticles free

in solution, due to the immobilisation of microparticles on the surface of the scaffold.

The degradation of microparticles is dependent on protein-polymer interactions,

influencing the release mechanisms. The novel composite scaffolds have a positive

biological effect in contact with precursor osteoblast cells up to 18 days in culture.

The scaffold design achieved with the techniques presented here makes these new

composite scaffolds promising templates for growth factor delivery.

6.2 KEYWORDS

Electrospraying, drug delivery, polymer-drug interactions, microstructures, tissue

engineering.

6.3 INTRODUCTION

Biodegradable polymeric scaffolds have become crucial in the arena of tissue

engineering (TE), where they provide a temporary porous structure for a specific

tissue to re-grow [213]. Fibre-based scaffolds have been extensively studied for this

purpose based on the popularity of electrospinning technologies for the fabrication of

fibres on the nano- to micron scale with good control over the physico-chemical

properties [143, 214]. Scaffold porosity and architecture, as well as fibre diameter

and fibre arrangement are essential variables enabling cell invasiveness [21, 131].

Such properties can reproducibly be controlled by electrospinning polymer solutions

and melts [215], in static and direct writing modes [216-218]. While electrospun

matrices may provide physical cues for cells, there is also a need to deliver biological


- 184 -

cues that effectively direct cellular growth and differentiation, for the regeneration of

complex tissues such as skin, cartilage and bone [219]. This can be achieved by

delivering exogenous bioactive molecules, for instance growth factors (GF), which

are embedded within fibres and available for cellular uptake upon matrix degradation

[22-24]. While many studies have considered this issue, as reviewed by Szentivanyi

et al. [20], the direct incorporation of proteins into fibres is not yet optimised for

load-bearing scaffolds, as required in bone TE for instance, where GFs should ideally

be sustainably released over many weeks, while the scaffold can still maintain its

structural function [145, 220]. The addition of proteins directly to electrospinning

solutions can also result in poor fibre mesh properties and instabilities in the cone-jet

[25, 104]. To overcome these limitations, one approach is to incorporate a separate

release system into the scaffold [24, 181].

Polymeric nano- and microparticles are suitable candidates for incorporation

into/onto TE constructs since both scaffold structure and GF release requirements

can be taken into account [18]. One advantage of encapsulating GFs is that multiple

GF release profiles can be achieved by judicious choice of polymers’ properties and

processing parameters [16]. For instance, bone morphogenetic protein-2 (BMP-2)

and insulin-like GF-1 (IGF-1) have been simultaneously delivered from

microparticles embedded in a hydrogel scaffold demonstrating different release

patterns [146]. The secure fixation of loaded particles to fibre-based scaffolds,

however, remains a challenge, but is an important consideration to prevent particle

loss during implantation [24]. Ionescu et al. have previously reported dual fabrication

techniques that enabled loaded microparticles to be successfully incorporated into

polycaprolactone (PCL) electrospun nanofibres [144]. To do this, they

simultaneously electrospun PCL with a sacrificial solution of polyethylene oxide

(PEO), containing the loaded particles. Upon removal of the PEO, the microparticles

remained entrapped within the PCL nanofibres [144]. Similarly, a co-

electrospraying/electrospinning strategy reported by Wang et al. allowed the direct

formation of a scaffold from polyurethaneurea nanofibres and poly(lactic-co-glycolic

acid) (PLGA) microcapsules containing an IGF-1 gelatin solution [29]. While this

approach provided interesting morphologies, only low densities of particles could be

loaded and thus minimal IGF-1 was released from the nanofibres [29].

Electrospraying is an emerging strategy to load therapeutic molecules into

polymeric particles [181]. While the electrospraying process follows the same


- 185 -

principles of solution electrospinning, the polymer concentration used is below a

critical concentration which would be required to form fibres and as such, this results

in the jet breaking up into droplets [102]. Electrospraying provides high control over

particle size distributions and morphology with control of polymer solution and

processing parameters [83, 221]. A large variety of molecules can be encapsulated in

electrosprayed particles, such as anti-cancer and inhalation drugs, antibiotics,

proteins and GFs [64]. In the context of fibre-scaffolds, we hypothesise that

electrospraying may be used to directly coat fibres with loaded particles without the

use of additives, while conferring high control over the characteristics of the final

particles.

Here we present unique composite scaffolds which incorporate a microparticle

protein-delivery system produced using electrospraying directly onto melt

electrospun scaffolds. The use of microfibre-scaffolds produced by melt

electrospinning allows for a greater degree of pore size and interconnectivity

available for particle loading than is possible with solution electrospun nanofibres

[218], meaning that high densities of loaded microparticles can be incorporated into

the final scaffolds. The steps involved in optimising a homogenous coating and high

yield of PLGA microparticles loaded with serum albumin (SA) onto PCL microfibres

and the final physical properties of the constructs are described hereafter. SA release

profiles and degradation characteristics of composite scaffolds in solution are

reported over 4 months and compared with free electrosprayed particles which are

not attached to the scaffolds. Preliminary positive biological effects of composites

scaffolds on precursor osteoblast cells are assessed in vitro up to 18 days.


6.4.1 Scaffold Fabrication

6.4.1.1 Materials

Poly(lactic-co-glycolic acid) 85:15 (Mn 41.3 kg/mol, PDI 1.6) was purchased from

Evonik Industries, USA. Dichloromethane (DCM) and serum albumin (SA) were

purchased from Sigma-Aldrich, Australia. Polycaprolactone (Mn 41 kg/mol, PDI

1.78) was donated by Perstorp Ltd, UK.


- 186 -

6.4.1.2 Raw Scaffold Fabrication

Melt PCL scaffolds (14 mm diameter) were produced using an in–house built

apparatus with a circulating water system as a heater described elsewhere [222]. PCL

was heated at 70°C, electrospun at a 10 µL/h flow rate from a plastic Luer-lock 2 mL

syringe (B-Braun, Australia) fitted with a blunt 23 G needle placed 30 mm above a

grounded aluminium plate while a voltage of 6 kV was applied to the needle tip. The

resultant scaffolds were placed in 70% ethanol (EtOH) for 30 min under vacuum,

followed by immersion in sodium hydroxide (NaOH) 2 M for 1 h at 37°C to reduce

surface hydrophobicity. The scaffolds were then rinsed with deionised water until the

pH level dropped to 7. Finally, scaffolds were placed in phosphate buffer saline

(PBS) for 30 min and dried.

6.4.1.3 Composite Scaffold Fabrication

Composite scaffolds were obtained by direct electrospraying onto melt scaffolds.

Five PCL scaffolds were fixed with conductive double-face carbon tape on a

grounded aluminium disc, 48 mm in diameter, or 150 × 150 mm2 aluminium plate

recovered with aluminium foil, placed 150 mm away from the tip (depicted

schematically in Figure S6.8, supporting information). Reproducible PLGA

microparticles were generated by working in the semi-dilute entangled regime of

electrospraying [181]. Briefly, PLGA was dissolved in DCM (10% wt/v), loaded in a

1 mL glass syringe (Hamilton, USA) and extruded through a 21 G blunt needle at a

rate of 0.8 mL/h from a syringe pump (WPI, USA), while a voltage of 10 kV was

applied to the needle tip. For SA loading, the PLGA solution was added to

lyophilised SA under magnetic stirring (10% wt/v PLGA loaded with 1% wt SA).

The resultant dispersion was probe sonicated for 60 s at 0.5 W (Misonix 3,000,

USA). Conductivity was measured using a conductivity meter (TPS 900C,

Australia). The dispersion was electrosprayed in the same conditions as the SA-free

polymer solution. All scaffolds and particles were placed in a dessicator under pump-

aided vacuum overnight and stored at -18°C until further use.


6.4.2.1 Size

Particle size was assessed with ImageJ analysis software (National Institutes of

Health (NIH)) based on scanning electron microscope (SEM) micrographs. 6 random

composite scaffolds were selected per condition and 30 particles were measured per


- 187 -

scaffold (n = 30/scaffold). 4 random samples of collected particles were selected and

45 particles were measured (n = 45/sample). Fibre size was measured identically on

raw PCL scaffolds, where 6 scaffolds were selected and 40 fibres were measured (n

= 40/scaffold).

6.4.2.2 Morphology and Microstructure

Morphology was assessed by SEM and micro-computed tomography (µCT). For

SEM, materials were taped on aluminium stubs and gold coated for 225 s at 30 mA

(SC500, Bio-Rad, Australia). The morphology and microstructure of scaffolds were

characterised with a FEI Quanta 200 SEM operating at 10 kV in high vacuum mode.

For µCT, scaffolds were scanned (µCT 40, Scanco Medical, Switzerland) in air at an

energy of 45 kVp and intensity of 177 µA with 300 ms integration time. The scans

were reconstructed to 3D datasets with an isotropic voxel size of 6 µm. After

segmentation, the average fibre spacing was determined by applying a bone

morphometric analysis algorithm using the distance transformation method [223]

with the scanner’s software (µCT Evaluation Program V6.5-1, Scanco Medical,

Switzerland). Scaffold porosity was obtained by determining the ratio of volume

occupied by the fibres and particles to the total volume scanned. The results were

expressed as means ± 1 standard deviation (SD).


6.4.3.1 SA Content

SA content in the particles was determined by an extraction procedure. Particles (20

mg) were dissolved in DCM (1 mL), n = 4. PBS (3 mL) was added to the dispersions

and tubes were vortexed for 30 s to extract the SA. The resultant emulsions were

centrifuged at 5,000 rpm for 15 min and left overnight. The aqueous phase was

collected and another extraction cycle was performed. SA content was determined by

the micro-bicinchoninic acid (µBCA) assay (Thermo Fisher Scientific, USA).

Encapsulation efficiency (%) was measured as: [Actual SA content

(µg)]/[Theoretical SA content]×100.

6.4.3.2 Release Studies

In vitro release studies were performed by placing particles and composites that

contained the same amount of particles (8.5 mg), SA-free and SA-loaded, in 2 mL

screw-capped microtubes with PBS (1 mL), n = 3. Tubes were agitated at a speed of


- 188 -

8 rpm at 37°C for 130 days. At specific time points, tubes were centrifuged at 14,000

rpm for 5 min to settle particles before 970 µL of supernatant was collected and

replaced by the same amount of fresh PBS. The supernatant was analysed by the

µBCA assay (Thermo Fisher Scientific, USA). A standard curve was prepared by

serial dilutions of the supplied SA from 40 to 0 µg/mL. A polynomial fit was

deducted from the corresponding absorbance readings at 562 nm (R2 = 0.9989)

(Microplate Manager V5.2, Benchmark Plus spectrophotometer, Bio-Rad, USA).

The absorbance measured from released SA was normalised to the reading from SA-

free particles/scaffolds degraded to the same time point, in order to account for the

presence of PLGA degradation products in the release media.

6.4.3.3 Polymer Degradation

Degradation was assessed by placing the same materials and amounts as for release

studies in PBS, under the same conditions. 3 scaffolds and 3 samples of collected

particles were assessed at each time point (n = 3) and uncoated raw PCL scaffolds

were assessed as well. While 970 µL of PBS was replaced for all samples at all time

points, the samples from each specific time point were washed twice with water,

centrifuged at 14,000 rpm for 5 min, and vacuum-dried. Mass loss was determined

gravimetrically as % Mass Loss = [(M1-M2)/M1] ×100, where M1 was the initial mass

and M2 the final mass of particles. Number-average molecular weight (Mn), weight-

average molecular weight (Mw) and polydispersity were determined by gel

permeation chromatography (GPC) in chloroform. All samples were dissolved in

chloroform and SA-loaded samples were further filtered (0.45 µm pore). Solutions

were injected (250 µL) onto Styragel HR columns with a flow rate operating at 1

mL/min. Calibration was done by the use of polystyrene standards ranging from

1,350 to 382.1 kg/mol. Glass transitions of free particles were measured by

differential scanning calorimetry (DSC) on a TA instruments Q100 DSC instrument.

Samples (1-6 mg) were scanned twice from 0°C to 250°C at a heating/cooling rate of

20°C/min. Degraded morphology was assessed by SEM under the same conditions as

non-degraded samples.

6.4.3.4 Statistical Analysis

Statistical analysis was performed with PASW Statistics 18. For size, analysis was

done on means using an independent Student’s t-test assuming equal variances, after

Levene’s test confirmed equality of variances (0.411). For SA content, a Mann-


- 189 -

Whitney non-parametric test was done on medians. The significance level was

determined for p < 0.05.

6.4.4 Biological Evaluation

6.4.4.1 Cell Culture

The biological effect of scaffolds produced here was assessed with a mouse

osteoblast precursor cell line, MC3T3-E1 (passage 9). Prior to cell seeding, MC3T3

cells were cultured in growth medium: α-MEM cell culture medium supplemented

with 10% foetal calf serum (FCS), and 1% penicillin/streptomycin (P/S) (all from

Invitrogen, Australia) at 37°C and 5% CO2. All scaffolds were cut into 3 symmetrical

samples and weighed. Scaffolds were rinsed with 70% EtOH and sterilised under

ultraviolet (UV) radiation for 20 min on each side. Dry, sterile scaffolds were

transferred into 24-well plates and incubated each for 1 h with 20 µL of serum-free

medium. MC3T3 cells (4,500) suspended in culture medium (40 µL) were equally

distributed onto each scaffold using a top seeding static method. Cell-scaffold

constructs were incubated for 2 h at 37°C, during which initial cell attachment to the

scaffolds occurred. Culture medium (1 mL) was then carefully added to each well.

MC3T3 cells were maintained in 5% CO2 at 37°C for 18 days in a humidified

incubator. Medium was changed every 2-3 days and selected constructs were fixed

and prepared for analysis as described below after 24 h, 9 days and 18 days of cell

culture.

6.4.4.2 Cell Viability

Cell viability was assessed by a LIVE/DEAD staining assay with fluorescein

diacetate (FDAC) and propidium iodide (PI) (both from Invitrogen).[224] Cell-

scaffold constructs were washed twice with PBS, followed by incubation in FDAC

(0.67 µg/mL) and PI (5 µg/mL) solution (1 mL) for 5 min at 37°C in the dark. After

washing with PBS, specimens were immediately imaged using a Zeiss Axio M2

Imager (Zeiss, Germany) fluorescent microscope.

6.4.4.3 Cell Morphology

The morphology of actin fibres and nuclei of MC3T3 cells on the scaffolds was

visualised using confocal laser scanning microscopy (CLSM). Cell-scaffold

constructs were removed from the media, washed twice with PBS (containing Mg2+

and Ca2+

), fixed in paraformaldehyde (4%) for 20 min at room temperature (RT),


- 190 -

permeabilised in Triton X-100 (0.2%) (Invitrogen)/PBS for 5 min at RT, and

incubated in SA/PBS (0.5%) (Sigma) for 10 min at 37°C to block non-specific

binding sites. Between each step, scaffolds were washed twice with PBS. Samples

were then incubated for 45 min with SA/PBS (0.5%) solution containing rhodamine-

conjugated phalloidin (200 U/ml) and 4’,6-diamino-2-phenylindole (DAPI) (5

μg/mL), (Invitrogen). After washing with PBS, constructs were stored in PBS until

imaging. Fluorescence images were captured using a Leica TCS SP5 confocal laser

scanning microscope (Leica Microsystems, Wetzlar, Germany). SEM was used to

investigate cellular attachment and morphology on the scaffolds. Cell-scaffold

constructs were fixed with glutaraldehyde (3%) in cacodylate buffer (0.1 M) (Sigma)

at 4°C overnight. Fixed specimens were washed in sodium cacodylate buffer (0.1 M),

osmium tetroxide (1%), deionised water, and dehydrated through a graded series of

EtOH before incubation in hexamethyldisilazane (HMDS) twice, for 30 min (all

reagents were supplied by ProSciTech, Australia). After full air-drying, constructs

were mounted and gold sputter-coated (SC500, Bio-Rad, Australia) prior to

visualisation with a FEI Quanta 200 SEM, using an accelerating voltage of 10 kV.


6.5.1 Fabrication and Physical Characterisation

6.5.1.1 Fabrication

The composite scaffolds were fabricated in a two-step process, both using

electrohydrodynamic techniques. Firstly, 14 mm diameter PCL microfibre scaffolds

were fabricated by melt electrospinning. The scaffolds were fixed on a static

conductive collector for subsequent coating with unloaded and loaded (1% wt SA)

electrosprayed PLGA particles. Traditionally, a static collector that is larger than the

electrospraying cone is used to collect the particles when electrospraying (Figure

S6.8, supporting information), in order to ensure a high yield. However, when

coating scaffolds with electrosprayed particles, a focused cone is more desirable for

reproducible and constant coating of scaffold areas and limited loss to the

surrounding collector. This can be challenging due to drifting of the electrospraying

cone and shielding effects of the polymer. Indeed, during electrospraying, the

collector is covered with increased amounts of polymer carrying electrical charge

from the high voltage supply. Over time, residual charge generates charge build-up,


- 191 -

resulting in an increased electrosprayed area [225]. The use of a secondary electrode

to focus the jet has been reported, however this only temporarily overcomes the

problem [226]. Charge build-up was evident here by non-reproducible

electrospraying patterns and low yields being deposited on the scaffolds. To

overcome this issue, a smaller 4.8 cm diameter, aluminium disc collector, shown in

Figure S6.8, was used to focus the electrospraying cone and improve stability.

Placing the PCL scaffolds onto a smaller disc collector enabled the cone to focus

better and deposit the microparticles more evenly, effectively minimising deposition

on the collector areas which did not contain a scaffold.

When electrospraying, the solvent contained in the polymer droplets evaporate

during flight towards the collector. Therefore by tailoring the polymer concentration

(10% wt/v in this instance), particles may still contain a small amount of residual

solvent when impacting onto the scaffolds, thus enabling attachment to the fibres,

before full solvent evaporation. Owing to the high porosity of the microfibre

scaffolds and the attractive grounded collector underneath, particles were able to

penetrate within the scaffolds and coat them through their entire depth, 415 µm on

average, when electrospraying for as short a period as one hour, equating to 80 mg of

electrosprayed PLGA (Figure 6.1). A comparison of the deposition obtained with the

traditional setup is shown in Figure S6.8e-f. Such a feature represents a significant

achievement in the field and has not previously been reported for equivalent

thickness solution electrospun nanofibres due to their lower porosity [159] hence our

new approach lends well to cell culture and should better enable cell penetration.

1000 µm 500 µm

5 µm

A B

C D

50 µm


- 192 -

Figure 6.1. Overview of a PCL scaffold coated with electrosprayed PLGA particles after 1 h of

electrospraying (80 mg PLGA), viewed by; (A) µCT, (B-D) SEM, at different magnifications.

6.5.1.2 Physical Properties

The raw PCL melt electrospun scaffolds had an average fibre diameter of 16.2 ± 1.8

µm and were 415 ± 59 µm thick, based on analysis using SEM/ImageJ and µCT,

respectively. PLGA particles electrosprayed on PCL scaffolds were 8.4 ± 1.6 µm and

6.1 ± 0.9 µm for 1% SA-loaded PLGA particles and non-loaded PLGA particles,

respectively, showing a size range within the same order for particles and fibres.

With the electrospraying parameters chosen here, spherical particles with quasi-

monodispersity were obtained (Figure S6.9). The increase in size for loaded particles

was due to a decrease of the solution conductivity after addition of protein (0.035

µS/cm compared to 0.07 µS/cm), yet no differences in surface morphology were

observed. Importantly, the size distributions of particles found either at the surface of

the scaffold and at the furthest distance from the surface (i.e. the underside) were

statistically equivalent (p = 0.39) showing an overall deposition of homogeneously-

sized particles throughout the scaffolds’ thickness. This is an important requirement

since size distribution ultimately dictates release profiles and thus reproducibility in

size and distribution is paramount.

There are two ways to vary the amount of protein loaded onto the composite

scaffolds, namely to change the amount of protein loaded into each microparticle, or

to change the total number of microparticles deposited onto the scaffold. To

investigate this second possibility we electrosprayed SA-loaded PLGA particles on

PCL scaffolds for durations of 1, 2, 3 and 8 h at a constant flow rate equating to 80

mg, 160 mg, 240 mg and 640 mg of PLGA microparticles, respectively (Figure 6.2).


- 193 -

Figure 6.2. Characteristics of composite scaffolds with alteration of electrospraying duration. (A-D)

SEM images showing composite morphology. (E-H) µCT scans of the central sections of scaffolds

showing interfibre distance. (I) Average porosity (n = 3), (J) Average pore size (n = 3), (K) Average

particle loading (n = 5). Error bars represent standard errors (SE).

Using the disc collector, the distribution of the particles was uniform throughout

the scaffold after both 1 and 2 h of electrospraying, due to the initial high porosity of

melt electrospun scaffolds, with the entire scaffold becoming coated (415 µm on

average thick), following 2 h of electrospraying. Additionally, both porosity and pore

size were still high with 83.1 ± 1.2% and 55.6 ± 27.3 µm, respectively, suitable for

enabling cell invasiveness [227]. Importantly, no deposition gradient was seen from

the front to the back of the scaffolds, maintaining homogeneous coating and porosity

throughout the scaffolds, as seen in Figure S6.10. When compared with

electrospraying for 1 h onto a traditional collector, the fibres were less

homogeneously coated due to the larger conductive area available for particle

deposition and thus reduced deposition on scaffolds. When electrospraying for more

than 3 h with the optimised disc collector, fibres became saturated with particles and

distribution was no longer uniform (Figure 6.2d). Additionally, a deposition gradient

was also observed (Figure S6.10) with the front surface becoming fully coated, thus

0

10

20

30

40

0 2 4 6 8

Electrospraying duration (hours)

10

30

50

70

90

110

130

0 2 4 6 8Electrospraying duration (hours)

Raw PCL fibres 1h electrospraying 2h electrospraying 3h electrospraying

0 228µm

Mo

rph

olo

gy

Inte

rfib

red

ista

nc

e

100 µm

1 mm 1 mm 1 mm 1 mm

0.6

0.7

0.8

0.9

1.0

0 2 4 6 8Electrospraying duration (hours)

A B C D

E F G H

I) Relative porosity J) Scaffold pore size (µm) K) Particle loading (mg)

100 µm 100 µm 100 µm


- 194 -

blocking the porosity which would otherwise enable further coating of deeper parts

of the scaffolds. This could be correlated with µCT measurements (Figure 6.2e-k)

where a decrease in porosity and pore size was observed for the first 2 hours of

electrospraying (160 mg of electrosprayed PLGA) but slowed following further

electrospraying, attributed to the surface porosity reduction not allowing particles to

penetrate through to the centre of the scaffolds. This was evidenced by measuring the

total yield YT, which refers to the amount of particles deposited on the collector

compared to the initial amount of electrosprayed material (Equation 6.1):

( ) ( )

( ) (6.1)

YT was similar for 1, 2 and 3 h of electrospraying, namely 68.9%, 72.6%, and 67.6%,

respectively, however, a yield of only 47.6% was measured for electrospraying for 8

h, showing that fewer particles were deposited onto the collector for prolonged

coating. This phenomenon was attributed to charge build-up, since the collector was

heavily coated with particles at this stage, resulting in particle deposition on non-

charged surfaces, i.e. apparatus housing and surrounding equipment. Importantly,

particle size and morphology remained identical for any given electrospraying

duration and were not affected by charge build-up.

The electrospraying process efficiency was assessed by considering the scaffold

yield, YS, which is the amount of particles on the scaffold area compared to the total

amount of particles deposited on the collector (Equation 6.2):

( ) ( )

( ) (6.2)

When electrospraying on a traditional 150 × 150 mm2 aluminium foil collector

(Figure S6.8b), YS was higher for SA-free particles (YS = 12.4%) compared with SA-

loaded particles (YS = 3.9%) after electrospraying for 1 h. These low yields illustrate

an effect from the type of solution electrosprayed (loaded, non-loaded) due to

conductivity differences. However, when the small disc collector was used, YS was

higher, namely 44.0 ± 5.1% and 42.0 ± 4.4% for SA-free and SA-loaded particles,

respectively, thus independent of solution conductivity and also independent of

electrospraying duration. This was consistent with the area occupied by scaffolds,

namely 42.5% of the collector, showing homogenous and non-preferential coating of

both scaffolds and collector areas.


- 195 -

Overall it was found that the optimised small disc collector configuration was

better suited to providing higher and more reproducible scaffolds yields than

standard, traditionally large collectors, independent of the conductivity of the

electrosprayed solution and electrospraying duration. The working window for

uniform coating of fibres was found from 0 to 160 mg of electrosprayed PLGA

(Table 6.1), above which, reduced deposition and lower reproducibility occurred, due

to charge build-up.

Table 6.1. Summary of the characteristics of the composite scaffolds. Averages are expressed as

means ± SD.

Electrospraying duration - 1 hour 2 hours

Amount of electrosprayed PLGA - 80 mg 160 mg

Total yield (YT) - 68.9% 72.6%

Amount of PLGA particles per PCL scaffold - 7.6 ± 1.0 mg 15.0 ± 2.0 mg

Scaffold porosity 92.9 ± 1.9% 90.6 ± 0.9% 83.1 ± 1.2%

Scaffold pore size 94.5 ± 37.6 µm 83.6 ± 38.2 µm 55.6 ± 27.3 µm

6.5.2 Protein Release and Polymer Degradation

PLGA undergoes degradation by bulk erosion upon exposure to aqueous media such

as phosphate buffer saline solution in vitro or physiological fluids in vivo, where

water penetrates faster than subsequent polymer chain cleavage [228]. While

enantiomeric and copolymer compositions of PLGA are critical factors in directing

the rate of degradation, additives also play an important role, although their role

remains controversial in existing literature [229]. While GF release from PLGA

matrices is a combination of GF diffusion and polymer degradation [18], polymer-

GF interactions can be critical considerations, although they are often overlooked

[230], The nature of a GF may influence degradation rate – acid GFs can accelerate

hydrolysis while basic GFs can neutralise acidic chains [229]. Similarly, the spatial

arrangement of a polymeric implant may lead to differences in degradation kinetics,

since size and shape are known to affect degradation, and increasing system size may

lead to more auto-catalysis effects [230]. In the next part of this study, we

investigated the in vitro degradation and protein release from composite scaffolds

over a 4 month study period and compared these with the results from free particles

not attached to the scaffolds. The use of the scaffold prevented aggregation of

particles, allowing us to study the influence of particle arrangement on release and

degradation. The influence of drug loading on polymer degradation was also


- 196 -

monitored and several characterisation methods were employed to measure the time-

dependent physical and morphological changes within the composite scaffolds

compared to free particles.

6.5.2.1 Protein Release

The release profiles of SA from composites and free particles are shown in Figure

6.3, both indicating a burst release within 10 days in solution, followed by a slower

sustained release up to 110 days. Importantly, the burst release from particles

immobilised on composites was only 18% compared to 55% of the total amount of

protein available for loading, for free particles (Figure 6.3a), corresponding to 58%

and 86% of the total amount released (Figure 6.3b), respectively. A higher release

rate from composites was then observed from 10 to 70 days which then stabilised

until 110 days. For free particles, very little release after the initial burst release was

observed.

Figure 6.3. Release profiles of SA from particles free in solution and particles from the composite

scaffolds. (A) Cumulative amount of SA released compared to the initial amount of SA used for

loading. (B) Cumulative amount of SA released compared to the total amount of SA released (mean ±

SE, n = 3).

Burst release is a common problem with protein-loaded PLGA devices [205] due

to diffusion of proteins near the surface of particles. Here, a lower burst release was

observed from particles which were attached to the composites since those particles

had an initial reduced surface area in contact with water compared to the free

particles. This may have served to seal some surface pores, hence leading to reduced

initial protein diffusivity. This is very important in the context of GF encapsulation

since bolus release of GFs can have detrimental side-effects, such as

supraphysiological doses, and should be avoided. These safety concerns have drawn

a lot of attention recently, with the clinical use of BMPs in particular, with increasing

numbers of reports of catastrophic complications associated with off label use and

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120 140

Cum

ula

tive S

A rele

ased (

%)

Time (Days)

Particles free in solutionParticles on composites

A B

40

50

60

70

80

90

100

110

0 20 40 60 80 100 120 140

Cum

ula

tive S

A rele

ased

norm

alis

ed

to tota

l S

A rele

ased (

%)

Time (Days)

Particles free in solutionParticles on composites


- 197 -

high doses [13, 14]. For instance, in a study by Mannion et al., one third of the

smallest commercially available dose of BMP-2 was used for lumbar fusion (1.4 mg

instead of 4.2 mg) and soaked in a collagen sponge before insertion in the disc space

[15]. However, even this lower dose was still too high and generated one case of

vertebral body osteolysis, two cases of asymptomatic heterotopic ossification and

two cases of perineural cyst formation, out of 36 patients [15]. This is not surprising

considering; that BMP is a potent stimulator of new bone formation, that every cell

in the body possesses a BMP receptor, and that only nanogram levels of BMP are

required for cellular stimulation [12]. The possibility of producing coatings using

electrospraying thus presents potential to load smaller quantities of GFs, specific to a

particular application, and reduce initial burst release.

On average, a composite scaffold loaded with 15 mg of electrosprayed PLGA

particles was able to release 46.5 µg of detectable SA, which would be considered

sufficient in most delivery devices that require only a few micrograms for therapeutic

effect. For instance the physiologically active range of vascular endothelial GF

(VEGF) necessary for angiogenesis is 2 to 6 ng/mL [121]. For chondrogenesis of

stem cells, only 5 ng/day of BMP-6 and transforming GF-ß3 are required [231].

However, in traditional fabrication techniques, a higher than necessary dose is used,

since there are issues with loss of bioactivity during processing [35]. Electrospraying

is more promising in this regard, since the process does not require water-in-oil

emulsions, known to be the main factor of protein denaturation in many

encapsulation approaches [35]. Thus, minute but efficient doses can be incorporated

when electrospraying which is a promising novel method for rapidly and

reproducibly adding therapeutic release systems to fibre-scaffolds.

6.5.2.2 Polymer Degradation

The composites scaffolds were made with two biodegradable polymers; PCL for the

fibres and PLGA 85:15 for the particles. While both polymers undergo bulk

degradation, PCL is semi-crystalline and PLGA amorphous, conferring a slower

(over years) and faster (over months) degradation pattern, respectively [91]. As a

result, PCL degradation was negligible over the period of study [232]. When placed

in PBS, all PLGA particles demonstrated a reduction in surface roughness by 10 days

and started coalescing from 30 days onwards, as shown with SEM observation

(Figure 6.4d and i). At each incubation time point, more coalescence and smoother


- 198 -

surfaces were observed, until a unique mass of polymer was shown at around 70 and

90 days for SA-free (Figure 6.4j) and SA-loaded (Figure 6.4f) formulations,

respectively, followed by significant auto-catalytic degradation (Figure 6.4l). No

differences in PLGA degradation were observed between free particles and particles

from the composites according to SEM and gel permeation chromatography analysis.

Figure 6.4. Degradation of free particles and composite scaffolds over 4 months in solution. (A-B)

Molecular weight (Mn) measured by GPC for; (A) free particles, (B) composite scaffolds. Since PLGA

and PCL had a similar Mn around 41 kg/mol, both peaks initially overlapped as one. As PLGA

degradation progressed, the unchanged PCL peak became visible. (C-I) SEM images of particles from

the composite scaffolds.

Strikingly, significant degradation differences were observed between SA-loaded

and SA-free formulations. Free particles and particles from the composites both

showed reduced degradation when loaded with SA, as shown by a delay in

morphological changes over time (Figure 6.4c-l) and a delay in the decrease of

PLGA molecular weight (Figure 6.4a-b). Similarly, larger shifts in the glass

BSA-FreeBSA-Loaded

Raw

70 days

90 days

110 days

130 days

100,000 10,000 1,000

Decreasing Mn (g mol-1)

A)

PLGA loaded

with SA

Raw

PLGA

Particles free in solution

PCL

PLGA loaded

with SARaw

PLGA

Raw 30 days 70 days 90 days

SA

-Load

ed

SA

-Fre

e

C

H

B) Composites

100,000 10,000 1,000

Decreasing Mn (g mol-1)

10 µm

D E F G

I J K L

10 µm

110 days


- 199 -

transition temperature (Tg) (Figure S6.11) were observed in the presence of SA in the

polymer structure reflecting more enthalpy relaxation processes [233-235], compared

to the particles without SA, with slower degradation and no significant Tg drop.

Since the degradation of PLGA microparticles was unaffected by the presence of

the scaffold, based on SEM and GPC results, the influence of SA on PLGA

degradation focused only on particles free in solution (the data on degradation of the

composite scaffolds can be found in supporting information (Table S6.3 to Table

S6.6)). When plotting molecular weight over time, two linear phases of degradation

were observed, confirming pseudo-first order degradation kinetics for PLGA (Figure

6.5a). The first phase lasted until 90 days and was slower than the following phase

from 90 to 130 days. For both phases, degradation was faster for SA-free particles as

shown by the higher degradation constants (kMw), which are presented in Table 6.2.

Polydispersity of polymer chains (Figure 6.5b) was steady until 30 days, with little

molecular weight decrease and individual particles were still distinguishable. From

30 to 90 days, the higher chain mobility, supported by particle coalescence observed

by SEM (Figure 6.4), corresponded to a moderate linear decrease in molecular

weight (Figure 6.5a). During this period, established degradation took place and

hydrolytic scission of the longer chains resulted in smaller oligomers, generating a

broader chain distribution, as reflected by increasing polydispersity (Figure 6.5b).

After 90 days, the degradation process accelerated due to auto-catalysis from the

degradation products, trapped within the polymer matrix. From SEM observations,

coalescence of particles peaked at this stage and the surface area was minimal

(Figure 6.4f and 4j). This resulted in increased degradation [236], represented by the

sharper degradation rate shown in Figure 6.5a from 90 to 130 days and the drastic

drop in Tg, especially for SA-free particles (Figure 6.5c). Here, the degradation rate

constant k2’ doubled for SA-free particles compared to SA-loaded particles (Table

6.2). Interestingly, little mass loss for any formulation was observed during the

period of study, with 82% mass still remaining after 130 days (Figure 6.5d). It is

known that although the molecular weight of PLGA decreases upon contact with

water, degraded polymer fragments need to obtain a critical molecular weight so that

they become soluble in the aqueous degradation media, causing mass loss [237, 238].

For example, in a degradation study by Blanco et al. the mass loss of PLGA

microspheres began from the critical Mn value of 5.4 kg/mol for PLGA 50:50 and 6.2

kg/mol for PLGA 75:25 [239]. The reason for such a delay in mass loss is attributed


- 200 -

to the hydrophobicity of PLGA 85:15 which reduced water uptake and thus reduced

the overall rate of in vitro degradation [240].

Figure 6.5. Physical characteristics of PLGA particles plotted over degradation time. a) Molecular

weight as measured by GPC, n = 2. Slopes are representative of degradation rate constants. b)

Polydispersity index as measured by GPC, n = 2. c) Glass transition temperature as measured by DSC,

n = 2. d) Relative mass remaining as measured by gravimetrical analysis, n = 3, mean ± SD.

Table 6.2. Degradation rate constants of PLGA particles in solution calculated from Figure 6.5a by

linear regression (R2 is the linear regression coefficient).

kMw value [10-3

/Day]

1st phase

(0 to 90 days)

2nd

phase

(90 to 130 days)

SA-loaded PLGA particles k1 = 6.3

(R2 = 0.944)

k2 = 22.7

(R2 = 0.998)

SA-free PLGA particles k1’ = 9.6

(R2 = 0.967)

k2’ = 45.5

(R2 = 0.982)

6.5.2.3 Protein-Polymer Interactions

We hypothesise that protein-polymer interactions took place, as evidenced by

incomplete SA release and slower PLGA degradation for particles containing the

protein. It is well accepted that protein aggregation/degradation can take place due to

possible acidification of the microenvironment or protein-polymer interactions

during particle formation, storage and release [35, 205]. Here, excessive acidification

of the environment caused by the accumulation of the degradation products within

the particles was not an issue since this phenomenon applies mostly to fast-degrading

75

80

85

90

95

100

0 20 40 60 80 100 120 140

Rela

tive m

ass r

em

ain

ing (

%)

SA-Free

SA-Loaded25

35

45

55

65

0 20 40 60 80 100 120 140

Onset T

g(°

C)

SA-FreeSA-Loaded

1.5

1.9

2.3

2.7

0 20 40 60 80 100 120 140

Poly

dis

pers

ity index SA-Free

SA-Loaded

-3.3

-2.3

-1.3

-0.3

0 20 40 60 80 100 120 140

ln(M

w(t

)/M

w(t

0))

SA-Free (-0.2 offset)

SA-Loaded

Time (Days)

Time (Days)

Time (Days)

A) Molecular weight

k1

k1’k2

k2’

B) Polydispersity index (Mw/Mn)

C) Glass transition temperature D) Mass remaining

Time (Days)


- 201 -

PLGAs. This was not the case here, since PLGA 85:15 showed overall slow

degradation (> 5 months) and little mass loss (< 20%) so we can negate this as

contributing factor. In parallel, the monitoring of pH confirmed that pH remained at

7.4 throughout the degradation study, thus confirming the protein-polymer

interactions hypothesis.

When encapsulating proteins with methods that involve water/oil interfaces such

as multiple emulsion processes, protein degradation through non-covalent

aggregation is the main interaction taking place [241]. Alternatively, when using

electrospraying, protein-loaded particles can be produced without the use of

emulsions, by dispersing the protein directly into the organic polymer solution.

Indeed, the encapsulation efficiency of the SA-loaded PLGA particles free in

solution, measured through an extraction process involving water/oil interface was

measured as 46.4 ± 3.1%. This was lower than the total detected SA released (63.5 ±

9.9%) showing evident discrepancies (p = 0.02). This difference was attributed to the

non-covalent aggregation of SA at the water/oil interface during the encapsulation

efficiency assay. This phenomenon did not take place during electrospraying or

while the SA was released in solution, consequently a higher amount of SA may be

measured during the release study.

Boury et al. have also observed slower degradation of PLGA 50:50 in the

presence of SA and were able to show that the released SA non-specifically adsorbed

back onto the PLGA surface [242]. This resulted in a hydrophobic coating on the

surface of the polymer, hence reducing water uptake and slowing hydrolytic

degradation. This same phenomenon may be occurring on the electrosprayed

particles, although contact angle measurements were not possible owing to their size,

and we propose this phenomenon to be responsible for incomplete SA release. An

approach to limit non-specific adsorption is to use an anionic surfactant, such as

sodium dodecyl sulfate, in the release medium which can displace adsorbed albumin

from the PLGA. Crotts et al. indeed showed that up to 20% more protein were

released over eight weeks of incubation from PLGA 50:50 microparticles, compared

to when PLGA microparticles were incubated in PBS alone [210].

Overall, the protein release appears to be governed by diffusion, polymer

degradation, and protein-polymer interactions. This study emphasises the in vitro

limitations and further studies are thus required that look at the in vivo situation


- 202 -

where different conditions, such as particle mobility, viscosity and enzymatic

participation, may trigger different behaviours.

6.5.3 Biological Effects

Here we evaluated the effect of PCL and composite scaffolds on MC3T3 precursor

osteoblast cells in vitro. MC3T3s have the potential to differentiate into osteoblasts

under osteogenic stimuli. To examine the influence of the scaffolds on cell viability

and morphology, we seeded MC3T3s onto the scaffolds and allowed them to culture

for 18 days.

The live/dead staining showed high viability and even distribution of cells across

both PCL and composite scaffolds after 1 day of culture (data not shown). After 9

and 18 days, high cell viability throughout the cell culture study showed that none of

the constructs were cytotoxic to MC3T3s (Figure 6.6a-b), and thus incorporation of

PLGA particles did not lead to reduced cell viability, an essential pre-requisite in TE

applications. Higher resolution confocal laser scanning microscopy (CLSM) (Figure

6.6c-d) and SEM images (Figure 6.7) showed a characteristic spreading morphology.

Cells attached to both fibres and particles and spanned adjacent fibres and particles.

The cells on raw PCL scaffolds exhibited a typical elongated morphology, and grew

along fibres. Interestingly, the cells on composite scaffolds showed a different

growth pattern, wrapping around the coated fibres rather than along their axes

(Figure 6.6c-d). This was likely due to the presence of particles which roughened the

topography [212]. The isotropic growth direction of cells was confirmed by SEM

where numerous filopodia of MC3T3 cells reached neighbouring particles (Figure

6.7b), while cells on the fibres alone showed preferential growth along the fibres axis

(Figure 6.7a) and across fibres at fibre crossover points (Figure 6.7d). After initial

contact with a small region on the top of neighbouring microparticles, cells then

spread around particles, covering several particles at a time, while also penetrating

between adjacent particles. Topographical features such as curvature are critical

parameters for cell locomotion. While cells prefer to extend in horizontal planes due

to limited flexibility of the cytoskeleton, they possess the ability to still make

successful protrusion and contact in any given direction [243, 244], and are more

stimulated by micropatterned surfaces than flat surfaces [212]. Importantly, here no

differences were seen between SA-loaded and SA-free composites, showing similar

cell morphology and attachment on both types of scaffold.


- 203 -

It can be concluded that the composite scaffolds loaded with SA are

biocompatible with MC3T3 cells and do no exert any negative biological effect,

allowing initial and subsequent cell attachment, with high cell viability up to 18 days

in vitro. Subsequent studies will look at proliferation and optimum infiltration of

cells within the constructs, by controlling the porosity of the scaffold and the pore

size, respectively tunable here by the duration of the coating, fibre size and particle

size.

Figure 6.6. (A-B) LIVE/DEAD staining and (C-D) CLSM 3D projections of MC3T3 cells cultured on

PCL raw scaffolds and composites after 18 days of cell culture. (A-B) FDAC (green) indicated live

cells, while PI (red) indicated dead cells. (C-D) Rhodamine-conjugated phalloidin stained cell f-actin

(red) while DAPI stained cell nuclei (blue). White arrows show cells spanning on adjacent fibres.

200 µm

Raw PCL scaffolds SA-free composites

Liv

e/D

ea

d s

tain

ing

Nu

cle

i/A

ctin

sta

inin

g

200 µm

25 µm 25 µm

A B

C D


- 204 -

Figure 6.7. SEM visualisation of MC3T3 cells cultured on PCL raw scaffolds and composites after 9

(A-C) and 18 (D-F) days of cell culture. Yellow arrows show cell filopodia reaching neighbouring

particles. Blue arrow shows cell spanning on two adjacent fibres. Green arrow shows microparticles

coalescing. Inset in (F) shows a CLSM 3D projection of cells growing on composites. Red stains f-

actin while blue stains nuclei.

6.6 CONCLUSIONS

In conclusion, the innovative technique of electrospraying on melt electrospun

scaffolds allowed the fabrication of novel biocompatible composites. By using a

novel collector configuration, we fabricated unique constructs comprising of PCL

microfibres coated with high densities of PLGA microparticles, containing SA

amounts comparable with most GF applications and released with reduced burst.

Overall, the technique enabled precise control over physical and protein release

parameters for a final composite scaffold which may be ultimately tailored to fit any

protein delivery purposes, and may be specifically well-suited as a template for GF

delivery therapies applied to skin, cartilage and bone.

10 µm

F

Raw PCL scaffolds SA-free composites SA-loaded composites9

days

18

da

ys

10 µm 10 µm 10 µm

10 µm 10 µm

A B C

D E

10 µm


- 205 -

6.7 SUPPORTING INFORMATION

6.7.1 Electrospraying Setup

Figure S6.8. (A) µCT image of a raw PCL melt electrospun microfibre scaffold. (B-C) Schematics of

the collector setups for electrospraying in real proportions; (B) traditional 15 × 15 cm aluminium foil

collector, (C) 4.8 cm aluminium disc collector. (D-F) Morphology of fibres; (D) from a raw PCL

scaffold, (E) after electrospraying for one hour on a traditional collector and (F) on the optimised

collector. (G) Schematic of the electrospraying setup (non-proportional).

6.7.2 Particle Size Distributions and Morphologies

Optimisation of the electrospraying parameters (0.8 mL/h flow rate and a 10% wt/v

PLGA concentration) resulted in the production of particles with uniform size

distributions and reproducible spherical morphologies by ensuring sufficient polymer

chain entanglements in solution. This is mainly controlled by increased polymer

concentrations and reduced flow rates, when assuming constant solution conductivity

and the stable cone-jet mode of electrospraying, where evaporating droplets are

unable to be disrupted by Coulomb fission [19]. The addition of 1% wt SA had no

100 µm 100 µm

B) Traditional collector setup

0.800

START

+ 10 kV

G) Electrospraying setup

Syringe pump

Voltage supply

Non-conductive PMMA

stand with disc collector

and PCL melt electrospun

microfibre scaffolds

Electrosprayed PLGA

microparticle droplets

Aluminium

disc

48 mm

Non-conductive PMMA stand

PCL melt electrospun


Non-conductive PMMA stand

PCL melt electrospun


14 mm

150 mm

15

0 m

m

Aluminium foil

C) Optimized collector setupA) Raw PCL scaffold

100 µm

D E F


- 206 -

discernible effect on the morphology of particles and was similar to SA-free

particles. Morphology of particles collected either on the collector or coated on the

fibres were identical, also no change was observed when moving to the new collector

configuration, either with or without the addition of SA, confirming the

electrospraying processing parameters and configuration chosen here as highly

homogenous.

Figure S6.9. (A) Comparison of particle size distributions of SA-loaded and SA-free particles within

scaffolds (µm). (B) Comparison of particle size distributions of SA-free particles collected on the

collector and SA-free particles embedded within the scaffolds (µm). (C-D) SEM images of SA-free

particles found; (C) within scaffolds and; (D) collected from the aluminium disc. The wider size

distribution for particles collected on the aluminium foil is due to the presence of smaller satellite

particles at the periphery of the cone formed when particles divide en route to the collector.

Conversely, the scaffolds were placed more centrally on the collector and as such deposition of

satellite particles was minimised.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

1.0 2.3 3.7 5.0 6.3 7.7 9.0

Fra

ctio

n o

f p

op

ula

tio

n

Particle diameter (µm)

BSA-free particles: From the collector

BSA-free particles: Embedded within scaffolds

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

1.0 3.0 5.0 7.0 9.0 11.0 13.0Fra

ctio

n o

f p

op

ula

tio

n

Particle diameter (µm)

Within scaffolds: BSA-loaded particlesWithin scaffolds: BSA-free particles

A B

C D

10 µm 10 µm


- 207 -

6.7.3 Morphology of Composite Scaffolds

Figure S6.10. SEM images of the front and back of composite scaffolds after 1, 2 and 3 h of

electrospraying of SA-free PLGA particles. Results show homogenous deposition of the PLGA

particles throughout the scaffold for both 1 and 2 h of electrospraying (A-B and C-D, respectively)

while a deposition gradient was seen following 3 h of electrospraying (E-F).

50 µm

25 µm

50 µm

50 µm 50 µm

25 µm

Scaffold front Scaffold back

1h

ele

ctr

osp

rayin

g2h

ele

ctr

osp

rayin

g3h

ele

ctr

osp

rayin

g

A B

C D

E F


- 208 -

6.7.4 Glass Transition

Figure S6.11. DSC chromatograms from the first runs of SA-free and SA-loaded PLGA particles,

upon time, showing the effect of incubation time on the glass transition. Glass transition temperature

(Tg) is an important indicator of PLGA degradation; with decreasing molecular weight, the degree of

polymer chain entanglement decreases, enabling an increased mobility of macromolecules, thus

lowering the Tg [38]. Upon immersion in PBS, Tg was found to be higher than non-degraded samples,

namely 54°C and 46°C after 10 days, for SA-free and SA-loaded, respectively, compared to 43°C for

the non-degraded samples. Although the overall Tg decreased with time, values remained mostly

higher than the non-degraded samples upon the whole range of degradation, due to lyophilisation of

samples before DSC scanning. While the Tg behaviour was initially similar for both formulations

(loaded and non-loaded), Tg dropped below 37°C for SA-free particles only after 90 days. By

annealing a polymer close to the glass transition region, as it is the case here with PLGA having a Tg

of 43°C, enthalpy relaxation processes cause Tg values to be somewhat higher [42-44]. The presence

of SA within the polymer structure in SA-loaded particles caused slightly more relaxation, responsible

for higher Tg of SA-loaded samples, while slower degradation prevented any significant Tg drop.

6.7.5 Molecular Weight, Polydispersity, Mass

Table S6.3. Number-average molecular weight (Mn) measured by GPC, expressed in g/mol. The

averages of two samples are presented.

Time

(Days)

Raw PCL

Fibres PLGA particles free in solution

Composite scaffolds (PLGA

particles + PCL fibres)

SA-free SA-loaded SA-free SA-loaded

0 56953 31620 31997 37906 39910

10 57935 32314 31490 38750 37120

30 58504 28321 29872 31925 31353

50 57613 19774 24766 27965 28831

70 56264 11026 16054 13329 19174

90 52145 10488 14019 11919 16140

110 53835 3398 8928 7379 9199

130 50923 2240 5933 3031 6651

Raw

10 days

30 days

70 days

110 days

He

at f

low

(e

xo

the

rmic

up

)

-3.5

-2.5

-1.5

-0.5

Hea

t Flo

w (W

/g)

25 45 65 85 105 125 145 165

Temperature (°C)Exo Up Universal V4.5A TA Instruments

30 40 50 60 70 T(ºC)

SA-freeSA-loaded

37 C


- 209 -

Table S6.4. Weight-average molecular weight (Mw) measured by GPC, expressed in g/mol. The

averages of two samples are presented.

Time

(Days)

Raw PCL





0 80315 56570 55940 61600 63649

10 81652 56121 53979 62587 60905

30 81579 48682 50070 55425 55427

50 81867 37131 44333 51487 53731

70 82755 28119 34082 45434 50450

90 81089 26340 33739 41169 48631

110 80883 8535 20676 38369 39665

130 79950 4271 13628 32394 38075

Table S6.5. Polydispersity index, representative of Mw/Mn, measured by GPC. The averages of two

samples are presented.

Time

(Days)

Raw PCL





0 1.41 1.59 1.59 1.79 1.75

10 1.41 1.64 1.64 1.74 1.71

30 1.39 1.74 1.77 1.72 1.68

50 1.42 1.84 1.86 1.88 1.79

70 1.47 3.41 2.63 2.56 2.15

90 1.57 3.49 3.01 2.69 2.41

110 1.50 5.20 4.33 2.17 2.32

130 1.57 10.69 5.73 1.90 2.30

Table S6.6. Average mass remaining (%) measured by gravimetrical analysis. The averages of three

samples are presented.

Time (Days)

0 10 30 50 70 90 110 130

Raw PCL fibres

100 100 100 100 100 100 100 100

PLGA particles free in

solution

SA-free 100 94.8 93.9 91.2 90.8 90.3 88.4 82.4

SA-loaded 100 98.3 97.1 93.5 92.3 88.2 86.2 82.6



SA-free 100 95.3 96.3 94.1 93.4 92.9 92.3 89.0

SA-loaded 100 97.8 97.3 95.2 93.6 93.8 92.7 90.3


- 210 -


The authors would like to thank Dr. Rachel Hancock (QUT) for biological sample

preparation for SEM analysis. N.B. acknowledges the financial support from QUT in

the form of an Australian Postgraduate Award scholarship, and top-up from the

Deputy Vice Chancellor. B.F. acknowledges support from the Wound Management

Innovation CRC. This work was supported by the Australian Research Council

(Linkage grant LP110200082).

- 211 -

Chapter 7: Summary and Future Directions

Here, the outcomes of this thesis are discussed, with respect to the research problem

and aims stated in Chapter 1. It was hypothesised that the electrospraying

technology may be used to produce biodegradable microparticles encapsulating and

delivering growth factors (GFs) relevant in bone tissue engineering.

The aims addressing this research problem were; to understand and tailor the

processing parameters involved in electrosprayed particle formation, to develop

electrosprayed particle formulations for reproducible and efficient GF encapsulation,

to characterise GF-loaded formulations for in vitro release and bioactivity, and to

investigate the potential of loaded electrosprayed microparticles used in association

with a porous fibre scaffold in vitro, as a suitable construct for tissue engineering.

Development of electrosprayed particles containing growth factors

A review of the literature (Chapter 2) collates and summarises the current body of

research surrounding the use of the electrospraying technique applied to the loading

of therapeutic molecules for delivery from biodegradable polymeric particles. The

chapter suggests that electrospraying is a promising technology for drug delivery

applications, yet is scarcely used and characterised for the encapsulation of proteins

and GFs for applications in tissue engineering.

A key to reproducible electrospraying is the understanding of the complex

interplay between materials and electrospraying processing parameters (Chapter 3

and 4), which direct the size distributions and morphology of particles – two

essential parameters to efficiently encapsulate and deliver GFs. Electrospraying

requires a strict control over processing parameters which are inter-dependently

linked (Chapter 3). The polymer entanglement regime taking place in the course of

the electrospraying process is essential to reproducible, spherical and narrowly

distributed particles, otherwise leading to shapeless particles, fibres, and

offspring/secondary droplets, responsible for irreproducible morphologies and

bimodal size distributions. The two strongest parameters to control the entanglement

regime are electrospraying flow rate and polymer concentration whereas less

significant variables include tip-to-collector distance, voltage and needle gauge. For

Chapter 7 Summary and Future Directions

- 212 -

a standard FDA-approved biodegradable polymer, such as polycaprolactone (PCL)

(84 kDa) dissolved in chloroform, the optimum conditions involve a 9-10% wt/v

concentration and flow rates above 0.5 mL/h, generating average particle sizes

between 10 and 20 µm.

While a single polymer/solvent combination is more easily optimisable, the

blending with an additional polymer can create instabilities during the

electrospraying process, influencing the polymer entanglements and final

characteristics of particles (Chapter 4). When PCL and poly(lactic-co-glycolic acid)

(PLGA) are mixed with 10% wt of poly(ethylene glycol) (PEG), a spread of particle

sizes and morphologies are obtained. Here, flow rate and concentration, along with

molecular weight of the additive are important variables to control in order to obtain

spherical and monodisperse particles. However, not all combinations are able to

provide sufficiently large particle sizes, which are required for homogeneous, dry

encapsulation of large GF clusters. PEG is an efficient protective agent for proteins

and GFs (Chapter 5) and good microniser (Chapter 4) prior to dry encapsulation in

electrosprayed particles, yet encapsulation of a model protein, serum albumin (SA),

is improved for increasing particle sizes. In general, higher molecular weight PEG,

close to the molecular weight of the matrix polymer and higher electrospraying flow

rates (here up to 3 mL/h) should be considered, for increased particle sizes. An

increase in flow rate, however, correlates with bimodal size distributions being

generated, comprising 50% of primary and 50% secondary droplets (which are

approximately half the size of the primary droplets).

Compared to the amount and molecular weight of PEG used here, particle size

distributions are more determinant in influencing the release profiles of SA, with

broader distributions generating more particle aggregation in solution. Aggregation

reduces burst release, providing more sustained but less overall release of the protein

during the diffusion stage of release. Hence, to avoid bimodal size distributions, flow

rates need to be reduced, in turn also lowering the final average particle size. When

applied to an FDA-approved polymer such as PLGA 85:15, which presents a

degradation profile more suited to GF delivery applications than PCL due to a faster

degradation, optimised parameters for GF encapsulation are obtained using a

polymer concentration of 11% wt/v and a flow rate of 0.8 mL/h (Chapter 5). This

generates quasi-monodisperse particles of around 6 µm in size, which allows


- 213 -

reproducible and efficient encapsulation of dry bone morphogenetic protein 7 (BMP-

7) and vascular endothelial GF (VEGF), following micronisation with PEG.

Characterisation of growth factors in vitro

The in vitro characterisation of a drug release system is helpful in understanding

performance in vivo, and is a pre-requisite to any expensive and complex experiment.

Here, a challenge lies in detecting protein drugs encapsulated and released from

polymeric electrosprayed particles in vitro, due to significant interactions with the

environment (Chapter 4, 5, 6). Due to the presence of hydrophobic domains in

proteins and growth factors, they tend to aggregate together and non-specifically

adsorb to the polymer matrix, which is hydrophobic in nature. This can happen at

several stages of processing, from micronisation up until delivery, but also during

post-encapsulation quantification.

An important finding here is that in the course of GF processing, especially

involving drying after contact with organic and aqueous solvents, a significant

degree of aggregation between GFs takes place. GF dissociation is not fully achieved

by re-dissolution in an aqueous solvent, as evidenced by lower GF concentrations in

solution measured by the enzyme-linked immunosorbent assay (ELISA) assay.

Strikingly, these processed GFs present a full bioactivity when placed in contact with

cells (Chapter 5). This result indicates that the nature of aggregation obtained

through processing relative to electrospraying, is non-covalent and does not affect

protein stability.

During characterisation of GFs in vitro, non-covalent aggregation is accompanied

by a degree of non-specific adsorption to the hydrophobic polymer matrix, which the

electrosprayed particles are made of. This is shown to lower GF detection but also

affect particle degradation during in vitro release, which in turn may alter the release

profiles (Chapter 6). This generates complications in characterisation; a strategy to

overcome this involves the use of surfactants. Here, Tween 20® and sodium dodecyl

sulphate (SDS) are efficient in displacing adsorbed growth factors from hydrophobic

matrices to some extent. However, they are unable to fully dissociate aggregated

proteins, as seen with SA (Chapter 4), and growth factors, such as VEGF and BMP-

7 (Chapter 5).


- 214 -

Assessing the encapsulation efficiency by extraction procedures, which involve

the dissolution of loaded electrosprayed particles in an organic solvent, is mostly

affected by GF aggregation, generating results lower than 50%, sometimes lower

than the total amounts detected during release assays, as measured by ELISA. The in

vitro release profiles of SA, VEGF and BMP-7 are highly dependent on the in vitro

conditions rather than particle formulations, involving the amount of particles studied

and the presence of surfactant in solution. Non-specific adsorption to polymer is

dominant during release and is better counterbalanced by the use of SDS, an anionic

surfactant, than Tween 20®, a non-ionic surfactant (Chapter 5). All profiles are

affected by incomplete release due to GF interactions with their environment, which

the presence of surfactant in solution but also the presence of PEG, SA, and trehalose

in the particle formulation do not fully impair. Trehalose, in particular (1% wt of the

particle formulation), leads to reduced burst release of BMP-7 from electrosprayed

particles, but also reduces overall release due to promotion of BMP-7 aggregates

during the micronisation step which further reduces detection in solution, as

measured by ELISA. Conversely, similar bioactivity is measured in vitro on pre-

osteoblasts when either the formulation with or without the trehalose is used,

showing the discrepancies of release assays in PBS and measured by ELISA, and the

actual effectiveness seen with direct contact with cells.

Here it becomes clear that in vitro assays may be limited in truly assessing

electrosprayed particles loaded and releasing GFs. An in vivo experiment may be

more suitable, whereby the environment offers further possible interactions and is the

only environment whereby the GF action can be truly measured via assessing the

new tissue formation/regeneration. This could be, for instance, assessed by an

ectopic bone model in mice.

Potential of electrospraying for tissue engineering

In order to validate electrosprayed particles loaded with GFs to be used for

applications in tissue engineering, particles must be non-toxic and minimise potential

GF denaturation during processing. The electrospraying process indeed requires

organic solvents that come in contact with GFs, which may be a potential source of

GF denaturation and toxicity for tissue engineering applications. Here, GF-free PCL

electrosprayed microparticles have no adverse effects on fibroblast cells up to 48 h


- 215 -

after direct exposure (Chapter 3), inferring that the organic solvent is fully removed

following the electrospraying process.

Next, when BMP-7 and VEGF are freeze-dried, micronised with PEG, and

vortexed with the organic solvent, which are the most critical steps during the

electrospraying process, it is shown that both GFs perform in an unaltered manner in

vitro in contact with cells (Chapter 5). When microparticles loaded with BMP-7 are

placed in vitro in direct contact with pre-osteoblasts, significant cell differentiation is

observed up to three weeks due to active BMP-7 being released from microparticles,

hence validating the technique for the specific bone tissue engineering application.

The electrospraying process is an atomization process, the final particles produced

are dry and do not need extra treatment after production. Due to the nature of the

fabrication method, particles are also delivered with a natural homogenous

distribution on a collector. This is a great advantage for applications in tissue

engineering, where scaffolds and other matrices can homogeneously be coated with

particles, when placed under the electrospraying cone (Chapter 6). Here, PCL melt

electrospun meshes provide a suitable substrate with high porosity and pore size

allowing scaffold coating throughout the structure. Interestingly, due to the

immobilisation of PLGA particles on the surface of the meshes, burst release of SA

is reduced, compared to particles free in solution. As seen with particle size

distributions (Chapter 4), particle aggregation is, again, paramount in tailoring the

release profiles of GF from electrosprayed particles. In terms of suitability of the

electrosprayed/electrospun composites developed here, they have a positive effect in

contact with precursor osteoblast cells up to 18 days in culture, which is an

encouraging result for further consideration as a construct for bone tissue

engineering.

Future Directions

In this thesis, the potential of the electrospraying technology for producing

biodegradable microparticles encapsulating and delivering growth factors relevant in

bone tissue engineering, namely VEGF and BMP-7, is discussed. While positive

results lean towards this goal with the use of simple polyester polymers such as

PLGA and PCL, several areas still need to be addressed. Particle size distributions

and particle configuration, involving aggregation in solution or immobilisation on a


- 216 -

scaffold, seem to be key in directing the release of GFs from electrosprayed particles

and may be an important variable to contend with for the tissue engineering

application. Importantly, these parameters seem more dominant over the actual

particle formulation, with the additives and amounts studied here (10% wt PEG and

1% wt trehalose). GF interactions are a significant issue in the in vitro

characterisation of electrosprayed particles loaded with VEGF and BMP-7, and may

be similar for any type of GF. The optimisation of the in vitro environment is

essential to achieve accurate quantification but may actually be difficult to achieve,

which here becomes a strong argument in favour of in vivo assessment instead,

although this remains controversial with increased costs.

While burst release is often regarded negatively in the field of drug delivery in

general, it is not the specific case of GF delivery, where burst delivery is as

beneficial and necessary as sustained delivery. Indeed, a complex cascade of

numerous growth factors need to occur to allow full tissue reconstruction and the

burst release of specific growth factors during this process is also essential, for

instance in triggering the recruitment of natural growth factors. Hence, in the field of

GF delivery, burst release should be regarded as complementary to sustained

delivery and rather than avoiding it, a thorough control and tailoring of all types of

profiles (burst and sustained) would be more useful.

Finally, the use of electrospraying with GFs on fibre scaffolds is established as a

successful solution for turning inert fibres into tissue-inductive materials, but will

require more characterisation to be fully translated to the bone tissue engineering

application. More specifically, the possibility of simultaneously electrospraying and

electrospinning the scaffold material should provide an ideal construct with more

tailoring possibilities for wider applications in the biomedical field.

- 217 -

Bibliography

[1] Williams, D. F., The Williams Dictionary of Biomaterials. Liverpool

University Press: 1999.

[2] MacNeil, S., Progress and opportunities for tissue-engineered skin. Nature

2007, 445, 874-880.

[3] Langer, R.; Vacanti, J. P., Tissue Engineering. Science 1993, 260, 920-926.

[4] Hutmacher, D. W.; Garcia, A. J., Scaffold-based bone engineering by using

genetically modified cells. Gene 2005, 347, 1-10.

[5] Whitaker, M. J.; Quirk, R. A.; Howdle, S. M.; Shakesheff, K. M., Growth

factor release from tissue engineering scaffolds. J. Pharm. Pharmacol. 2001,

53, 1427-1437.

[6] Babensee, J. E.; McIntire, L. V.; Mikos, A. G., Growth factor delivery for

tissue engineering. Pharm. Res. 2000, 17, 497-504.

[7] Chen, R. R.; Mooney, D. J., Polymeric growth factor delivery strategies for

tissue engineering. Pharm. Res. 2003, 20, 1103-1112.

[8] Nagai, M. K.; Embil, J. M., Becaplermin: recombinant platelet derived

growth factor, a new treatment for healing diabetic foot ulcers. Exp. Op.

Biolog. Ther. 2002, 2, 211-218.

[9] Cahill, K. S.; Chi, J. H.; Day, A.; Claus, E. B., Prevalence, Complications,

and Hospital Charges Associated With Use of Bone-Morphogenetic Proteins

in Spinal Fusion Procedures. JAMA 2009, 302, 58-66.

[10] Lysaght, M. J.; Jaklenec, A.; Deweerd, E., Great expectations: Private sector

activity in tissue engineering, regenerative medicine, and stem cell

therapeutics. Tissue Eng. A 2008, 14, 305-357.

[11] Place, E. S.; Evans, N. D.; Stevens, M. M., Complexity in biomaterials for

tissue engineering. Nat. Mater. 2009, 8, 457-470.

[12] Bessa, P. C.; Casal, M.; Reis, R. L., Bone morphogenetic proteins in tissue

engineering: the road from the laboratory to the clinic, part I (basic

concepts). J. Tissue Eng. Regen. M. 2008, 2, 1-13.

[13] Kang, J. D., Commentary: Another complication associated with rhBMP-2?

Spine J. 2011, 11, 517-519.

[14] Carragee, E. J.; Hurwitz, E. L.; Weiner, B. K., A critical review of

recombinant human bone morphogenetic protein-2 trials in spinal surgery:

emerging safety concerns and lessons learned. Spine J. 2011, 11, 471-491.

[15] Mannion, R. J.; Nowitzke, A. M.; Wood, M. J., Promoting fusion in

minimally invasive lumbar interbody stabilization with low-dose bone

morphogenic protein-2-but what is the cost? Spine J. 2011, 11, 527-533.

[16] Vasita, R.; Katti, D. S., Growth factor-delivery systems for tissue

engineering: a materials perspective. Expert Rev. Med. Dev. 2006, 3, 29-47.

[17] Sokolsky-Papkov, M.; Agashi, K.; Olaye, A.; Shakesheff, K.; Domb, A. J.,

Polymer carriers for drug delivery in tissue engineering. Adv. Drug Deliver.

Rev. 2007, 59, 187-206.

[18] Freiberg, S.; Zhu, X., Polymer microspheres for controlled drug release. Int.

J. Pharm. 2004, 282, 1-18.

- 218 -

[19] Smith, L. A.; Liu, X. H.; Ma, P. X., Tissue engineering with nano-fibrous

scaffolds. Soft Matter 2008, 4, 2144-2149.

[20] Szentivanyi, A.; Chakradeo, T.; Zernetsch, H.; Glasmacher, B., Electrospun

cellular microenvironments: Understanding controlled release and scaffold

structure. Adv. Drug Deliver. Rev. 2011, 63, 209-220.

[21] Venugopal, J.; Low, S.; Choon, A. T.; Ramakrishna, S., Interaction of cells

and nanofiber scaffolds in tissue engineering. J. Biomed. Mater. Res. B 2008,

84B, 34-48.

[22] Chung, H. J.; Park, T. G., Surface engineered and drug releasing pre-

fabricated scaffolds for tissue engineering. Adv. Drug Deliver. Rev. 2007,

59, 249-262.

[23] Hadjiargyrou, M.; Chiu, J. B., Enhanced composite electrospun nanofiber

scaffolds for use in drug delivery. Exp. Op. Drug Deliv. 2008, 5, 1093-1106.

[24] Tessmar, J. K.; Goepferich, A. M., Matrices and scaffolds for protein

delivery in tissue engineering. Adv. Drug Deliver. Rev. 2007, 59, 274-291.

[25] Chew, S. Y.; Wen, J.; Yim, E. K. F.; Leong, K. W., Sustained Release of

Proteins from Electrospun Biodegradable Fibers. Biomacromolecules 2005,

6, 2017-2024.

[26] Freitas, S.; Merkle, H. P.; Gander, B., Microencapsulation by solvent

extraction/evaporation: reviewing the state of the art of microsphere

preparation process technology. J. Control. Release 2005, 102, 313-332.

[27] Jaworek, A.; Sobczyk, A. T., Electrospraying route to nanotechnology: An

overview. J. Electrostat. 2008, 66, 197-219.

[28] Jaworek, A., Electrostatic micro- and nanoencapsulation and

electroemulsification: A brief review. J. Microencapsul. 2008, 25, 443-468.

[29] Wang, F.; Li, Z. Q.; Tamama, K.; Sen, C. K.; Guan, J. J., Fabrication and

Characterization of Prosurvival Growth Factor Releasing, Anisotropic

Scaffolds for Enhanced Mesenchymal Stem Cell Survival/Growth and

Orientation. Biomacromolecules 2009, 10, 2609-2618.

[30] Ekaputra, A. K.; Prestwich, G. D.; Cool, S. M.; Hutmacher, D. W., The three-

dimensional vascularization of growth factor-releasing hybrid scaffold of

poly (ɛ-caprolactone)/collagen fibers and hyaluronic acid hydrogel.

Biomaterials 2011, 32, 8108-8117.

[31] Sukarto, A.; Amsden, B. G., Low melting point amphiphilic microspheres for

delivery of bone morphogenetic protein-6 and transforming growth factor-β3

in a hydrogel matrix. J. Control. Release 2012, 158, 53-62.

[32] Sukarto, A.; Yu, C.; Flynn, L. E.; Amsden, B. G., Co-delivery of Adipose-

Derived Stem Cells and Growth Factor-Loaded Microspheres in RGD-

Grafted N-Methacrylate Glycol Chitosan Gels for Focal Chondral Repair.

Biomacromolecules 2012, 13, 2490-2502.

[33] Soran, Z.; Aydin, R. S. T.; Gumusderelioglu, M., Chitosan scaffolds with

BMP-6 loaded alginate microspheres for periodontal tissue engineering. J.

Microencapsul. 2012, 29, 770-780.

[34] Putney, S. D.; Burke, P. A., Improving protein therapeutics with sustained-

release formulations. Nat. Biotechnol. 1998, 16, 153-157.

[35] Ye, M.; Kim, S.; Park, K., Issues in long-term protein delivery using

biodegradable microparticles. J. Control. Release 2010, 146, 241-260.

[36] Chen, F.-M.; Zhang, M.; Wu, Z.-F., Toward delivery of multiple growth

factors in tissue engineering. Biomaterials 2010, 31, 6279-6308.

- 219 -

[37] Li, H. F.; Zhang, H. M.; Guo, X. L., Preparation of Micro-/Nano-

Nonspherical Polymer Particles. Prog. Chem. 2011, 23, 1196-1210.

[38] Chan, H.-K.; Kwok, P. C. L., Production methods for nanodrug particles

using the bottom-up approach. Adv. Drug Deliver. Rev. 2011, 63, 406-416.

[39] Peltonen, L.; Valo, H.; Kolakovic, R.; Laaksonen, T.; Hirvonen, J.,

Electrospraying, spray drying and related techniques for production and

formulation of drug nanoparticles. Exp. Op. Drug Deliv. 2010, 7, 705-719.

[40] Lassalle, V.; Ferreira, M. L., PLA nano- and microparticles for drug delivery:

An overview of the methods of preparation. Macromol. Biosci. 2007, 7, 767-

783.

[41] Ciach, T., Application of electro-hydro-dynamic atomization in drug delivery.

J. Drug. Deliv. Sci. Tech. 2007, 17, 367-375.

[42] Hartman, R. P. A.; Brunner, D. J.; Camelot, D. M. A.; Marijnissen, J. C. M.;

Scarlett, B., Jet break-up in electrohydrodynamic atomization in the cone-jet

mode. J. Aerosol Sci. 2000, 31, 65-95.

[43] Brandenberger, H.; Nussli, D.; Piech, V.; Widmer, F., Monodisperse particle

production: A method to prevent drop coalescence using electrostatic forces.

J. Electrostat. 1999, 45, 227-238.

[44] Gomez, A.; Bingham, D.; Juan, L. d.; Tang, K., Production of protein

nanoparticles by electrospray drying. J. Aerosol Sci. 1998, 29, 561-574.

[45] Yue, H.; Wei, W.; Yue, Z.; Lv, P.; Wang, L.; Ma, G.; Su, Z., Particle size

affects the cellular response in macrophages. Eur. J. Pharm. Sci. 2010, 41,

650-657.

[46] Xie, Y. B.; Castracane, J., High-Voltage, Electric Field-Driven

Micro/Nanofabrication for Polymeric Drug Delivery Systems Electrostatic

Techniques and Biomedical Application. IEEE Eng. Med. Biol. Mag. 2009,

28, 23-30.

[47] Amsden, B. G.; Goosen, M. F. A., An examination of factors affecting the

size, distribution and release characteristics of polymer microbeads made

using electrostatics. J. Control. Release 1997, 43, 183-196.

[48] Chakraborty, S.; Liao, I. C.; Adler, A.; Leong, K. W., Electrohydrodynamics:

A facile technique to fabricate drug delivery systems. Adv. Drug Deliver.

Rev. 2009, 61, 1043-1054.

[49] Almeria, B.; Deng, W. W.; Fahmy, T. M.; Gomez, A., Controlling the

morphology of electrospray-generated PLGA microparticles for drug

delivery. J. Colloid Interf. Sci. 2010, 343, 125-133.

[50] Almería, B.; Fahmy, T. M.; Gomez, A., A multiplexed electrospray process

for single-step synthesis of stabilized polymer particles for drug delivery. J.

Control. Release 2011, 154, 203-210.

[51] Arya, N.; Chakraborty, S.; Dube, N.; Katti, D. S., Electrospraying: A Facile

Technique for Synthesis of Chitosan-Based Micro/Nanospheres for Drug

Delivery Applications. J. Biomed. Mater. Res. B 2009, 88B, 17-31.

[52] Hong, Y. L.; Li, Y. Y.; Yin, Y. Z.; Li, D. M.; Zou, G. T.,

Electrohydrodynamic atomization of quasi-monodisperse drug-loaded

spherical/wrinkled microparticles. J. Aerosol Sci. 2008, 39, 525-536.

[53] Ranganath, S. H.; Kee, I.; Krantz, W. B.; Chow, P. K. H.; Wang, C. H.,

Hydrogel Matrix Entrapping PLGA-Paclitaxel Microspheres: Drug Delivery

with Near Zero-Order Release and Implantability Advantages for Malignant

Brain Tumour Chemotherapy. Pharm. Res. 2009, 26, 2101-2114.

- 220 -

[54] Naraharisetti, P. K.; Ong, B. Y. S.; Xie, J. W.; Lee, T. K. Y.; Wang, C. H.;

Sahinidis, N. V., In vivo performance of implantable biodegradable

preparations delivering Paclitaxel and Etanidazole for the treatment of

glioma. Biomaterials 2007, 28, 886-894.

[55] Xie, J.; Marijnissen, J. C. M.; Wang, C.-H., Microparticles developed by

electrohydrodynamic atomization for the local delivery of anticancer drug to

treat C6 glioma in vitro. Biomaterials 2006, 27, 3321-3332.

[56] Xie, J. W.; Lim, L. K.; Phua, Y. Y.; Hua, J. S.; Wang, C. H.,

Electrohydrodynamic atomization for biodegradable polymeric particle

production. J. Colloid Interf. Sci. 2006, 302, 103-112.

[57] Xie, J. W.; Tan, J. C.; Wang, C. H., Biodegradable films developed by

electrospray deposition for sustained drug delivery. J. Pharm. Sci. 2008, 97,

3109-3122.

[58] Nie, H.; Dong, Z.; Arifin, D. Y.; Hu, Y.; Wang, C.-H., Core/shell

microspheres via coaxial electrohydrodynamic atomization for sequential

and parallel release of drugs. J. Biomed. Mater. Res. A 2010, 95A, 709-716.

[59] Ding, L.; Lee, T.; Wang, C. H., Fabrication of monodispersed Taxol-loaded

particles using electrohydrodynamic atomization. J. Control. Release 2005,

102, 395-413.

[60] Ciach, T., Microencapsulation of drugs by electro-hydro-dynamic

atomization. Int. J. Pharm. 2006, 324, 51-55.

[61] Wu, Y. Q.; MacKay, J. A.; McDaniel, J. R.; Chilkoti, A.; Clark, R. L.,

Fabrication of Elastin-Like polypeptide Nanoparticles for Drug Delivery by

Electrospraying. Biomacromolecules 2009, 10, 19-24.

[62] Xie, J. W.; Tan, R. S.; Wang, C. H., Biodegradable microparticles and fiber

fabrics for sustained delivery of cisplatin to treat C6 glioma in vitro. J.

Biomed. Mater. Res. A 2008, 85A, 897-908.

[63] Ijsebaert, J. C.; Geerse, K. B.; Marijnissen, J. C. M.; Lammers, J. W. J.;

Zanen, P., Electro-hydrodynamic atomization of drug solutions for inhalation

purposes. J. Appl. Physiol. 2001, 91, 2735-2741.

[64] Valo, H.; Peltonen, L.; Vehvilainen, S.; Karjalainen, M.; Kostiainen, R.;

Laaksonen, T.; Hirvonen, J., Electrospray Encapsulation of Hydrophilic and

Hydrophobic Drugs in Poly(L-lactic acid) Nanoparticles. Small 2009, 5,

1791-1798.

[65] Bohr, A.; Kristensen, J.; Stride, E.; Dyas, M.; Edirisinghe, M., Preparation of

microspheres containing low solubility drug compound by

electrohydrodynamic spraying. Int. J. Pharm. 2011, 412, 59-67.

[66] Midhun, B. T.; Shalumon, K. T.; Manzoor, K.; Jayakumar, R.; Nair, S. V.;

Deepthy, M., Preparation of Budesonide-Loaded Polycaprolactone

Nanobeads by Electrospraying for Controlled Drug Release. J. Biomater. Sci.

Polym. Ed. 2011, 22, 2431-2444.

[67] Yu, D.-G.; Williams, G. R.; Yang, J.-H.; Wang, X.; Yang, J.-M.; Li, X.-Y.,

Solid lipid nanoparticles self-assembled from electrosprayed polymer-based

microparticles. J. Mater. Chem. 2011, 21, 15957-15961.

[68] Enayati, M.; Ahmad, Z.; Stride, E.; Edirisinghe, M., Size mapping of electric

field-assisted production of polycaprolactone particles. J. R. Soc. Interface

2010, 7, S393-S402.

[69] Trotta, M.; Cavalli, R.; Trotta, C.; Bussano, R.; Costa, L., Electrospray

technique for solid lipid-based particle production. Drug Dev. Ind. Pharm.

2010, 36, 431-438.

- 221 -

[70] Cavalli, R.; Bisazza, A.; Bussano, R.; Trotta, M.; Civra, A.; Lembo, D.;

Ranucci, E.; Ferruti, P., Poly(amidoamine)-Cholesterol Conjugate

Nanoparticles Obtained by Electrospraying as Novel Tamoxifen Delivery

System. J. Drug Deliv. 2011, 2011, 587604.

[71] Xie, J. W.; Wang, C. H., Encapsulation of proteins in biodegradable

polymeric microparticles using electrospray in the Taylor Cone-Jet mode.

Biotechnol. Bioeng. 2007, 97, 1278-1290.

[72] Xu, Y. X.; Hanna, M. A., Electrosprayed bovine serum albumin-loaded

tripolyphosphate cross-linked chitosan capsules: Synthesis and

characterization. J. Microencapsul. 2007, 24, 143-151.

[73] Xu, Y. X.; Hanna, M. A., Electrospray encapsulation of water-soluble

protein with polylactide - Effects of formulations on morphology,

encapsulation efficiency and release profile of particles. Int. J. Pharm. 2006,

320, 30-36.

[74] Xu, Y. X.; Skotak, M.; Hanna, M., Electrospray encapsulation of water-

soluble protein with polylactide. I. Effects of formulations and process on

morphology and particle size. J. Microencapsul. 2006, 23, 69-78.

[75] Rayleigh, L., XX. On the equilibrium of liquid conducting masses charged

with electricity. Philo. Mag. S5 1882, 14, 184-186.

[76] Rezvanpour, A.; Attia, A. B. E.; Wang, C. H., Enhancement of Particle

Collection Efficiency in Electrohydrodynamic Atomization Process for

Pharmaceutical Particle Fabrication. Ind. Eng. Chem. Res. 2010, 49, 12620-

12631.

[77] Xie, J. W.; Ng, W. J.; Lee, L. Y.; Wang, C. H., Encapsulation of protein

drugs in biodegradable microparticles by co-axial electrospray. J. Colloid

Interf. Sci. 2008, 317, 469-476.

[78] Pareta, R.; Edirisinghe, M. J., A novel method for the preparation of

biodegradable microspheres for protein drug delivery. J. R. Soc. Interface

2006, 3, 573-582.

[79] Ahmad, Z.; Zhang, H. B.; Farook, U.; Edirisinghe, M.; Stride, E.; Colombo,

P., Generation of multilayered structures for biomedical applications using a

novel tri-needle coaxial device and electrohydrodynamic flow. J. R. Soc.

Interface 2008, 5, 1255-1261.

[80] Hwang, Y. K.; Jeong, U.; Cho, E. C., Production of uniform-sized polymer

core-shell microcapsules by coaxial electrospraying. Langmuir 2008, 24,

2446-2451.

[81] Park, S.; Hwang, S.; Lee, J., pH-responsive hydrogels from moldable

composite microparticles prepared by coaxial electro-spray drying. Chem.

Eng. J. 2011, 169, 348-357.

[82] Kim, W.; Kim, S. S., Multishell Encapsulation Using a Triple Coaxial

Electrospray System. Anal. Chem. 2010, 82, 4644-4647.

[83] Bock, N.; Woodruff, M. A.; Hutmacher, D. W.; Dargaville, T. R.,

Electrospraying, a Reproducible Method for Production of Polymeric

Microspheres for Biomedical Applications. Polymers 2010, 3, 131-149.

[84] Cohen, S.; Yoshioka, T.; Lucarelli, M.; Hwang, L. H.; Langer, R., Controlled

Delivery Systems for Proteins Based on Poly(Lactic Glycolic Acid)

Microspheres. Pharm. Res. 1991, 8, 713-720.

[85] Zhu, G. Z.; Mallery, S. R.; Schwendeman, S. P., Stabilization of proteins

encapsulated in injectable poly (lactide-co-glycolide). Nat. Biotechnol. 2000,

18, 52-57.

- 222 -

[86] Li, L.; Schwendeman, S. P., Mapping neutral microclimate pH in PLGA

microspheres. J. Control. Release 2005, 101, 163-173.

[87] Nie, H. M.; Fu, Y. L.; Wang, C. H., Paclitaxel and suramin-loaded core/shell

microspheres in the treatment of brain tumors. Biomaterials 2010, 31, 8732-

8740.

[88] Woodruff, M. A.; Hutmacher, D. W., The return of a forgotten polymer--

Polycaprolactone in the 21st century. Prog. Polym. Sci. 2010, 35, 1217-1256.

[89] Cipitria, A.; Skelton, A.; Dargaville, T. R.; Dalton, P. D.; Hutmacher, D. W.,

Design, fabrication and characterization of PCL electrospun scaffolds-a

review. J. Mater. Chem. 2011, 21, 9419-9453.

[90] Sinha, V. R.; Bansal, K.; Kaushik, R.; Kumria, R.; Trehan, A., Poly-epsilon-

caprolactone microspheres and nanospheres: an overview. Int. J. Pharm.

2004, 278, 1-23.

[91] Seyednejad, H.; Ghassemi, A. H.; van Nostrum, C. F.; Vermonden, T.;

Hennink, W. E., Functional aliphatic polyesters for biomedical and

pharmaceutical applications. J. Control. Release 2011, 152, 168-176.

[92] Wu, Y.; Liao, I. C.; Kennedy, S. J.; Du, J.; Wang, J.; Leong, K. W.; Clark, R.

L., Electrosprayed core-shell microspheres for protein delivery. Chem.

Commun. 2010, 46, 4743-4745.

[93] Anumolu, R.; Gustafson, J. A.; Magda, J. J.; Cappello, J.; Ghandehari, H.;

Pease, L. F., Fabrication of Highly Uniform Nanoparticles from Recombinant

Silk-Elastin-like Protein Polymers for Therapeutic Agent Delivery. ACS

Nano 2011, 5, 5374-5382.

[94] Wu, Y. Q.; Clark, R. L., Electrohydrodynamic atomization: a versatile

process for preparing materials for biomedical applications. J. Biomater. Sci.

Polym. Ed. 2008, 19, 573-601.

[95] Songsurang, K.; Praphairaksit, N.; Siraleartmukul, K.; Muangsin, N.,

Electrospray Fabrication of Doxorubicin-Chitosan-Tripolyphosphate

Nanoparticles for Delivery of Doxorubicin. Arch. Pharm. Res. 2011, 34, 583-

592.

[96] Kim, S. Y.; Lee, H.; Cho, S.; Park, J. W.; Park, J.; Hwane, J., Size Control of

Chitosan Capsules Containing Insulin for Oral Drug Delivery via a

Combined Process of Ionic Gelation with Electrohydrodynamic Atomization.

Ind. Eng. Chem. Res. 2011, 50, 13762-13770.

[97] Park, C. H.; Lee, J., Electrosprayed Polymer Particles: Effect of the Solvent

Properties. J Appl Polym Sci 2009, 114, 430-437.

[98] Wu, Y. Q.; Kennedy, S. J.; Clark, R. L., Polymeric Particle Formation

Through Electrospraying at Low Atmospheric Pressure. J. Biomed. Mater.

Res. B 2009, 90B, 381-387.

[99] Chang, M. W.; Stride, E.; Edirisinghe, M., Controlling the thickness of

hollow polymeric microspheres prepared by electrohydrodynamic

atomization. J. R. Soc. Interface 2010, 7, S451-S460.

[100] Meng, F. Z.; Jiang, Y.; Sun, Z. H.; Yin, Y. Z.; Li, Y. Y.,

Electrohydrodynamic Liquid Atomization of Biodegradable Polymer

Microparticles: Effect of Electrohydrodynamic Liquid Atomization Variables

on Microparticles. J. Appl. Polym. Sci. 2009, 113, 526-534.

[101] Xue, L. W.; Mao, L. X.; Cai, Q.; Yang, X. P.; Jin, R. G., Preparation of

amino acid ester substituted polyphosphazene microparticles via

electrohydrodynamic atomization. Polym. Adv. Technol. 2011, 22, 2009-

2016.

- 223 -

[102] Shenoy, S. L.; Bates, W. D.; Frisch, H. L.; Wnek, G. E., Role of chain

entanglements on fiber formation during electrospinning of polymer

solutions: good solvent, non-specific polymer-polymer interaction limit.

Polymer 2005, 46, 3372-3384.

[103] Gupta, P.; Elkins, C.; Long, T. E.; Wilkes, G. L., Electrospinning of linear

homopolymers of poly(methyl methacrylate): exploring relationships between

fiber formation, viscosity, molecular weight and concentration in a good

solvent. Polymer 2005, 46, 4799-4810.

[104] Pareta, R.; Brindley, A.; Edirisinghe, M. J.; Jayasinghe, S. N.; Luklinska, Z.

B., Electrohydrodynamic atomization of protein (bovine serum albumin). J.

Mater. Sci.-Mater. Med. 2005, 16, 919-925.

[105] GananCalvo, A. M.; Davila, J.; Barrero, A., Current and droplet size in the

electrospraying of liquids. Scaling laws. J. Aerosol Sci. 1997, 28, 249-275.

[106] Ramakrishna, S., An Introduction to Electrospinning and Nanofibers. World

Scientific: 2005.

[107] Zhang, S. L.; Kawakami, K., One-step preparation of chitosan solid

nanoparticles by electrospray deposition. Int. J. Pharm. 2010, 397, 211-217.

[108] Yao, J.; Lim, L. K.; Xie, J. W.; Hua, J. S.; Wang, C. H., Characterization of

electrospraying process for polymeric particle fabrication. J. Aerosol Sci.

2008, 39, 987-1002.

[109] Iserson, K. V., The origins of the gauge system for medical equipment. JEM

1987, 5, 45-48.

[110] Hao, X. F.; Lu, X. F.; Li, Z. Y.; Zhao, Y. Y.; Shang, T. C.; Yang, Q. B.;

Wang, C.; Li, L. J., Effects of the electrospray ionization parameters on the

formation and morphology of colloidal microspheres of polyacrylonitrile. J.

Appl. Polym. Sci. 2006, 102, 2889-2893.

[111] Enayati, M.; Farook, U.; Edirisinghe, M.; Stride, E., Electrohydrodynamic

preparation of polymeric drug-carrier particles: Mapping of the process. Int.

J. Pharm. 2011, 404, 110-115.

[112] Delamora, J. F.; Loscertales, I. G., The current emitted by highly conducting

Taylor cones. J. Fluid Mech. 1994, 260, 155-184.

[113] Kinam, P., Albumin: A versatile carrier for drug delivery. J. Control. Release

2012, 157, 3.

[114] Yoo, J. Y.; Kim, M.; Lee, J., Electrospraying of micro/nano particles for

protein drug delivery. Polymer-Korea 2007, 31, 215-220.

[115] Jaklenec, A.; Hinckfuss, A.; Bilgen, B.; Ciombor, D. M.; Aaron, R.;

Mathiowitz, E., Sequential release of bioactive IGF-I and TGF-beta(1) from

PLGA microsphere-based scaffolds. Biomaterials 2008, 29, 1518-1525.

[116] Bertram, J. P.; Rauch, M. F.; Chang, K.; Lavik, E. B., Using Polymer

Chemistry to Modulate the Delivery of Neurotrophic Factors from

Degradable Microspheres: Delivery of BDNF. Pharm. Res. 2010, 27, 82-91.

[117] Humphrey, W.; Dalke, A.; Schulten, K., VMD: Visual molecular dynamics. J.

Mol. Graph. Model. 1996, 14, 33-38.

[118] Harwood, L. M.; Moody, C. J., Experimental Organic Chemistry Principles

and Practice. Blackwell Science: 1989.

[119] Zaky, A.; Elbakry, A.; Ehmer, A.; Breunig, M.; Goepferich, A., The

mechanism of protein release from triglyceride microspheres. J. Control.

Release 2010, 147, 202-210.

[120] Mundargi, R. C.; Babu, V. R.; Rangaswamy, V.; Patel, P.; Aminabhavi, T.

M., Nano/micro technologies for delivering macromolecular therapeutics

- 224 -

using poly(D,L-lactide-co-glycolide) and its derivatives. J. Control. Release

2008, 125, 193-209.

[121] King, T. W.; Patrick, C. W., Development and in vitro characterization of

vascular endothelial growth factor (VEGF)-loaded poly(DL-lactic-co-

glycolic acid)/poly(ethylene glycol) microspheres using a solid

encapsulation/single emulsion/solvent extraction technique. J. Biomed.

Mater. Res. 2000, 51, 383-390.

[122] Liao, I.; Chew, S.; Leong, K., Aligned core–shell nanofibers delivering

bioactive proteins. Nanomedicine 2006, 1, 465-471.

[123] Johnson, P. J.; Skornia, S. L.; Stabenfetdt, S. E.; Willits, R. K., Maintaining

bioactivity of NGF for controlled release from PLGA using PEG. J. Biomed.

Mater. Res. A 2008, 86A, 420-427.

[124] Lu, L.; Stamatas, G. N.; Mikos, A. G., Controlled release of transforming

growth factor beta 1 from biodegradable polymer microparticles. J. Biomed.

Mater. Res. 2000, 50, 440-451.

[125] Bilati, U.; Allemann, E.; Doelker, E., Strategic approaches for overcoming

peptide and protein instability within biodegradable nano- and

microparticles. Eur. J. Pharm. Biopharm. 2005, 59, 375-388.

[126] Yang, Y. Y.; Chung, T. S.; Bai, X. L.; Chan, W. K., Effect of preparation

conditions on morphology and release profiles of biodegradable polymeric

microspheres containing protein fabricated by double-emulsion method.

Chem. Eng. J. 2000, 55, 2223-2236.

[127] Morita, T.; Sakamura, Y.; Horikiri, Y.; Suzuki, T.; Yoshino, H., Protein

encapsulation into biodegradable microspheres by a novel S/O/W emulsion

method using poly(ethylene glycol) as a protein micronization adjuvant. J.

Control. Release 2000, 69, 435-444.

[128] Mann, B. K.; Schmedlen, R. H.; West, J. L., Tethered-TGF-beta increases

extracellular matrix production of vascular smooth muscle cells. Biomaterials

2001, 22, 439-444.

[129] Zeugolis, D. I.; Khew, S. T.; Yew, E. S. Y.; Ekaputra, A. K.; Tong, Y. W.;

Yung, L.-Y. L.; Hutmacher, D. W.; Sheppard, C.; Raghunath, M., Electro-

spinning of pure collagen nano-fibres - Just an expensive way to make

gelatin? Biomaterials 2008, 29, 2293-2305.

[130] Smith, L. A.; Ma, P. X., Nano-fibrous scaffolds for tissue engineering.

Colloids Surf B Biointerfaces 2004, 39, 125-131.

[131] Xie, J. W.; Li, X. R.; Xia, Y. N., Putting Electrospun Nanofibers to Work for

Biomedical Research. Macromol. Rapid Commun. 2008, 29, 1775-1792.

[132] Pham, Q. P.; Sharma, U.; Mikos, A. G., Electrospinning of polymeric

nanofibers for tissue engineering applications: A review. Tissue Eng. 2006,

12, 1197-1211.

[133] Min, B. M.; Lee, G.; Kim, S. H.; Nam, Y. S.; Lee, T. S.; Park, W. H.,

Electrospinning of silk fibroin nanofibers and its effect on the adhesion and

spreading of normal human keratinocytes and fibroblasts in vitro.

Biomaterials 2004, 25, 1289-1297.

[134] Shin, M.; Ishii, O.; Sueda, T.; Vacanti, J. P., Contractile cardiac grafts using

a novel nanofibrous mesh. Biomaterials 2004, 25, 3717-3723.

[135] Piskin, E.; Isoglu, I. A.; Bolgen, N.; Vargel, I.; Griffiths, S.; Cavusoglu, T.;

Korkusuz, P.; Guzel, E.; Cartmell, S., In vivo performance of simvastatin-

loaded electrospun spiral-wound polycaprolactone scaffolds in

- 225 -

reconstruction of cranial bone defects in the rat model. J. Biomed. Mater.

Res. A 2009, 90A, 1137-1151.

[136] Zhao, L.; He, C. G.; Gao, Y. J.; Cen, L.; Cui, L.; Cao, Y. L., Preparation and

cytocompatibility of PLGA scaffolds with controllable fiber morphology and

diameter using electrospinning method. J. Biomed. Mater. Res. B 2008, 87B,

26-34.

[137] Sanders, E. H.; Kloefkorn, R.; Bowlin, G. L.; Simpson, D. G.; Wnek, G. E.,

Two-phase electrospinning from a single electrified jet: Microencapsulation

of aqueous reservoirs in poly(ethylene-co-vinyl acetate) fibers.

Macromolecules 2003, 36, 3803-3805.

[138] Jiang, H.; Hu, Y.; Li, Y.; Zhao, P.; Zhu, K.; Chen, W., A facile technique to

prepare biodegradable coaxial electrospun nanofibers for controlled release

of bioactive agents. J. Control. Release 2005, 108, 237-243.

[139] Nie, H.; Soh, B. W.; Fu, Y. C.; Wang, C. H., Three-dimensional fibrous

PLGA/HAp composite scaffold for BMP-2 delivery. Biotechnol. Bioeng.

2008, 99, 223-234.

[140] Li, C. M.; Vepari, C.; Jin, H. J.; Kim, H. J.; Kaplan, D. L., Electrospun silk-

BMP-2 scaffolds for bone tissue engineering. Biomaterials 2006, 27, 3115-

3124.

[141] Fu, Y. C.; Nie, H.; Ho, M. L.; Wang, C. K.; Wang, C. H., Optimized bone

regeneration based on sustained release from three-dimensional fibrous

PLGA/HAp composite scaffolds loaded with BMP-2. Biotechnol. Bioeng.

2008, 99, 996-1006.

[142] Choi, J. S.; Leong, K. W.; Yoo, H. S., In vivo wound healing of diabetic

ulcers using electrospun nanofibers immobilized with human epidermal

growth factor (EGF). Biomaterials 2008, 29, 587-596.

[143] Sill, T. J.; von Recum, H. A., Electrospinning: Applications in drug delivery

and tissue engineering. Biomaterials 2008, 29, 1989-2006.

[144] Ionescu, L. C.; Lee, G. C.; Sennett, B. J.; Burdick, J. A.; Mauck, R. L., An

anisotropic nanofiber/microsphere composite with controlled release of

biomolecules for fibrous tissue engineering. Biomaterials 2010, 31, 4113-

4120.

[145] Porter, J. R.; Ruckh, T. T.; Popat, K. C., Bone Tissue Engineering: A Review

in Bone Biomimetics and Drug Delivery Strategies. Biotechnol. Prog. 2009,

25, 1539-1560.

[146] Chen, F. M.; Chen, R.; Wang, X. J.; Sun, H. H.; Wu, Z. F., In vitro cellular

responses to scaffolds containing two microencapulated growth factors.

Biomaterials 2009, 30, 5215-5224.

[147] Biondi, M.; Ungaro, F.; Quaglia, F.; Netti, P. A., Controlled drug delivery in

tissue engineering. Adv. Drug Deliver. Rev. 2008, 60, 229-242.

[148] Richardson, T. P.; Peters, M. C.; Ennett, A. B.; Mooney, D. J., Polymeric

system for dual growth factor delivery. Nat. Biotechnol. 2001, 19, 1029-1034.

[149] Qi, H. X.; Hu, P.; Xu, J.; Wang, A. J., Encapsulation of drug reservoirs in

fibers by emulsion electrospinning: Morphology characterization and

preliminary release assessment. Biomacromolecules 2006, 7, 2327-2330.

[150] Dong, B.; Smith, M. E.; Wnek, G. E., Encapsulation of Multiple Biological

Compounds Within a Single Electrospun Fiber. Small 2009, 5, 1508-1512.

[151] Wang, Y. Z.; Qiao, W. L.; Wang, B. C.; Zhang, Y. Q.; Shao, P. Y.; Yin, T.

Y., Electrospun composite nanofibers containing nanoparticles for the

programmable release of dual drugs. Polymer J. 2011, 43, 478-483.

- 226 -

[152] Baker, B. M.; Gee, A. O.; Metter, R. B.; Nathan, A. S.; Marklein, R. A.;

Burdick, J. A.; Mauck, R. L., The potential to improve cell infiltration in

composite fiber-aligned electrospun scaffolds by the selective removal of

sacrificial fibers. Biomaterials 2008, 29, 2348-2358.

[153] Stankus, J. J.; Guan, J.; Fujimoto, K.; Wagner, W. R., Microintegrating

smooth muscle cells into a biodegradable, elastomeric fiber matrix.

Biomaterials 2006, 27, 735-744.

[154] Hong, Y.; Fujimoto, K.; Hashizume, R.; Guan, J.; Stankus, J. J.; Tobita, K.;

Wagner, W. R., Generating elastic, biodegradable polyurethane/poly(lactide-

co-glycolide) fibrous sheets with controlled antibiotic release via two-stream

electrospinning. Biomacromolecules 2008, 9, 1200-1207.

[155] Ekaputra, A. K.; Prestwich, G. D.; Cool, S. M.; Hutmacher, D. W.,

Combining electrospun scaffolds with electrosprayed hydrogels leads to

three-dimensional cellularization of hybrid constructs. Biomacromolecules

2008, 9, 2097-2103.

[156] Kim, S. J.; Jang, D. H.; Park, W. H.; Min, B. M., Fabrication and

characterization of 3-dimensional PLGA nanofiber/microfiber composite

scaffolds. Polymer 2010, 51, 1320-1327.

[157] Bonani, W.; Maniglio, D.; Motta, A.; Tan, W.; Migliaresi, C., Biohybrid

nanofiber constructs with anisotropic biomechanical properties. J. Biomed.

Mater. Res. B 2011, 96B, 276-286.

[158] Malda, J.; Klein, T. J.; Upton, Z., The roles of hypoxia in the In vitro

engineering of tissues. Tissue Eng. 2007, 13, 2153-2162.

[159] Jaworek, A.; Krupa, A.; Lackowski, M.; Sobczyk, A. T.; Czech, T.;

Ramakrishna, S.; Sundarrajan, S.; Pliszka, D., Nanocomposite fabric

formation by electrospinning and electrospraying technologies. J. Electrostat.

2009, 67, 435-438.

[160] Gupta, D.; Venugopal, J.; Mitra, S.; Giri Dev, V. R.; Ramakrishna, S.,

Nanostructured biocomposite substrates by electrospinning and

electrospraying for the mineralization of osteoblasts. Biomaterials 2009, 30,

2085-2094.

[161] Francis, L.; Venugopal, J.; Prabhakaran, M. P.; Thavasi, V.; Marsano, E.;

Ramakrishna, S., Simultaneous electrospin-electrosprayed biocomposite

nanofibrous scaffolds for bone tissue regeneration. Acta Biomater. 2010, 6,

4100-4109.

[162] Stankus, J. J.; Soletti, L.; Fujimoto, K.; Hong, Y.; Vorp, D. A.; Wagner, W.

R., Fabrication of cell microintegrated blood vessel constructs through

electrohydrodynamic atomization. Biomaterials 2007, 28, 2738-2746.

[163] Park, J. S.; Lee, J. H.; Shin, H. S.; Lee, T. W.; Kim, M. S.; Khang, G.; Rhee,

J. M.; Lee, H. K.; Lee, H. B., Biodegradable polymer microspheres for

controlled drug release. Tissue Eng. Regen. Med. 2007, 4, 347-359.

[164] Kempen, D. H. R.; Kim, C. W.; Lu, L.; Dhert, W. J. A.; Currier, B. L.;

Yaszemski, M. J., Controlled release from poly(lactic-co-glycolic acid)

microspheres embedded in an injectable, biodegradable scaffold for bone

tissue engineering. In Thermec'2003, Pts 1-5, Chandra, T.; Torralba, J. M.;

Sakai, T., Eds. Trans Tech Publications Ltd: Zurich-Uetikon, 2003; 426-4,

3151-3156.

[165] Carrascosa, C.; Torres-Aleman, I.; Lopez-Lopez, C.; Carro, E.; Espejo, L.;

Torrado, S.; Torrado, J. J., Microspheres containing insulin-like growth

- 227 -

factor I for treatment of chronic neurodegeneration. Biomaterials 2004, 25,

707-714.

[166] Zeleny, J., The Electrical Discharge from Liquid Points, and a Hydrostatic

Method of Measuring the Electric Intensity at Their Surfaces. Physic. Rev.

1914, 3, 69.

[167] Taylor, G., Disintegration of Water Drops in an Electric Field. Proceedings

of the Royal Society of London. Series A, Mathematical and Physical

Sciences 1964, 280, 383-397.

[168] Fantini, D.; Zanetti, M.; Costa, L., Polystyrene microspheres and

nanospheres produced by electrospray. Macromol. Rapid Commun. 2006,

27, 2038-2042.

[169] Loscertales, I. G.; Barrero, A.; Guerrero, I.; Cortijo, R.; Marquez, M.; Ganan-

Calvo, A. M., Micro/nano encapsutation via electrified coaxial liquid jets.

Science 2002, 295, 1695-1698.

[170] Crescenzi, V.; Manzini, G.; Calzolari, G.; Borri, C., Thermodynamics of

fusion of poly-[beta]-propiolactone and poly-[epsilon]-caprolactone.

comparative analysis of the melting of aliphatic polylactone and polyester

chains. Eur. Polym. J. 1972, 8, 449-463.

[171] Wu, F.; Jin, T., Polymer-Based Sustained-Release Dosage Forms for Protein

Drugs, Challenges, and Recent Advances. AAPS PharmSciTech 2008, 9,

1218-1229.

[172] Morita, T.; Horikiri, Y.; Yamahara, H.; Suzuki, T.; Yoshino, H., Formation

and isolation of spherical fine protein microparticles through lyophilization

of protein-poly(ethylene glycol) aqueous mixture. Pharm. Res. 2000, 17,

1367-1373.

[173] Kasper, J. C.; Winter, G.; Friess, W., Recent advances and further challenges

in lyophilization. Eur. J. Pharm. Biopharm. 2013, 85, 162-169.

[174] Carpenter, J. F.; Crowe, J. H., The Mechanism of Cryoprotection of Proteins

by Solutes. Cryobiology 1988, 25, 244-255.

[175] Rawat, S.; Kohli, N.; Suri, C. R.; Sahoo, D. K., Molecular Mechanism of

improved Structural Integrity of Protein in Polymer Based Microsphere

Delivery System. Mol. Pharm. 2012, 9, 2403-2414.

[176] Giteau, A.; Venier-Julienne, M. C.; Aubert-Pouëssel, A.; Benoit, J. P., How

to achieve sustained and complete protein release from PLGA-based

microparticles? Int. J. Pharm. 2008, 350, 14-26.

[177] Morita, T.; Horikiri, Y.; Suzuki, T.; Yoshino, H., Preparation of gelatin

microparticles by co-lyophilization with poly(ethylene glycol):

characterization and application to entrapment into biodegradable

microspheres. Int. J. Pharm. 2001, 219, 127-137.

[178] Al-Azzam, W.; Pastrana, E. A.; Griebenow, K., Co-lyophilization of bovine

serum albumin (BSA) with poly(ethylene glycol) improves efficiency of BSA

encapsulation and stability in polyester microspheres by a solid-in-oil-in-oil

technique. Biotechnol. Lett. 2002, 24, 1367-1374.

[179] Mok, H.; Park, T. G., Water-free microencapsulation of proteins within

PLGA microparticles by spray drying using PEG-assisted protein

solubilization technique in organic solvent. Eur. J. Pharm. Biopharm. 2008,

70, 137-144.

[180] Jiang, W. L.; Schwendeman, S. P., Stabilization and controlled release of

bovine serum albumin encapsulated in poly(D, L-lactide) and poly(ethylene

glycol) microsphere blends. Pharm. Res. 2001, 18, 878-885.

- 228 -

[181] Bock, N.; Dargaville, T. R.; Woodruff, M. A., Electrospraying of polymers

with therapeutic molecules: State of the art. Prog. Polym. Sci. 2012, 37,

1510-1551.

[182] Bock, N.; Woodruff, M. A.; Steck, R.; Hutmacher, D. W.; Farrugia, B. L.;

Dargaville, T. R., Composites for Delivery of Therapeutics: Combining Melt

Electrospun Scaffolds with Loaded Electrosprayed Microparticles.

Macromol. Biosci. 2013, 14, 202-214.

[183] Flory, P. J., Citation Classic - Thermodynamics of High Polymer-Solutions.

Current Contents/Engineering Technology & Applied Sciences 1985, 18-18.

[184] Huggins, M. L., Thermodynamic Properties of Solutions of Long-chain

Compounds. New York Academy of Sciences: 1942.

[185] Hansen, C. M., Hansen solubility parameters : a user's handbook. CRC

Press: Boca Raton, Fla., 2000.

[186] Hansen, C. M.; Abbott, S., Hansen Solubility Parameters in Practice.

Hansen-Solubility.com: 2008; p 111.

[187] Hildebrand, J. H., Solubility of Non-electrolytes. Reinhold Publishing

Corporation: 1936.

[188] Matsumoto, A.; Matsukawa, Y.; Suzuki, T.; Yoshino, H.; Kobayashi, M., The

polymer-alloys method as a new preparation method of biodegradable

microspheres: principle and application to cisplatin-loaded microspheres. J.


[189] Masuelli, M. A., Study of Bovine Serum Albumin Solubility in Aqueous

Solutions by Intrinsic Viscosity Measurements. Advances in Physical

Chemistry 2013, 2013, 8.

[190] Vay, K.; Scheler, S.; Friess, W., Application of Hansen solubility parameters

for understanding and prediction of drug distribution in microspheres. Int. J.

Pharm. 2011, 416, 202-209.

[191] Gander, B.; Johansen, P.; NamTran, H.; Merkle, H. P., Thermodynamic

approach to protein microencapsulation into poly(D,L-lactide) by spray

drying. Int. J. Pharm. 1996, 129, 51-61.

[192] Barton, A. F. M., CRC Handbook of Solubility Parameters and Other

Cohesion Parameters. CRC Press: 1991.

[193] Schenderlein, S.; Luck, M.; Muller, B. W., Partial solubility parameters of

poly(D,L-lactide-co-glycolide). Int. J. Pharm. 2004, 286, 19-26.

[194] Bordes, C.; Fréville, V.; Ruffin, E.; Marote, P.; Gauvrit, J. Y.; Briançon, S.;

Lantéri, P., Determination of poly(ɛ-caprolactone) solubility parameters:

Application to solvent substitution in a microencapsulation process. Int. J.

Pharm. 2010, 383, 236-243.

[195] Thakral, S.; Thakral, N. K., Prediction of drug-polymer miscibility through

the use of solubility parameter based flory-huggins interaction parameter and

the experimental validation: PEG as model polymer. J. Pharm. Sci. 2013,

102, 2254-2263.

[196] Adamska, K.; Voelkel, A., Hansen solubility parameters for polyethylene

glycols by inverse gas chromatography. J. Chromatogr. A 2006, 1132, 260-

267.

[197] Pean, J. M.; Boury, F.; Venier-Julienne, M. C.; Menei, P.; Proust, J. E.;

Benoit, J. P., Why does PEG 400 co-encapsulation improve NGF stability

and release from PLGA biodegradable microspheres? Pharm. Res. 1999, 16,

1294-1299.

- 229 -

[198] Sah, H. K., A new strategy to determine the actual protein content of

poly(lactide-co-glycolide) microspheres. J. Pharm. Sci. 1997, 86, 1315-1318.

[199] Lochmann, A.; Nitzsche, H.; von Einem, S.; Schwarz, E.; Mäder, K., The

influence of covalently linked and free polyethylene glycol on the structural

and release properties of rhBMP-2 loaded microspheres. J. Control. Release

2010, 147, 92-100.

[200] Anderson, J. M.; Shive, M. S., Biodegradation and biocompatibility of PLA

and PLGA microspheres. Adv. Drug Deliver. Rev. 2012, 64, Supplement, 72-

82.

[201] Arakawa, T.; Prestrelski, S. J.; Kenney, W. C.; Carpenter, J. F., Factors

affecting short-term and long-term stabilities of proteins. Adv. Drug Deliver.

Rev. 2001, 46, 307-326.

[202] Cleland, J. L.; Duenas, E. T.; Park, A.; Daugherty, A.; Kahn, J.; Kowalski, J.;

Cuthbertson, A., Development of poly-(D,L-lactide-coglycolide) microsphere

formulations containing recombinant human vascular endothelial growth

factor to promote local angiogenesis. J. Control. Release 2001, 72, 13-24.

[203] Amsden, B. G.; Timbart, L.; Marecak, D.; Chapanian, R.; Tse, M. Y.; Pang,

S. C., VEGF-induced angiogenesis following localized delivery via

injectable, low viscosity poly(trimethylene carbonate). J. Control. Release

2010, 145, 109-115.

[204] Expert Opinion on Drug DeliveryWang, W., Protein aggregation and its

inhibition in biopharmaceutics. Int. J. Pharm. 2005, 289, 1-30.

[205] van de Weert, M.; Hennink, W. E.; Jiskoot, W., Protein Instability in

Poly(Lactic-co-Glycolic Acid) Microparticles. Pharma. Res. 2000, 17, 1159-

1167.

[206] Bock, N.; Dargaville, T. R.; Woodruff, M. A., Controlling

Microencapsulation and Release of Micronized Proteins using Poly(Ethylene

Glycol) and Electrospraying Eur. J. Pharm. Biopharm. 2014, n/a.

[207] Rui, J.; Dadsetan, M.; Runge, M. B.; Spinner, R. J.; Yaszemski, M. J.;

Windebank, A. J.; Wang, H., Controlled release of vascular endothelial

growth factor using poly-lactic-co-glycolic acid microspheres: In vitro

characterization and application in polycaprolactone fumarate nerve

conduits. Acta Biomater. 2012, 8, 511-518.

[208] Yeh, L. C. C.; Tsai, A. D.; Lee, J. C., Osteogenic protein-1 (OP-1, BMP-7)

induces osteoblastic cell differentiation of the pluripotent mesenchymal cell

line C2C12. J. Cell. Biochem. 2002, 87, 292-304.

[209] Wang, C.-K.; Ho, M.-L.; Wang, G.-J.; Chang, J.-K.; Chen, C.-H.; Fu, Y.-C.;

Fu, H.-H., Controlled-release of rhBMP-2 carriers in the regeneration of

osteonecrotic bone. Biomaterials 2009, 30, 4178-4186.

[210] Crotts, G.; Park, T. G., Stability and release of bovine serum albumin

encapsulated within poly(D,L-lactide-co-glycolide) microparticles. J.


[211] van der Walle, C. F.; Sharma, G.; Kumar, M. N. V. R., Current approaches

to stabilising and analysing proteins during microencapsulation in PLGA.

Exp. Op. Drug Deliv. 2009, 6, 177-186.

[212] Miyaki, M.; Fujimoto, K.; Kawaguchi, H., Cell response to micropatterned

surfaces produced with polymeric microspheres. Colloid Surface A 1999,

153, 603-608.

[213] Hollister, S. J., Porous scaffold design for tissue engineering. Nat. Mater.

2005, 4, 518-524.

- 230 -

[214] Murugan, R.; Ramakrishna, S., Nano-featured scaffolds for tissue

engineering: A review of spinning methodologies. Tissue Eng. 2006, 12, 435-

447.

[215] Reneker, D. H.; Yarin, A. L.; Zussman, E.; Xu, H., Electrospinning of

nanofibers from polymer solutions and melts. In Adv. Appl. Mech., Aref, H.;

VanDerGiessen, E., Eds. 2007; 41, 43-195.

[216] Brown, T. D.; Dalton, P. D.; Hutmacher, D. W., Direct Writing By Way of

Melt Electrospinning. Adv. Mater. 2011, 23, 5651-5657.

[217] Farrugia, B. L.; Brown, T. D.; Upton, Z.; Hutmacher, D. W.; Dalton, P. D.;

Dargaville, T. R., Dermal fibroblast infiltration of poly(ε-caprolactone)

scaffolds fabricated by melt electrospinning in a direct writing mode.

Biofabrication 2013, 5, 025001.

[218] Dalton, P. D.; Vaquette, C.; Farrugia, B. L.; Dargaville, T. R.; Brown, T. D.;

Hutmacher, D. W., Electrospinning and additive manufacturing: converging

technologies. Biomater. Sci. 2013, 1, 171-185.

[219] Lee, S.-H.; Shin, H., Matrices and scaffolds for delivery of bioactive

molecules in bone and cartilage tissue engineering. Adv. Drug Deliver. Rev.

2007, 59, 339-359.

[220] Blackwood, K. A.; Bock, N.; Dargaville, T. R.; Woodruff, M. A., Scaffolds

for Growth Factor Delivery as Applied to Bone Tissue Engineering.

International Journal of Polymer Science 2012.

[221] Bhaskar, S.; Pollock, K. M.; Yoshida, M.; Lahann, J., Towards Designer

Microparticles: Simultaneous Control of Anisotropy, Shape, and Size. Small

2010, 6, 404-411.

[222] Brown, T. D.; Slotosch, A.; Thibaudeau, L.; Taubenberger, A.; Loessner, D.;

Vaquette, C.; Dalton, P. D.; Hutmacher, D. W., Design and Fabrication of

Tubular Scaffolds via Direct Writing in a Melt Electrospinning Mode.

Biointerphases 2012, 7.

[223] Hildebrand, T.; Ruegsegger, P., A new method for the model-independent

assessment of thickness in three-dimensional images. J. Microsc. 1997, 185,

67-75.

[224] Jones, K. H.; Senft, J. A., An improved method to determine cell viability by

simultaneous staining with fluorescien diacetate propidium iodide. J.

Histochem. Cytochem. 1985, 33, 77-79.

[225] Vaquette, C.; Cooper-White, J., A simple method for fabricating 3-D

multilayered composite scaffolds. Acta Biomater. 2013, 9, 4599-4608.

[226] Vaquette, C.; Cooper-White, J., The use of an electrostatic lens to enhance

the efficiency of the electrospinning process. Cell Tissue Res. 2012, 347, 815-

826.

[227] Vaquette, C.; Cooper-White, J. J., Increasing electrospun scaffold pore size

with tailored collectors for improved cell penetration. Acta Biomater. 2011,

7, 2544-2557.

[228] von Burkersroda, F.; Schedl, L.; Göpferich, A., Why degradable polymers

undergo surface erosion or bulk erosion. Biomaterials 2002, 23, 4221-4231.

[229] Li, S. M.; GirodHolland, S.; Vert, M., Hydrolytic degradation of poly(DL-

lactic acid) in the presence of caffeine base. J. Control. Release 1996, 40, 41-

53.

[230] Siepmann, J.; Elkharraz, K.; Siepmann, F.; Klose, D., How Autocatalysis

Accelerates Drug Release from PLGA-Based Microparticles: A Quantitative

Treatment. Biomacromolecules 2005, 6, 2312-2319.

- 231 -

[231] Diekman, B. O.; Rowland, C. R.; Lennon, D. P.; Caplan, A. I.; Guilak, F.,

Chondrogenesis of Adult Stem Cells from Adipose Tissue and Bone Marrow:

Induction by Growth Factors and Cartilage-Derived Matrix. Tissue Eng. A

2010, 16, 523-533.

[232] Nair, L. S.; Laurencin, C. T., Biodegradable polymers as biomaterials. Prog.

Polym. Sci. 2007, 32, 762-798.

[233] Ivan S. Gutzow, O. V. M., Jürn W. P. Schmelzer, Snejana V. Todorova, Boris

B. Petroff, Alexander I. Priven, Glasses and the Glass Transition. 2011.

[234] Wunderlich, B., Thermal Analysis of Polymeric Materials. Springer: 2005.

[235] Castelli, F.; Gilbert, S. M.; Caruso, S.; MacCarrone, D. E.; Fisichella, S.,

Thermoanalytical characterization of high molecular weight glutenin

subunits: Water effect on their glass transition. Thermochim. Acta 2000, 346,

153-160.

[236] Dunne, M.; Corrigan, O. I.; Ramtoola, Z., Influence of particle size and

dissolution conditions on the degradation properties of polylactide-co-

glycolide particles. Biomaterials 2000, 21, 1659-1668.

[237] Hausberger, A. G.; DeLuca, P. P., Characterization of biodegradable

poly(d,l-lactide-co-glycolide) polymers and microspheres. J. Pharmaceut.

Biomed. 1995, 13, 747-760.

[238] Park, T. G., Degradation of poly(lactic-co-glycolic acid) microspheres: effect

of copolymer composition. Biomaterials 1995, 16, 1123-1130.

[239] Blanco, M. D.; Sastre, R. L.; Teijon, C.; Olmo, R.; Teijon, J. M., Degradation

behaviour of microspheres prepared by spray-drying poly(D,L-lactide) and

poly(D,L-lactide-co-glycolide) polymers. Inter. J. Pharm. 2006, 326, 139-147.

[240] Lemoine, D.; Francois, C.; Kedzierewicz, F.; Preat, W.; Hoffman, M.;

Maincent, P., Stability study of nanoparticles of poly(epsilon-caprolactone),

poly(D,L-lactide) and poly(D,L-lactide-co-glycolide). Biomaterials 1996, 17,

2191-2197.

[241] Lu, W.; Park, T. G., Protein release from poly(lactic-co-glycolic acid)

microspheres: Protein stability problems. PDA J. Pharm. Sci. Tech. 1995, 49,

13-19.

[242] Boury, F.; Marchais, H.; Proust, J. E.; Benoit, J. P., Bovine serum albumin

release from poly(α-hydroxy acid) microspheres: effects of polymer

molecular weight and surface properties. J. Control. Release 1997, 45, 75-86.

[243] Clark, P.; Connolly, P.; Curtis, A. S. G.; Dow, J. A. T.; Wilkinson, C. D. W.,

Topographical control of cell behavior .1. Simple step cues. Development

1987, 99, 439-448.

[244] Curtis, A.; Wilkinson, C., Topographical control of cells. Biomaterials 1997,

18, 1573-1583.

Documents

MOLECULES USING ELECTROSPRAYEDeprints.qut.edu.au/74514/1/Nathalie_Bock_Thesis.pdfMicrospheres Loaded with Growth Factors (Poster) - vi - List of Postgraduate Conferences Australian