Citation
Deformable Microparticles with Multiple Functions for Drug Delivery and Device Testing

Material Information

Title:
Deformable Microparticles with Multiple Functions for Drug Delivery and Device Testing
Creator:
THULA, TAILI T. ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alginates ( jstor )
Blood ( jstor )
Capsules ( jstor )
Crosslinking ( jstor )
Diameters ( jstor )
Erythrocytes ( jstor )
Ions ( jstor )
Particle size distribution ( jstor )
Polymers ( jstor )
Zinc ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Taili T. Thula. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/12/2007
Resource Identifier:
659898557 ( OCLC )

Downloads

This item has the following downloads:

thula_t ( .pdf )

thula_t_Page_069.txt

thula_t_Page_012.txt

thula_t_Page_090.txt

thula_t_Page_111.txt

thula_t_Page_106.txt

thula_t_Page_156.txt

thula_t_Page_007.txt

thula_t_Page_108.txt

thula_t_Page_020.txt

thula_t_Page_076.txt

thula_t_Page_044.txt

thula_t_Page_105.txt

thula_t_Page_013.txt

thula_t_Page_081.txt

thula_t_Page_080.txt

thula_t_Page_119.txt

thula_t_Page_077.txt

thula_t_Page_057.txt

thula_t_Page_150.txt

thula_t_Page_051.txt

thula_t_Page_030.txt

thula_t_Page_067.txt

thula_t_Page_088.txt

thula_t_Page_050.txt

thula_t_Page_093.txt

thula_t_Page_146.txt

thula_t_Page_005.txt

thula_t_Page_130.txt

thula_t_Page_122.txt

thula_t_Page_142.txt

thula_t_Page_063.txt

thula_t_Page_115.txt

thula_t_Page_152.txt

thula_t_Page_114.txt

thula_t_Page_098.txt

thula_t_Page_121.txt

thula_t_Page_041.txt

thula_t_Page_147.txt

thula_t_Page_038.txt

thula_t_Page_052.txt

thula_t_Page_084.txt

thula_t_Page_143.txt

thula_t_Page_145.txt

thula_t_Page_095.txt

thula_t_Page_025.txt

thula_t_Page_037.txt

thula_t_Page_054.txt

thula_t_Page_061.txt

thula_t_Page_046.txt

thula_t_Page_159.txt

thula_t_Page_035.txt

thula_t_Page_034.txt

thula_t_Page_016.txt

thula_t_Page_024.txt

thula_t_Page_032.txt

thula_t_Page_074.txt

thula_t_Page_028.txt

thula_t_Page_010.txt

thula_t_Page_118.txt

thula_t_Page_082.txt

thula_t_Page_116.txt

thula_t_Page_065.txt

thula_t_Page_079.txt

thula_t_Page_086.txt

thula_t_Page_092.txt

thula_t_Page_157.txt

thula_t_Page_047.txt

thula_t_Page_070.txt

thula_t_Page_125.txt

thula_t_Page_097.txt

thula_t_Page_029.txt

thula_t_Page_004.txt

thula_t_Page_123.txt

thula_t_Page_101.txt

thula_t_Page_058.txt

thula_t_Page_059.txt

thula_t_Page_141.txt

thula_t_Page_158.txt

thula_t_Page_120.txt

thula_t_Page_019.txt

thula_t_Page_040.txt

thula_t_Page_003.txt

thula_t_Page_131.txt

thula_t_Page_155.txt

thula_t_Page_089.txt

thula_t_Page_128.txt

thula_t_Page_001.txt

thula_t_Page_011.txt

thula_t_Page_026.txt

thula_t_Page_113.txt

thula_t_Page_107.txt

thula_t_Page_160.txt

thula_t_Page_161.txt

thula_t_Page_134.txt

thula_t_Page_154.txt

thula_t_Page_102.txt

thula_t_pdf.txt

thula_t_Page_027.txt

thula_t_Page_071.txt

thula_t_Page_002.txt

thula_t_Page_021.txt

thula_t_Page_060.txt

thula_t_Page_049.txt

thula_t_Page_017.txt

thula_t_Page_066.txt

thula_t_Page_129.txt

thula_t_Page_075.txt

thula_t_Page_053.txt

thula_t_Page_153.txt

thula_t_Page_078.txt

thula_t_Page_014.txt

thula_t_Page_045.txt

thula_t_Page_042.txt

thula_t_Page_133.txt

thula_t_Page_068.txt

thula_t_Page_031.txt

thula_t_Page_062.txt

thula_t_Page_127.txt

thula_t_Page_110.txt

thula_t_Page_064.txt

thula_t_Page_144.txt

thula_t_Page_008.txt

thula_t_Page_043.txt

thula_t_Page_036.txt

thula_t_Page_023.txt

thula_t_Page_126.txt

thula_t_Page_137.txt

thula_t_Page_124.txt

thula_t_Page_104.txt

thula_t_Page_009.txt

thula_t_Page_055.txt

thula_t_Page_135.txt

thula_t_Page_140.txt

thula_t_Page_072.txt

thula_t_Page_149.txt

thula_t_Page_039.txt

thula_t_Page_136.txt

thula_t_Page_094.txt

thula_t_Page_091.txt

thula_t_Page_022.txt

thula_t_Page_151.txt

thula_t_Page_099.txt

thula_t_Page_048.txt

thula_t_Page_018.txt

thula_t_Page_138.txt

thula_t_Page_100.txt

thula_t_Page_132.txt

thula_t_Page_109.txt

thula_t_Page_033.txt

thula_t_Page_112.txt

thula_t_Page_103.txt

thula_t_Page_006.txt

thula_t_Page_056.txt

thula_t_Page_148.txt

thula_t_Page_117.txt

thula_t_Page_087.txt

thula_t_Page_096.txt

thula_t_Page_139.txt

thula_t_Page_085.txt

thula_t_Page_073.txt

thula_t_Page_015.txt

thula_t_Page_083.txt


Full Text





DEFORMABLE MICROPARTICLES WITH MULTIPLE FUNCTIONS FOR DRUG
DELIVERY AND DEVICE TESTING




















By

TAILI T. THULA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007

































2007 Taili T. Thula
































To my parents for their unconditional love









ACKNOWLEDGMENTS

I thank God for giving me the opportunity to accomplish my goals and to meet wonderful

people who supported and helped me. I thank my parents for their support and endless love. I

thank Asdrubal for always being there for me, especially when I most needed support. I thank

my little brother, Jeancho, for all his help counting endless batches of particles. Also, I thank my

two other musketeers: Jompo and John. You guys made it a fun journey and I could not have

done without you. Finally, I thank my professors (Dr. Christopher Batich, Dr. Gregory Schultz,

Dr. Roger Tran-Son-Tay and Dr. Malisa Sarntinoranont) for all of their support throughout this

lengthy project. I thank them for sharing their knowledge and for having faith in me.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

LIST O F TA BLE S ......... .... ........................................................................... 8

LIST OF FIGURES .................................. .. .... ..... ................. 10

A B S T R A C T ............ ................... ............................................................ 13

CHAPTER

1. INTRODUCTION ............... ..................................................... ..... 15

2. B A CK G R O U N D .......................................... .... ............ ....................... .... 19

Blood: Composition, Properties, and Functions .............. ..............................................19
C om p o sition ............................................................................. 19
P ro p e rtie s ................................................................................................................... 2 0
Functions ....... .... ............ ............. ..... .............. 20
Red Blood Cell: Morphology, Functions, Rheology, and Hematopoiesis ...........................21
M orphology and Functions........................................... ................... ............... 21
R heological Properties............ ... .......................................................... ................... 23
Hem atopoiesis ............................................. ..... 23
Red Blood Cell as a Model for Drug Delivery.....................................................25
Encapsulation M methods ................................ .. ... ........................... ...............25
Advantages and Disadvantages of Red Blood Cells in Drug Delivery ...........................26
Drug D delivery System s....................................................... .......... ......... ..... 27
M icrospheres and M icrocapsules .............................................................................28
M icroencapsulation Techniques......................................................... ................ ..... 29
Immune System Response to Foreign Elements .................................. ...............32

3. CHARACTERIZATION OF THE DEFORMABLE BEHAVIOR OF PLURONIC-
PLA AND POLYELECTROLYTE CHITOSAN-ALGINATE MICROPARTICLES .........40

In tro d u ctio n ................... ...................4...................0..........
Poly(Lactic A cid) ..................................... .............................41
P lu ro n ic s ................................................................................................................. 4 2
A lg in a te ..................................................................................................................... 4 3
C h ito s a n ..................................................................................................................... 4 5
Materials and Methods ................................................ 46
Preparation of PLA-Pluronic Particles ...................... .................. ............ 46
Preparation of Alginate Particles ......................................... .. ......................46
Coating of A lginate Particles............... .................................................. ..................... 47
D ecom position of the Particles' Core......................................... ......................... 48
Measurement of Particles' Deformability .............. ..................... .................48









Scanning Electron Microscopy (SEM) Analysis................................ ...............49
Results and Discussion ...................................... .. ......... ....... ..... 49
PLA -Pluronic Particles .................................................................. ............... 49
Chitosan-A lginate Particles .............................................................................52
C o n c lu sio n ................................ .............. .......................................... 5 4

4. SYNTHESIS AND CHARACTERIZATION OF PHYSICO-CHEMICAL AND
RHEOLOGY PROPERTIES OF POLYELECTROLYTE CHITOSAN-ALGINATE
MICROPARTICLES CROSSLINKED WITH CALCIUM, ZINC, OR COPPER IONS .....64

In tro d u ctio n ................... ...................6...................4..........
M materials and M methods ................................... ... .. .......... ....... ...... 65
Preparation of PEC Particles ...................................... ........... ............................. 65
A ir-spray crosslinking technique ........................................ ........................ 65
W/O emulsion gelation technique ................... .. ................. ............... 66
C oating of alginate particles.......................................................... ............... 67
D ecom position of the particles' core ............................................ ............... 67
Characterization of PEC Particles ................................................_...... ............... .... 68
Surface m orphology analysis ............................................................................68
Particle size distribution ...................... .. .. ................... ................... ............... 68
Particle stability analysis........................... ........ ............................ 69
M easurement of particles' deformability ...................................... ............... 70
A ssessm ent of particles' viscosity............................................................... .. 70
Coating adsorption analysis .................................. .....................................71
Ion exchange analysis........................... ................... .... ........... 72
A ssessm ent of particle cytotoxicity................................... .................................... 72
Data Interpretation................ ......... .. .... ..... ..............74
Results and Discussion ...................................... .. ......... ....... ..... 74
C o n clu sio n ................... ...................8...................9..........

5. ENCAPSULATION AND RELEASE OF BOVINE SERUM ALBUMIN FROM
POLYELECTROLYTE CHITOSAN-ALGINATE MICROPARTICLES
CROSSLINKED WITH ZINC OR COPPER IONS ............................... .....................120

Introduction ............ ....... .................... ...............120
M materials and M methods ............................................................................. ........ 121
Preparation of FITC-Labeled BSA (F-BSA)............... ........................................121
M icro-C ore P reparation ........................................................................ ....... ........... 122
C eating of A lginate Particles ......... ..................................................... ............... .. 122
Characterization ofF-BSA-Loaded Microcapsules ............................................... 123
Surface m orphology analysis ...........................................................................123
Particle size distribution .......................................................... ............... 123
Coating adsorption analysis ............................................................................124
Determination of protein loading efficiency ............................... ....................124
In vitro release kinetics of F-BSA microcapsules............................................... 125
Bovine serum albumin ELISA ....... .............................. ..................... 125
D ata Interpretation ........ ......... ........ ....... .......... ............ 126


6









R esu lts an d D iscu ssion ............................................................................. .. .................. 12 6
C o n clu sio n ................... ...................1...................3.........2

6. CONCLUSION AND FUTURE WORK ........................................................................ 144

L IST O F R E F E R E N C E S .............................................................................. ..........................15 1

B IO G R A PH IC A L SK E T C H .......................................................................... ....................... 161















































7









LIST OF TABLES


Table page

2-1 Plasma protein concentrations (mg/100 mL) ....................................... ...............35

2-2 Physicochemical properties of the major plasma proteins ............................................36

2-3 Concentrations of major electrolytes (mEq/L) in whole blood and plasma .....................37

2-4 Concentrations of various organic compounds (mg/100mL) in whole blood and
p la sm a ......................................................... ................................... 3 7

2-5 Controlled release drug delivery system s ............................................... ............... 39

2-6 Examples of commercialized PLGA copolymers microspheres ....................................39

3-1 List of commercially available Pluronics ............................... .. ....................... 56

3-2 List of Pluronic-PLA particles made. ............................. .......................... ........ 58

3-3 List of chitosan-alginate particles m ade ........................................ ........................ 59

3-4 Composition and deformability properties of Pluronic-PLA microspheres ...................61

3-5 Composition and deformability properties of PEC chitosan-alginate microcapsules. ......63

4-1 The PEC chitosan-alginate particle samples made ................................. ..................... 91

4-2 Concentrations of major electrolytes (mmol/L) in human plasma and cell culture
m edia......... ........................... ..................................... ......... ...... 9 1

4-3 Particle size diameter of copper-crosslinked alginate particles as a function of the
n eed le g au g e ......................................................................... 9 2

4-4 Particle size diameter of copper-crosslinked alginate particles as a function of
alg in ate flow rate...................................................... ................ 9 3

4-5 Particle size diameter of copper-crosslinked alginate particles as a function of
alg in ate flow rate...................................................... ................ 9 5

4-6 Particle size diameter of copper-crosslinked alginate particles as a function of the
cro sslinking ion ...........................................................................97

4-7 Particle size distribution of uncoated Ca-, Zn-, and Cu-crosslinked alginate particles ...101

4-8 Composition and deformability properties of PEC chitosan-alginate microcapsules ..... 110

4-9 Viscosity values of PEC chitosan-alginate microcapsules ............................................. 113









4-10 Zeta potential values of PEC chitosan-alginate microcapsules. .................................... 115

4-11 Concentration values of copper ions released from PEC chitosan-alginate
m icro c ap su le s ................................................... ................. ................ 1 16

4-12 Fibroblast cell survival and recovery percentages as a function of extract
concentrations from PEC m icroparticles ................................. ..................................... 118

5-1 The PEC chitosan-alginate particle samples made ......................................................... 133

5-2 Zeta potential values of PEC F-BSA-loaded chitosan-alginate microcapsules............. 140

5-3 Encapsulation efficiency of F-BSA loaded microparticles prepared from zinc- and
copper-crosslinked alginate gels.............................................. ............................ 142

5-4 Release kinetic profile of F-BSA from uncoated and multiple PEC coated zinc- and
copper-alginate microspheres during a 48-hour study.....................................................143









LIST OF FIGURES


Figure page

2-1 Scanning electron micrograph of red blood cell aggregates, rouleaux..............................38

2-2 Shear rate dependence of normal human blood viscoelasticity.............. ...................38

3-1 PL A m molecular structure. ......................................................................... .....................56

3-2 A lginate m molecular structure .............................................................................. ....... 57

3-3 C hitosan m molecular structure .............................................................................. ....... 57

3-4 Optical micrograph of 20% Pluronic-PLA (1:2) microspheres................................60

3-5 Optical micrograph of 20% P105-PLA (1:2) microspheres. ........................ ..................61

3-6 Scanning electron micrograph of 20% P105-PLA Particles.................. ..................61

3-7 Optical and scanning electron micrographs of 0.3% COS Ca-crosslinked alginate
p article s. ........ ........ ......................................................................... 6 2

3-8 Optical and scanning electron micrograph of 3% COS Cu-crosslinked alginate
particles, 5 coatings..................... ......... .. ... ....... ....... ..... 62

4-1 Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of the needle gauge. ............................ 92

4-2 Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of alginate flow rate.......................................93

4-3 Optical photomicrographs of copper-crosslinked alginate particles made with the air-
spray crosslinking technique ........................... ....................................... ............... 94

4-4 Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of air pressure. .............................................95

4-5 Optical photomicrographs of copper-crosslinked alginate particles made with the air-
spray crosslinking technique ........................... ....................................... ............... 96

4-6 Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of the crosslinking ion. .............................97

4-7 Optical photomicrographs of calcium-crosslinked alginate particles made with the
air-spray crosslinking technique. .............................................. ............................. 98

4-8 Optical micrographs of alginate microspheres made with the W/O emulsion gelation
tech n iqu e ........ ........ ........................................................................9 9









4-9 Particle size distribution. Percentage of total number of uncoated microspheres
versus particle diam eter. ........................................................................ ....................100

4-10 Optical photographs of Ca-, Zn-, and Cu-crosslinked alginate particles made with the
air-spray crosslinking technique. ............................................ ............................. 102

4-11 Scanning electron micrographs of alginate microspheres made with the W/O
em ulsion gelatin technique. ................................................ ............................... 103

4-12 Particle size distribution of uncoated and multiple-layer PEC particles..........................104

4-13 Stability studies of PEC m icroparticles. ............ ....................................... ................. 105

4-14 Optical photomicrographs of bilayer copper-crosslinked alginate particles incubated
2 4 h ou rs......... ....... .. ........... .............. ...............................................10 6

4-15 Optical photographs of uncoated zinc-crosslinked alginate particles..............................107

4-16 Optical micrographs of zinc-crosslinked PEC microparticles exposed to a shear rate
of 200-sec-1 for one m inute. ................................................... ..................................... 108

4-17 Optical micrographs of copper-crosslinked PEC microparticles exposed to a shear
rate of 200-sec 1 for one m minute. .......................................................................... ...... ..109

4-18 Optical photomicrograph of hexa-layer alginate-chitosan capsule during micropipette
a sp iratio n .................................................................................... ....... 1 1 1

4-19 Viscosity values of zinc- and copper-crosslinked PEC microparticles. .........................1112

4-20 Zeta potential values of A) zinc- and B) copper-crosslinked multiple layer PEC
particles. ......... .... ....... ........... ...................... 114

4-21 Concentration of copper ions released from chitosan-alginate PEC microparticles........ 116

4-22 Survival and recovery percentages of human dermal fibroblasts...............................117

5-1 Optical micrographs of blank zinc-alginate microparticles made with the W/O
em ulsion gelatin technique. ................................................ ............................... 134

5-2 Optical micrographs of blank copper-alginate microspheres made with the W/O
em ulsion gelatin technique. ................................................ ............................... 135

5-3 Optical micrographs ofF-BSA-loaded alginate microparticles coated with ten PEC
lay e rs. ........ ........ ........................................................................... 1 3 6

5-4 Scanning electron micrographs of zinc-crosslinked alginate microparticles made with
the W /O em ulsion gelation technique ......................................................... ............... 137









5-5 Scanning electron micrographs of copper-crosslinked alginate microparticles made
with the W /O emulsion gelation technique.............. ............................. .... ............. 138

5-6 Zeta potential values of A) zinc- and B) copper-crosslinked multiple layer PEC
particles encapsulating F-B SA ............ .................................................. ............... 139

5-7 Encapsulation efficiency of zinc- and copper- crosslinked alginate microparticles........141

5-8 In vitro release of F-BSA from uncoated and multiple PEC coated zinc- and copper-
alginate microspheres over a 48-hour period .... ........... ........ ........................ 142









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DEFORMABLE MICROPARTICLES WITH MULTIPLE FUNCTIONS FOR DRUG
DELIVERY AND DEVICE TESTING

By

Taili T. Thula

May 2007

Chair: Christopher Batich
Major: Biomedical Engineering

Since the HIV epidemic of the 1990s, researchers have attempted to develop a red blood

cell analog. Even though some of these substitutes are now in Phase III of clinical trials, their

use is limited by side effects and short half-life in the human body. As a result, there is still a

need for an effective erythrocyte analog with minimum immunogenic and side effects, so that it

can be used for multiple applications. Finding new approaches to develop more efficient blood

substitutes will not only bring valuable advances in the clinical approach, but also in the area of

in vitro testing of medical devices.

We examined the feasibility of creating a deformable multi-functional, biodegradable,

biocompatible particle for applications in drug delivery and device testing. As a preliminary

evaluation, we synthesized different types of microcapsules using natural and synthetic

polymers, various cross-linking agents, and diverse manufacturing techniques. After fully

characterizing of each system, we determined the most promising red blood cell analog in terms

of deformability, stability and toxicity. We also examined the encapsulation and release of

bovine serum albumin (BSA) within these deformable particles.

After removal of cross-linkers, zinc- and copper-alginate microparticles surrounded by

multiple polyelectrolyte layers of chitosan oligosaccharide and alginate were deformable and









remained stable under physiological pressures applied by the micropipette technique. In

addition, multiple coatings decreased toxicity of heavy-metal crosslinked particles. BSA

encapsulation and release from chitosan-alginate microspheres were contingent on the

crosslinker and number of polyelectrolyte coatings, respectively. Further theological studies are

needed to determine how closely these particles simulate the behavior of erythrocytes. Also,

studies on the encapsulation and release of different proteins, including hemoglobin, are needed

to establish the desired controlled release of bioactive agents for the proposed delivery system.









CHAPTER 1
INTRODUCTION

For the past century there has been an interest in blood substitutes. However, it was not

until the mid 1980s that efforts to find a commercially available product scaled up. Numerous

groups tried to develop a red blood cell analog due to the HIV epidemic of the time [1, 2].

Another reason that drove many researchers into the area was US Army concern about blood

supplies in the battlefield [1, 3]. Large investments were made by the military in search for the

ideal red blood cell analog. This "ideal blood analog" would carry oxygen, provide rapid

expansion of the blood space, eliminate infectious risks of transfusion, have universal

compatibility, and be available for immediate use without need for special storage [4].

However, about 10 years, all efforts and support were dropped since the developed analogs

did not meet the desired requirements [1]. Nowadays, some of those developed blood substitutes

are commercially available or in Phase III of clinical trials. Although many of them can fulfill

essential functions of transfused blood, providing expansion of the plasma volume and carrying

oxygen; their use is still limited due to side effects and short half-life time within the human

body [4].

The blood substitutes that have been studied more extensively can be grouped into two

classes: modified hemoglobin solutions and perfluorocarbon emulsions. The native human

hemoglobin molecule has a number of advantages as an oxygen carrier, including high capacity

for oxygen, the lack of antigens after purification, a prolonged shelf life, and the ability to

withstand harsh purification procedures [2, 4]. However, it needs to be modified in order to

decrease its oxygen affinity and to prevent rapid dissociation of the native a2-p2 tetramer into a-

p Oyperbo which are very toxic [4, 2]. Modifications of the hemoglobin molecule have included

inter- and intra-molecular cross-linking, conjugation to polymers, and more recently lipid or









polymeric encapsulation. Even though modification of hemoglobin prevents toxicity due to

breakdown of the molecule, other side effects have been observed with all modified hemoglobin

solutions. The main side effect, due to nitric oxide scavenging, is vasoconstriction resulting in

an increase in the mean arterial blood pressure and decrease in the cardiac index [4].

Perfluorocarbons have a molecular structure very similar to hydrocarbons', differing in the

replacement of hydrogen by fluorine atoms and their electron affinity [5]. Perfluorocarbons are

excellent carriers of oxygen and carbon dioxide; in addition, they can be mass-produced fairly

easily and source-independently. However, since they are not miscible in aqueous systems, they

have to be prepared as emulsions which leads to unwanted side effects due to surfactants [5, 6].

Complement activation is the major problem with perfluorocarbon-based products. Patients

present flu-like symptoms, including increased body temperatures and decreased platelet counts

[6, 5]. In addition to side effects, perfluorocarbon emulsions have shown poor efficacy and very

short half-life within the bloodstream [6].

New strategies to develop more efficient blood substitutes are necessary since all current

approaches lack effectiveness and patients normally depend on donated blood as the only source

for transfusion. Upcoming new potential blood-borne pathogens are always a concern regarding

blood transfusions; even though today that blood supply is much safer in developed countries

due to improved screenings [3, 5, 7]. Another ongoing problem is shortage of blood reserve due

to an annual increase of blood demand at a much faster pace than blood donations [5, 8]. In

addition to aging, worldwide population has been affected in the recent years by an increased

number of natural disasters, acts of terrorism, and civil and international conflicts, leading to a

steep increment in blood demand. Indeed, according to the American Association of Blood

Banks, a shortage of packed red blood cells is estimated by 2030 [3]. Other limitations to blood









collected for transfusion include: recipient immune suppression, short-term storage, and

decreased oxygenation capabilities [7].

The use of artificial erythrocytes could solve all the problems related to traditional

transfused blood. In addition to carrying oxygen, the new generation of erythrocyte analogs

could also be used as controlled-release delivery systems for the intravenous administration of

therapeutic agents. This new class of therapeutics should meet the following criteria: be non-

immunogenic, have long-term storage stability and an intravascular half life of weeks to months,

have no significant side effects, and be simple to manufacture and sterilize at permissible wide-

scale production costs.

Finding new approaches for the development of more efficient blood substitutes will not

only bring valuable advances in the clinical approach, but also in the area of in vitro testing of

medical devices and products. Certainly, many medical equipment and products have been

engineered with the purpose to interact directly with blood, e.g., dialysis machines, ventricular

assist devices, heart-lung machines, heart valves, catheters, grafts, stents, among others. Since

blood will be in direct contact and, sometimes, even subject to mechanical stress with these types

of devices and products, it is necessary to test their effect on blood in vitro before moving on to

clinical trials. However, in vitro testing with real blood carries a whole set of problems such as

cost, increment in rules and regulations, availability, variability among donors, and inconsistency

with storage time. As a result, particles with deformability properties and dispersed in fluids

with viscosities similar to blood will represent an invaluable innovation for the testing of devices

and products.

This study proposes to examine the feasibility of creating a multi-functional,

biodegradable, biocompatible particle with deformable properties for applications in in-vitro









testing of medical devices and products, and in drug delivery. The proposed method is based on

the hypothesis that: (1) polymeric microspheres with a highly porous inner structure or

microcapsules with a dissolvable gel core can be designed and synthesized so that they attain

appropriate deformability properties, and (2) that the ability of encapsulating and releasing

proteins within these microparticles, while still keeping some deformability capability, will allow

practical use. The aim of this dissertation is to conduct feasibility studies on the synthesis and

preliminary testing of different types of microcapsules using natural and synthetic polymers,

various cross-linking agents and solvents, and diverse manufacturing techniques. The aim will

also include discerning the most promising red blood cell analog, in terms of deformable

capability, to eventually be tested in an in vivo study. The deformability property of the

microcapsules will be based on the well-established micropipette aspiration technique.

In addition to finding microcapsules with correct deformability properties, another aim is

to develop a novel drug delivery system by using these deformable particles. The focus of this

second aim is on examining the delivery system feasibility: encapsulation, storage, and release of

proteins from microcapsules without affecting the protein bioactivity and the particle elastic

property. All components of the system will be designed to be non-toxic, and use materials

already approved for human contact. They will be combined in a novel way to test the

hypothesis through two specific aims. Each aim will have defined thresholds for measuring

acceptable outcomes, and if successful, will demonstrate the feasibility of developing a simple

technology for the manufacturing of a red blood cell analog for in vitro testing of medical

devices and products, and as a drug delivery system.









CHAPTER 2
BACKGROUND

Blood: Composition, Properties, and Functions

Composition

Blood is a specialized type of connective tissue composed of a suspension of living cells in

an aqueous solution of electrolytes and non-electrolytes called plasma [9, 10]. The cellular

portion of blood is composed of erythrocytes, leukocytes, and platelets. Erythrocytes normally

composed about 45 percent of the total blood volume; this percentage is known as hematocrit

[10]. In adults, the hematocrit ranges from 0.38 to 0.54: in females is slightly lower compared to

males, ranging between 0.38 0.46 and 0.42 0.54 respectively [11]. Conversely, not even one

percent of the total blood volume is constituted of leukocytes and platelets, while the remaining

volume (about 55%) is made up by plasma [10].

Blood plasma is a straw-colored fluid composed of about 90% (w) H20, 8% (w/v)

proteins, and 2% (w/v) inorganic and organic substances [9, 10]. Plasma makes approximately

4% of an individual's total body weight (40 to 45ml/Kg) [11]. The majority of proteins found in

plasma are produced in the liver and this makes about 7% of total plasma solutes (6.5 to 8g/dl)

[11]. The three main plasma proteins include: albumin, accounting for 60%; globulins; and

fibrinogen (Tables 2-1 and 2-2) [10, 11, 12]. Most of the blood functions are actually performed

by plasma proteins. The rest of plasma solutes include nutrients, gases, hormones, various cell

metabolic waste products, and ions [10]. Table 2-3 shows a list of all the electrolytes present in

plasma whose normal concentration is essential for nerve conduction, muscle contraction, blood

clotting, fluid balance and acid-base regulation [11, 12]. Other organic compounds found in

plasma are listed in Table 2-4 [12].









Properties

Although, the total amount of blood varies with age, sex, weight, and body build among

other factors; on average, blood accounts for approximately eight percent of the total body

weight. Also, the normal blood volume in healthy adult males is 5 6 liters whereas it is slightly

less for adult females (4 5 liters) [11]. The pH of blood ranges between 7.35 and 7.45 and its

temperature of 380C (100.40F) is a little higher than the body temperature [10]. Blood is denser

than water and has a viscosity approximately five times higher [10]. Although plasma is

considered a Newtonian viscous fluid, having a linear relationship between the shear stress and

strain rate; whole blood is considered non-Newtonian, showing both viscous and elastic

properties [9, 13, 14]. Thus, the non-Newtonian property of blood is acquired from the cellular

portion, and more specifically from the red blood cells or erythrocytes. The viscoelastic property

of blood is affected by blood compositional parameters, such as hematocrit and certain plasma

proteins. Other factors affecting blood viscoelasticity include changes in osmotic pressure, pH

and temperature; as well as administration of blood volume expanders and pathologies such as

myocardial infarction, peripheral vascular disease, cancer and diabetes [9, 15, 16, 17].

Functions

Blood performs several functions which can all be classified in three categories:

distribution, regulation, and protection. Blood transports oxygen and nutrients to all cells,

metabolic waste products to elimination sites, and hormones to target organs [10, 3, 11].

Diffusion and partial pressure are fundamental processes involved in the transport and

exchange of oxygen and carbon dioxide [11]. While the oxygen partial pressure (pO2) in the

alveoli is approximately 100 mm Hg, pulmonary capillaries have a pO2 of 40 mm Hg [10, 11];

resulting in a 60-mm-Hg diffusion gradient in favor of pulmonary capillaries. On the contrary,

systemic capillaries have a pO2 of 100 mm Hg as opposed to 40 mm Hg pO2 in tissues. In this









case, the 60-mm-Hg diffusion gradient is in favor of tissues, resulting in oxygen diffusion from

systemic capillaries to tissues.

In addition to distribution of substances, blood is in charge of maintaining appropriate

body temperatures, pH, osmotic pressure and fluid volume in the circulatory system [11, 10].

Blood disseminates heat throughout the body and deals with the excess by transporting it to the

skin surface. Regarding normal pH maintenance, many blood proteins act as buffers providing

an alkaline reserve of bicarbonate atoms [10]. Besides prevention of sudden pH changes, blood

proteins in conjunction with platelets also help stopping excessive fluid loss from the

bloodstream, maintaining an optimal fluid volume in the circulatory system. The protecting

action of blood includes prevention of blood loss and infections. Antibodies, complement

proteins, and white blood cells circulating in the blood help defend the body against foreign

aggressors [10].

Red Blood Cell: Morphology, Functions, Rheology, and Hematopoiesis

Morphology and Functions

Red blood cells (RBCs), or erythrocytes, are fully differentiated, anucleated, non-dividing

cells present in normal blood at high concentrations, with a hematocrit ratio (cell volume/blood

volume) of approximately 0.45 in large vessels, and 0.25 in small arterioles or venules [9, 10].

Matured erythrocytes are bound by plasma membrane and have essentially no organelles. They

are composed of approximately 97% hemoglobin, not including water, and they have other

proteins whose mainly function is to maintain the plasma membrane or support changes in the

RBC shape [11, 10]. The biconcave shape of erythrocytes is sustained by a net of fibrous

proteins (mainly spectrin) which is deformable, giving erythrocytes enough flexibility to change

shape as needed [10, 9].









The red cell's structural characteristics contribute to its extraordinary flexibility and

respiratory functions. Its fibrous-protein membrane is unstressed in the normal biconcave

configuration and large deformations of the cell can occur without stretching of the membrane

and without any change in the pressure differential between the interior and the exterior [10, 11,

9]. In addition to the RBC deformable membrane, its small size and biconcave disk shape, with

a diameter of approximately 7.5 [m and a thickness of 2 [m, provide a vast surface area ideal for

gas exchange [11, 9, 10]. Due to the lack of mitochondria, red blood cells generate ATP by an

anaerobic mechanism; as a result, they do not use any of the oxygen they transport, making them

very efficient oxygen carriers [10].

As mentioned earlier, most of the erythrocytes' content is composed of hemoglobin.

Hemoglobin (Hb) exists in a tetrameric form, consisting of two alpha (a) and two beta (3)

polypeptide chains each bound to a ringlike heme group [18, 11]. Each heme group carries an

atom of iron which can combine reversibly with one molecule of oxygen. Since a single red

blood cell contains about 250 million hemoglobin molecules, each cell can carry about 1 billion

molecules of oxygen [18].

Erythrocytes provide a protective environment for hemoglobin, preventing it from

breaking down into Oyperbo which adversely affects the kidneys, blood viscosity, and osmotic

pressure [18, 11]. In addition, erythrocytes protect the Hb molecule from undergoing an

unhindered oxidation process of its iron center, resulting in the transition of Hb from the ferrous

(HbFe2+) functional to the ferric nonfunctional form (HbFe3+) [19]. Along with the oxidative

process, damaging and toxic species can form, including the ferryl protein (HbFe4+) which can

peroxidize lipids, degrade carbohydrates, and cross-link proteins [19]. Indeed, the red blood cell









offers a reductase system rich in the enzymes catalase and superoxide dismutase which takes

care of the spontaneous oxidation of the ferrous iron [11, 19, 18, 1].

Rheological Properties

Erythrocytes are the major factor contributing to blood viscosity. G. B. Thurston was the

first to show that human blood exhibits elasticity and viscosity [20]. While blood viscoelasticity

depends on the elastic behavior of erythrocytes, blood rheology is governed by cell aggregation,

flow-induced cell organization and deformability [21]. One of the defining characteristic of

RBCs is that they form aggregates, known as rouleaux (Figure 2-1) [22], depending on the

presence of certain plasma proteins and on blood flow rates [9]. These aggregates are space

efficient since, at normal hematocrit levels, the available plasma space is extremely limited for

free cell motion without deformation. At very slow blood flows, shear rates on the cells are very

small and human blood becomes a big aggregate with the properties of a viscoelastic solid [9].

Although increasing blood flows will break up the rouleaux and reduce blood viscosity, further

rearrangement of erythrocytes is needed to optimize the plasma space for cell motion.

RBC deformation becomes important in reducing blood viscosity even further at shear

rates greater than 100 s-1 (Figure 2-2) [9, 23]. From a mechanical viewpoint, the red blood cell is

composed of an elastic membrane surrounding an incompressible Newtonian viscous fluid [24].

Since it lacks a nucleus and organelles, the intracellular matrix is considered as a protein rich,

low viscous solution which facilitates deformation [25]. As a result, the cell can undergo an

unlimited number of large deformations without changing its volume, surface area, and

stretching or tearing of the membrane [24].

Hematopoiesis

Hematopoiesis refers to the process of blood cell formation which occurs in the bone

marrow [10]. This process is specific to each type of blood cell and depends on the body needs









and different regulatory factors [10]. All of the cellular elements in blood arise from the

pleuripotent hematopoietic stem cell or hemocytoblast [11]. Once a cell is committed to a

specific blood cell type by the appearance of membrane surface receptors, its maturation

pathway is unique to the cell type [10, 11]. As cells mature, they migrate through the thin walls

of the sinusoids to enter the blood, resulting in an average of one ounce of new blood produced

daily [10].

Erythrocyte production, known as erythropoiesis, involves three distinct phases:

preparation for hemoglobin production, hemoglobin synthesis and accumulation, and nucleus

ejection [10]. During the first two phases, cells produce large amounts of ribosomes

(proerythroblast to erythroblast) and divide many times erythroblastt to normoblast) [10]. After

the cell reaches a hemoglobin concentration of approximately 35%, its nuclear function ends and

its nucleus degenerates and is ejected, causing the cell to collapse inward and adopt the

biconcave shape (reticulocyte) [10]. Reticulocytes still contain a small amount of clumped

ribosomes and rough endoplasmic reticulum which are degraded later by intracellular enzymes

[10]. The entire process from hemocytoblast to erythrocyte takes approximately five to seven

days [10, 11].

In healthy people, new erythrocytes are produced at a rapid rate of more that 2 million per

second [11]. Erythropoiesis is controlled hormonally and depends on appropriate supplies of

iron and certain B vitamins; however, its direct stimulus is provided by erythropoietin (EPO) [11,

10]. EPO stimulates stem cells in the bone marrow to produce red cells blood and the kidneys

play the major role in EPO synthesis [10, 11]. When there is a drop in normal blood oxygen

levels due to reduced number of red blood cells, decreased oxygen availability or increased

tissue demands for oxygen; kidneys accelerate their EPO release [11, 10]. Kidneys are so









important in the red blood cell formation process that a renal pathology reflects directly in the

erythrocyte production. To cite an example, dialysis patients do not produce enough

erythropoietin to support normal erythropoiesis, resulting in RBC counts less than half that of

healthy individuals [10].

Matured erythrocytes are unable to synthesize proteins, grow, and divide and they lose

their flexibility, becoming increasingly rigid and fragile as they age [10]. Red blood cells have a

useful life span of approximately 120 days. Mature erythrocytes swell and are engulfed and

destroyed by macrophages in the spleen. The heme of their hemoglobin is separated from the

globin and it is degraded to bilirubin while the globin is metabolized and broken down into

amino acids which are released back to the circulation [10]. At the same time, the iron core is

recycled, bound to protein (as ferritin or hemosiderin) and stored for reuse [10, 11].

Red Blood Cell as a Model for Drug Delivery

Erythrocytes are major candidates for drug delivery applications due to their abundance

and some unique characteristics such as long life-span in circulation, excellent biocompatibility

and biodegradability, and non-immunogenecity [26, 27]. As a result, they have been explored

extensively for two potential applications: (1) the sustained delivery of therapeutic agents in the

blood stream for a relative long term; and (2) the continuous and targeted delivery of drugs or

enzymes to organs of the reticulendothelial system (RES) [26, 28]. Certainly, using erythrocytes

as biological carriers offers an alternative to other carrier systems such as liposomes or

polymeric micro- and nano-particles that have been used for encapsulation of various drugs,

enzymes and peptides with therapeutic activity [29].

Encapsulation Methods

In general, the steps to prepare carrier erythrocytes include blood collection, erythrocyte

separation, drug encapsulation, resealing of the RBC carrier, and finally re-injection to the









organism [26]. Different techniques have been suggested to accomplish drug encapsulation

within erythrocytes along with a proper delivery. The osmotic methods, which are based on the

encapsulation under reduced osmotic pressure conditions, are the most widely used. Some of

these methods include hypotonic dilution, hypotonic pre-swelling, osmotic pulse, hypotonic

hemolysis and hypotonic dialysis [29, 26, 28]. Other techniques used for drug encapsulation

within erythrocytes consist of endocytosis and chemical and electrical processes [26, 29].

Advantages and Disadvantages of Red Blood Cells in Drug Delivery

As mentioned, erythrocytes present unique characteristics that made them a desirable

system for drug delivery. In addition to being a natural, biocompatible, biodegradable, and non-

immunogenic system; carrier RBCs offer the chance of loading a fairly high amount of drug in a

small volume, assuring dose sufficiency using a limited volume of erythrocyte samples [27, 30].

Other advantages in using erythrocytes as drug delivery systems include: their abundance, size,

morphology and inert intracellular environment; providing drug protection from endogenous

factors and cell metabolic activities, as well as the organism protection against toxic drug effects

[26, 27]. However, the two main advantages of carrier erythrocytes are that they act as a true

drug delivery system by modifying the drug's pharmacokinetic and pharmacodynamic

parameters, and their selective distribution to the RES organs [26, 28, 29]. The latter property of

the carrier RBCs is of great therapeutic importance in drugs such as antibiotics, enzymes or anti-

HIV peptides, among others [28, 26].

The clinical application of carrier erythrocytes has been limited by two main factors. The

lack of reliable and appropriate in-vitro storage methods for maintenance of cell survival and

drug content has become a major limiting factor [26]. Furthermore, autologous applications

might be limited depending on the disease state since the RBC morphology is directly affected

by certain pathologies. Although, the use of allocarriers could solve the problem, a whole set of









complications arises from this solution such as loss of the non-immunogenic property and

insufficient blood donors.

Drug Delivery Systems

The Food and Drug Administration (FDA) has approved a number of proteins, including

monoclonal antibodies, growth factors, cytokines, soluble receptors, and hormones, to treat a

variety of diseases [31, 32, 33]. However, conventional oral and intravenous (IV) delivery of

these drugs is usually not effective because of the inherent instability of many proteins [33, 31,

32, 34, 35]. Proteins have a very short in vivo half-life, are incapable of diffusing through

biological membranes and are unstable in the body environment [33, 36, 37]. Although

intravenous protein administration is most effective, daily injections and high protein

concentrations are required to achieve an effective local concentration for a prolonged time [33,

38]. Frequent systemic doses increase treatment cost, patient discomfort, and side effects.

To improve delivery of proteins, many controlled-release delivery systems composed of

polymeric biomaterials have been developed [39, 40, 41]. The main goal of developing these

systems is to control the release of drugs so that a therapeutic level is achieved for long periods

of time. Besides the therapeutic advantage, there is a business aspect for the great interest in

controlled-delivery systems. Due to increasing FDA regulations, pharmaceutical companies

need to invest more than $800 million for introducing a new drug in the market in addition to

spending more than 10 years of research and development work [42]. Therefore, creating new

devices or systems that deliver the same drug in a controlled manner is an economical strategy of

extending the license of the same drugs [42].

Controlled-release delivery systems are classified depending on the mechanism controlling

the drug release (Table 2-5) [42]. The most promising delivery approach is the encapsulation of

protein drugs within biodegradable polymers processed in a form that facilitates administration









through a syringe needle (particulate systems) [33, 41, 43]. Currently, three injectable polymer

configurations are used: nano- or microspheres, which are spherical matrix particles with the

drug uniformly distributed in the matrix; nano- or microcapsules which are conformed of a well-

defined core containing the therapeutic agent and a polymer membrane surrounding the core; and

cylindrical implants of approximately 0.8-1.5 mm in diameter [42, 43]. Microspheres and

microcapsules have several advantages over cylindrical implants, including less painful and a

more simplified administration [43].

Besides decreasing cost and frequency of injections, encapsulation of proteins and peptides

within biodegradable polymeric particulates has three key advantages over conventional drug

delivery systems [44]:

localization of the drug at the site of action,

continued and prolonged release of the therapeutic drug,

and protection of proteins and peptides against chemical or enzymatic degradation from

the physiological environment.

Microspheres and Microcapsules

Microencapsulation has been widely used not only to develop controlled-release drug

delivery systems; but to masquerade tastes and odors, reduce toxicity, and protect cells from the

host immune response in the absence of immunosuppression drugs [45, 46]. Among all the

microencapsulation systems, biodegradable polymeric nano- and micro-sized particulates are the

most promising ones for controlled delivery of different drugs, either hydrophobic or hydrophilic

ones. Microspheres are considered polymeric matrices with no superficial membrane. In this

system, the drug is relatively distributed homogeneously through the entire polymer matrix,

resulting in a release kinetic governed by erosion and diffusion [42]. Although microspheres

have been made with different kinds of polymers such as polyesters, polymethacrylates, and









celluloses; poly(lactide) or poly(lactide-co-glicolide)-based microspheres have been the most

studied systems due to the excellent biocompatibility and biodegradability properties of the

polymers [42, 47]. Table 2-6 shows a number of commercially available poly(lactide-co-

glycolide) (PLGA) microspheres used for drug delivery [42].

On the contrary, microcapsules present a well-defined core-shell structure where the core

is loaded with the therapeutic agent and the shell provides the pharmacokinetic-release limiting

factor [42, 48]. The release kinetics for this type of system is governed by diffusion from the

core through the degrading shell [48]. To date a diversity of materials has been employed as

shell components (i.e. synthetic and natural polyions, proteins, nucleic acids, lipids,

nanoparticles, etc) while biological cells, latex and inorganic particles, oil dispersion, and

organic crystals have been used as core templates [49, 48].

Microencapsulation Techniques

Several microencapsulation approaches with biodegradable polymers have been

developed, which are currently used in numerous applications in industry, agriculture, medicine,

pharmacy, and biotechnology [50]. Some of these methods include emulsion solvent

evaporation, solvent extraction, coacervation, spray-drying, interfacial complexation, coating,

and hot melt coating [46, 51]. Although, each method has both advantages and disadvantages in

the elaboration of polymeric microparticles, a common problem with all the approaches is the

preparation of truly efficient sustained-release delivery systems.

The most commonly used methods of preparing protein-loaded microspheres are the water-

in-oil-in-water (W/O/W) double emulsion and the oil-in-water (O/W) single emulsion solvent

evaporation techniques [47, 52]. Both methods involve the polymer dissolution in an organic

solution; mixing (shake, vortex, or sonication) of the drug, either in the powder or liquid form,









with the organic phase; and dispersion of the emulsion/solution into a continuous aqueous phase

followed by solvent evaporation or extraction. Although these techniques are fairly simple, cost

effective, and easily up-scalable for mass production; they lack effectiveness in producing a

linear controlled-release system. The main drawbacks of these methodologies include

denaturing of some encapsulated proteins due to the manufacturing process conditions (such as

intimate contact with organic solvent, heating, and mechanical forces), poor encapsulation

efficiency for hydrophilic drugs, and the polydispersed size of prepared microspheres generally

ranging from 10 to 100 [tm. Although some of these technological problems have been

addressed by the use of milder organic solvents and surfactants, the process still needs

improvement.

In addition to the emulsion solvent evaporation methods, the polymer phase separation and

spray-drying techniques have also being used for the preparation of protein-loaded microspheres.

The advantages of the polymer phase separation method are that no aqueous phase is involved,

eliminating the loss of protein through the aqueous phase as in the emulsion techniques, and that

the whole process takes place at room temperature, which avoids heat-induced denature of the

protein [47]. However, elevated concentrations of residual solvents have been found in the

microspheres [47], making them highly toxic and immunogenetic. Spray-drying has also being

used to encapsulate hydrophilic and hydrophobic drugs within several polymers [53, 54]. The

major advantages of this technique are the one-step process and ease of parameter control as well

as of scale-up [54].

On the other hand, the preparation of micro- and nano-sized capsules involves a wide

variety of manufacturing techniques. For this kind of system, fabrication of the core is not as

important as the shell preparation since the system's defining properties will depend on the latter.









While any of the techniques mentioned earlier could be applied to manufacture the core, the most

common approach used to fabricate the shell involves self-assembly strategies [55]. This

method, initially applied to planar surfaces, employs oppositely charged nanocomposite

multilayer films assembled onto templates of charged colloidal particles [55, 56].

The layer-by-layer (LbL) absorption technique, originally introduced by Decher [57], is the

method currently used to fabricate the core-shell particle system known as polyelectrolyte

complex (PEC). The LbL method is based on the electrostatic interaction between polyanions

and polycations that are consecutively absorbed on a charged planar or spherical surface [49, 58,

55]. The LbL method provides an effective and simple approach to manufacture PEC systems

with customized chemical and physical properties significantly different to those of the colloidal

template [51, 58]. In addition to customization of particle properties, the LbL approach permits

one to control the thickness of PEC layers with nanometer precision [55, 56, 48, 59, 7].

An important feature of PEC particles is the successive dissolution of the colloidal

template, resulting in free-standing, polyelectrolyte-shell capsules with the shape and size

determined by the template [7, 56, 58, 48]. The core removal is attained in a case-specific

manner since it depends on the template's chemical nature [51]. The major disadvantage of

polyelectrolyte capsules is their instability, depending on pH, temperature, and salt

concentrations [51, 60]. However, this disadvantage can be used to develop controlled-release

drug delivery capsules. In addition, capsules with prolonged release properties can be obtained

by increasing the number of polyelectrolyte layers [58, 48]. Another significant property of

these capsules is the selective permeability of their polyelectrolyte shell: permeable for small

molecules and ions and impermeable for higher molecular weight compounds [56, 58, 61].

Certainly, PEC capsules offer very attractive properties for the encapsulation of proteins,









peptides, oligonucleotides, genes, and cells for many applications in biotechnology, medicine,

pharmaceutics, cosmetics, and the food industry.

Immune System Response to Foreign Elements

The immune system is a complex network composed of many proteins, cells, and a few

well-defined organs [62]. Its main function is to protect us against pathogenic agents and

diseases by recognizing bacteria, fungi, viruses, parasites, cancerous cells, and foreign elements

[42, 62]. The immune system actually recognizes macromolecules (such as proteins or

polysaccharides) of foreign elements and the degree of immunogenecity of these structures will

depend on their foreignness, molecular size, chemical composition and complexity, and the

ability to be processed and presented with a major histological complex molecule [62].

Depending on the foreign agent's location in the body, different organs and cells will be involved

in the immune response.

The immune system response represents the major limiting factor in the efficacy of long-

term, controlled- release delivery particulates. Intravenously administered particulate carriers are

rapidly recognized by cells of the reticuloendothelial system (RES) and, consequently, removed

from the systemic circulation within minutes [62, 63, 64, 65]. The efficient elimination of

particulate systems by the RES is known for a long time. Depending on the surface chemistry,

charge and hydrophilicity, the removal process is initiated by opsonization. A set of plasma

proteins, known as opsonins, absorbs onto the surface of particulates and make them

recognizable to the RES [64]. Classical opsonin molecules include immunoglobulins,

complement proteins (such as Clq, C3b, and iC3b), apolipoproteins, von Willebrand factor,

thrombospondin, fibronectin, and mannose-binding protein [63, 64]. Opsonized particulates are

then phagocytosed by hepatic midzonal and periportal Kupffer cells [63, 66]. The spleen and

bone marrow might also participate in the particle clearance process from the bloodstream









depending on the pathophysiological conditions and the physicochemical characteristics of the

particulate carriers [66].

When particulate carriers are small and have a neutral-charged surface, they are not

efficiently opsonized and, therefore, they are poorly recognized by Kupffer cells [63]. However,

they might still go through a clearance process from the vasculature by fenestrations in the

hepatic sinusoidal endothelium, the spleen or bone marrow [63, 66]. Particles with diameters of

less than 100 nm get trapped by extrusion through endothelial fenestrations in the space of Disse

and the hepatic parenchyma [63]. The size, deformability, and sphericity of drug delivery

particulates also play a crucial role in their removal by the sinusoidal spleen [63]. Particles

larger than 200 nm and their aggregates can be physically trapped in the spleen fenestrations,

unless they are deformable as in the case of erythrocytes [63, 65]. On the other hand, the particle

elimination mechanism by the bone marrow is more complex and species-dependent, capable of

removing particles from the circulation by transcellular and intercellular paths [66].

Certainly, physicochemical factors of drug delivery particulates are critical for their

recognition and removal from the bloodstream by the RES. For the past 30 years, it has been

known that hydrophilic particles remained in the circulation longer than hydrophobic ones due to

rapid opsonization of the latter [67]. The particle's surface charge is another factor influencing

the clearance process due to electrostatic interactions with blood components and cell surfaces.

Although, there are conflicting viewpoints regarding the surface charge, it is believed that neutral

charged particulates have a longer half-life in the bloodstream [67, 64]. The other major factor

determining the RES removal of particulate systems is the size of the particles. A narrow

particle diameter range, between 100 and 200 nm, is preferred to avoid particle entrapment in

hepatic and splenic fenestrations [66]; unless, the particulates are deformable in which case









larger diameter particles could be administered, increasing the drug loading potential. Indeed, a

fuller understanding of the physicochemical properties of drug delivery particulates and their

effects on the immune response will make it possible to design particle carriers with reduced

affinity to the cells of the RES.










Table 2-1. Plasma protein concentrations (mg/100 mL)
Protein Plasma

Total 6500-8000

Albumin 4000-4800

ai-globulins 380-870

a2-globulins 570-940

P-globulins 730-1380

y-globulins 590-1450

Fibrinogen 200-400










Table 2-2. Phvsicochemical properties of the maior plasma proteins


Concentration
(mg/ml)


Mol
weight
(Da)


pI Sedimentation
const in water
at 200C (10-3
cm/dyn*s)


Diffusion
coeff in
water at
200C (10-7
cm2/s)


Partial
specific
volume of
protein at
200C (ml/g)


Prealbumin

Albumin

Al-seromucoid

Al-antitrypsin

A2-macroglobulin

A2-haptoglobin

Type 1.1

Type 2.1

Type 2.2

A2-Ceruloplasmin

Transferrin

Lipoproteins

LDL (p<1.019)

LDL (p<1.063)

HDL2 (p=1.093)

HDL3 (p=1.149)

IgA (monomer)

IgG

IgM

Clq

C3

C4

Fibrinogen


10-40

35-45

0.5-1.5

2.0-4.0

1.5-4.5



1.0-2.2

1.6-3.0

1.2-2.6

0.15-0.60

2.0-3.2


1.5-2.3

2.8-4.4

.37-1.17

2.17-2.70

1.4-4.2

6-17

0.5-1.9

0.1-0.25

1.5-1.7

0.2-0.5

2.0-4.0


5.5x104

6.6x104

4.4x104

5.4x104

7.2x105



1.0x105

2.0x105

4.0x105

1.6x105

7.6x104



1.5x107

3.2x106

4.4x105

2.0x105

1.6x105

1.5x105

9.5x105

4.0x105

1.8x105

2.1x105

3.4x105


4.2

4.6

3.1

3.5

19.6



4.4

4.3-6.5

7.5

7.08

5.5


--- >12

--- 0-12

--- 4-8

--- 2-4

7

6.8 6.5-7.0

--- 18-20

11.1

6.4 9.55

--- 10.1

5.5 7.6


Species


0.733

0.675

0.646

0.735



0.766





0.713

0.758


3.76

5.0


0.725

0.739

0.724


1.97


0.723









Table 2-3. Concentrations of major electrolytes (mEq/L) in whole blood and plasma
Electrolyte Whole blood Plasma

Bicarbonate 19-23 24-30

Calcium 4.8 4.0-5.5

Chloride 77-86 100-110

Magnesium 3.0-3.8 1.6-2.2

Phosphate 0.76-1.1 1.6-2.7

Potassium 40-60 4.0-5.6

Sodium 79-91 130-155

Sulfate 0.1-0.2 0.7-1.5


Table 2-4. Concentrations of various organic compounds (mg/100mL) in whole blood and
plasma
Species Whole blood Plasma

Amino acids 38-53 35-65

Bilirubin 0.2-1.4 0.2-1.4

Cholesterol 115-225 120-200

Creatine 2.9-4.9 2.5-3.0

Creatinine 1-2 0.6-1.2

Fat, neutral 85-235 25-260

Fatty acids 250-390 150-500

Glucose 80-100 60-130

Lipids, total 445-610 285-675

Nonprotein N 25-50 19-30

Phospholipids 225-285 150-250

Urea 20-40 20-30

Uric acid 0.6-4.9 2.0-6.0

Water 81-86g 93-95g































Figure 2-1. Scanning electron micrograph of red blood cell aggregates, rouleaux.


10 -

(L




10 -2


S10
(A
LU

" 10 -



o 10
lo
l>

S10


10 1 10 2
Shear Rate (1lsec)


Figure 2-2. Shear rate dependence of normal human blood viscoelasticity at 2 Hz and 22 C.









Table 2-5. Controlled release drug delivery systems
* Diffusion-controlled
Reservoir and monolithic systems
Water penetration-controlled
Osmotic and swelling-controlled systems
Chemically-controlled
Biodegradable reservoir and monolithic systems
Biodegradable polymer backbones with pendant drugs
Responsive
Physically- and chemically-responsive systems (T, solvents, pH, ions)
Mechanical, magnetic- or ultrasound-responsive systems
Biochemically-responsive; self-regulated systems
Particulate
Microparticulates
Polymer-drug conjugates
Polymeric micelle systems
Liposome systems



Table 2-6. Examples of commercialized PLGA copolymers microspheres
Trade Name Drug
Decapeptyl Depot Triptorelin
Enantone LP Leuprorelin
Somatulin LP Lanreotide
Parlodel LAR Bromocriptine
Sandostatin-LAR Ocreotide
Nutropin Recombinant Human Growth Factor
Lupron Leuprolide Acetate









CHAPTER 3
CHARACTERIZATION OF THE DEFORMABLE BEHAVIOR OF PLURONIC-PLA AND
POLYELECTROLYTE CHITOSAN-ALGINATE MICROPARTICLES

Introduction

Since the1980s numerous groups have tried to develop a red blood cell analog due to the

HIV epidemic of the time [1, 2]. Even though some of these substitutes are now in phase III of

clinical trials, their use is very limited due to side effects and short half-life time within the

human body [4]. As a result, there is still a need for an effective erythrocyte analog with

minimum immunogenic and side effects, so that it can be used for multiple applications. Besides

the imperative need of a blood substitute for in vivo use, there is also a need of it for in-vitro

testing of medical devices and products.

Many types of medical equipment and products have been engineered with the purpose to

interact directly with blood. To illustrate, some of the these devices include dialysis machines,

ventricular assist devices, heart-lung machines, heart valves, catheters, grafts, stents, among

others. Since blood will be in direct contact and, sometimes, even subject to mechanical stress

with this type of devices, it is necessary to test their effect on blood in vitro before moving on to

clinical trials. However, testing with real blood involves an array of complications such as cost,

increment in rules and regulations, availability, variability among donors, and inconsistency with

storage time.

Particles with deformability properties and dispersed in fluids with viscosities similar to

blood will represent a valuable advance for the testing of medical devices and products. In this

study we investigated the use of synthetic and natural biodegradable polymers as possible

materials for the development of biodegradable, biocompatible, and multi-functional particles

with deformability properties for applications in device testing and drug delivery. The main goal









is to attain particles capable to deform under micropipette suction pressures used to aspirate

erythrocytes.

Poly(Lactic Acid)

In the development of medical devices, especially for controlled-release delivery systems,

synthetic biodegradable polymers are frequently used as carriers for protein drugs. Synthetic

polymers are preferred over biological materials because of their biocompatibility, minimal

immunogenecity, biodegradability, and high manufactured reproducibility. The polymers most

often used for the fabrication of drug delivery systems are poly(glycolic acid) (PGA), poly(lactic

acid) (PLA), and the copolymer poly(lactide-co-glycolide) (PLGA) [33, 68] due to their good

biocompatibility, variable mechanical processability, and a wide range of biodegradable

properties [33, 41]. Among these three polyesters, PLA (Figure 3-1) with the chemical formula

(C3H402)n is the most hydrolytically stable in addition to its long history of safe and

biodegradable use as resorbable suture materials [69, 43, 70, 71].

PLA degrades by a well-known erosion process into natural lactic-acid metabolites that are

easily eliminated by the body. Hydrolysis of the ester bond initiates the erosion process;

followed by a reduction in molecular weight and an increase in the acidic environment which, in

turns, accelerates degradation [42]. The PLA erosion rate is controlled by varying the molecular

weight and type of polylactide monomer used [71, 68, 72, 42]. These factors determine the

hydrophilicity and crystallinity, which govern the rate of water penetration [68, 44]. The higher

the molecular weight, the longer the polymer retains its structural integrity, the slower its

degradation rate [68]. In this study, DL-PLA was the polymer used since it is completely

amorphous and its lack of crystallinity causes it to degrade faster than L-PLA. In addition, it has

a lower tensile strength which makes an attractive property when trying to manufacture

deformable microspheres.









Pluronics

Surfactants are an important constituent for many colloidal suspensions, governing

theological properties and stability against phase separation of many commercial dispersions in

the cosmetic, food, and medical field [73]. Poloxamers, also known by the commercial name

Pluronics (BASF, Wyandote, USA), are commonly used as stabilizers in the preparation of

emulsions and colloids. Pluronics are triblock ABA-type copolymers composed of

poly(ethylene oxide)-poly(propyleneoxide)-poly(ethylene oxide) (PEO-PPO-PEO) blocks

arranged in a basic structure. Table 3-1 shows a list of the commercially available Pluronics

which can be classified as hydrophobic or hydrophilic depending on the EO/PO ratio [74].

Pluronics are soluble in water and polar solvents. In aqueous solutions, they exhibit

temperature-dependent theological properties due to their amphiphilic structures [74]. Gelation

of Pluronics solutions will occur above a certain temperature and it will also depend on the

polymer concentration and EO/PO ratio [74]. Moreover, Pluronics in aqueous solutions are

capable of self-assembling into multi-molecular aggregates micelless), which makes them

attractive for drug delivery applications [75]. Micelle formation will also depend on block

copolymer concentrations, EO and PO unit lengths as well as temperature and type of solvent

used [76, 77].

In addition to unique gelation and micelle formation properties, some of the Pluronics

exhibit minimal toxicities in vivo, allowing their clinical use [78]. Furthermore, modification of

polymeric surfaces with PEO to reduce non-specific protein adsorption and undesired

bioadhesion in biological environments has been proven effective for the past 30 years [64, 36].

The valuable properties of PEO surface modifications are associated to the distinctive structure

of PEO molecules which show greatly hydrated, non-ionic, and mobile chains in an aqueous

environment [64]. However, using Pluronic copolymers, instead of the PEO homopolymer, for









surface modification is preferred due to the improved stabilization of the system, provided by

anchoring of the PPO units to the solid surface [36, 64]. In this study, conjugations of DL-PLA

and Pluronics with different compositions and molecular weights were used to investigate the

system's feasibility in the development of deformable particles.

Alginate

Alginates are random, anionic, linear, polysaccharides derived from brown algae. These

natural block copolymers are conformed by varying ratios of unbranched chain 1,4'-linked P-D-

mannuronic acid (M) and a-L-guluronic acid (G) residues (Figure 3-2) [79, 80, 81, 82].

Frequency and distribution of these monomers along the polymer chain are irregular and

dependent on the source of origin [83]. Although alginates have been isolated from bacteria,

three species of brown algae (Laminaria Oyperborean, Ascophyllum nodosum, and Macrocystis

pyrifera) are the primary source for commercially available alginates [84]. Approximately 40%

of the algae dry weight is alginate which is found in the intracellular matrix as a mixed salt of

diverse cations from the sea water (i.e. Mg2+, Ca2+, Sr2+, Ba2+, and Na+) [84].

Though alginates have several unique properties, gelation and swellability are the two most

significant. Alginate forms a gel in the presence of divalent cations and calcium-crosslinked gels

are the most abundant in nature [84]. The affinity of alginates for divalent cations depends

mostly on the electronic structure of the cation, with the highest affinity for copper ions and the

least for manganese ions [85]. Metal ions, especially calcium, are more commonly used to

crosslink gel alginates since toxicity has been a limitant factor in the use of transition metals.

The sol-gel transformation begins with the exchange of monovalent ions from the G residues of

water-soluble alginate salts with divalent cations [80, 86]. Binding of divalent cations to G

blocks is highly cooperative, forming stacks of more than 20 monomers [84]. Martinsen et al.

described the physical orientation of the stacks as the egg-box structure, being more abundant for









high G alginates than for low G ones [83]. However, Wang et al. found that transition metal ions

bind to both G and M residues with a binding distribution dependent on solvent conditions [85].

As a result, alginate composition as well as cation affinity will play a significant role in the

properties of the crosslinked gel matrix.

The swelling property of crosslinked alginate gels depends on solutes present in the

solution [85]. Sequestering agents form stable complexes with bound cations, resulting in a

polymer chain relaxation and volume expansion [61]. Calcium-alginate gel matrices start

swelling and are further destabilized upon contact with solutions containing chelators such as

phosphate, lactate, citrate, ethylenediamine tetraacid (EDTA) or high concentrations of non-

gelling cations like sodium or magnesium ions [61, 84]. Swelling and degradation of alginate

gels due to removal of crosslinker ions, time, pH, or temperature are determining features in the

development of controlled-release systems. As a result, alginate matrices have been commonly

used for a variety of delivery systems including gels, films, beads, microparticles, and sponges.

In addition to alginates' unique properties, their chemistry and relatively mild crosslinking

conditions, fairly easy processability, source abundant, low price, biocompatibility, and

biodegradability have enabled their used in a wide variety of biomedical applications [80, 81,

82]. Some of these applications include immunoprotective containers in cell transplantation, cell

scaffolds, controlled-release drug delivery systems, and surgical dressings for the treatment of

adhesion problems in tissue repair, capillary hemorrhage blockage, and burns [45, 61, 83, 84].

Certainly, a wide range of matrices with different morphologies, pore size, water content, and

release rates can be manufactured by selecting determined alginate compositions, crosslinkers,

additives, gelation conditions, and coating agents.









Chitosan

Chitosan is a cationic, linear, naturally occurring polysaccharide with structural

characteristics similar to glycosaminoglycans [87, 88]. This polycationic biopolymer is

fundamentally composed of (1-4)-linked D-glucosamine units with some percentage of N-acetyl-

D-glucosamine units (Figure 3-3) [50]. Chitosan is commonly obtained by alkaline

deacetylation of chitin since it is rarely found in nature [89]. Conversely, chitin is the second

most abundant natural polysaccharide [90], present in the exoskeleton of crustaceans, mollusks,

the cell walls of fungi, and the cuticle of insects [87]. Chitin is a homopolymer consisting of

P(1-4)-linked N-acetyl-D-glucosamine units, which can be specifically modified by controlled

chemical reactions [50].

The degree of deacetylation of chitin will determine the functional properties of chitosan.

A low-molecular-weight, highly-deacetylated chitosan, known as chitosan oligosaccharide

(COS), is obtained when chitosan is further hydrolyzed [89]. COS has very promising properties

which can be utilized in a wide range of applications. While chitosan is only biologically active

in acidic environments due to poor solubility, COS is water soluble at neutral pH as it maintains

its cationic nature. This property allows COS to electrostatically interact with polyanionic

polymers and molecules in diverse aqueous environments, forming polyelectrolyte materials

optimal for drug delivery [83, 91].

Besides the physicochemical features, many useful biological properties of chitosan have

been recognized, including biocompatibility; biodegradability; low toxicity; mucoadhesion; and

antifungal, antimicrobial, anticoagulant, antitumoral and hipolipidemic activity [50, 87, 92].

Chitosan is metabolized and degraded into non-toxic products by enzymes, such as lysozyme,

lipase and chitosanase [50]. While the latter is found in plants and insects, the other two are

present in mammals [50]. All these interesting characteristics have led to the recognition of









chitosan and its derivatives as potential materials in numerous applications in agriculture,

environment, food industry, medicine, pharmacy, and biotechnology. Some of the applications

in the medicine and pharmaceutical fields include: surgical sutures, sponges and bandages,

matrices and coatings for drug delivery systems, orthopedic and dentistry materials, cell

scaffolds, and immunoprotective barriers for cell transplantation [50, 88].

To attain microcapsules with deformable properties, we studied different concentrations of

alginate and chitosan for the development a polyelectrolyte complex (PEC) system consisting of

multiple-layer microcapsules with a dissolvable gel core.

Materials and Methods

Preparation of PLA-Pluronic Particles

Polyester microspheres were prepared by the water-in-oil-in-water (W/O/W) double-

emulsion solvent extraction/evaporation technique. The polymer used was D,L-poly lactic acid

(DL-PLA) (Mw 350 kDa) from Birmingham Polymers. In addition to PLA, Pluronics with

different HLB (kindly donated by BASF) were used as surfactants (Table 3-2). Pluronic was

dissolved at different concentrations in phosphate buffer saline (PBS), pH 7.4, and emulsified in

methylene chloride containing 2% (w/w) DL-PLA. This first water-in-oil emulsion was

generated by ultrasonication for 60 seconds in an ice water bath. This emulsion was added to

1.5% polyvinyl alcohol (PVA) solution at 40C under continuous stirring at 1500 rpm for 30 min.

Microspheres were collected by filtration, through a 300-tpm nylon mesh, and centrifugation.

PLA-Pluronic microparticles were washed three times with deionized water, lyophilized, and

stored at -200C until use.

Preparation of Alginate Particles

Natural polymeric microspheres were prepared by a water-in-oil emulsion crosslinking-

gelation technique [93]. The polymer used was alginate from Keltone (LV, food grade).









Calcium chloride and copper nitrate at different concentrations were used as crosslinking

solutions (Table 3-3). The aqueous phase consisted of sodium alginate dissolved in deionized

water at different concentrations. The oil phase consisted of soybean oil. The first emulsion was

obtained by ultrasonication of two phases at 60W for 1 min in an ice bath. A second aqueous

solution containing different concentrations of crosslinking agent was added to the emulsion by

air-spray (40 psi, 20ml/hr) at a 4-cm dropping distance while stirring the whole medium slowly

with a magnetic stirrer. Particles were allowed to cure for ten minutes under continuous stirring.

Then medium was allowed to rest for 24 hr so that particles would drop to the bottom of the

container while the oil phase was left at the top. After separation of the two phases, particles

were collected by filtration through a 45-[tm mesh and washed copiously to remove the oil.

Particles were stored at 40C until future use.

Coating of Alginate Particles

Ca- and Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption

technique. The coating always started and ended with the positive polymer. Microcapsules with

zero, one, five, and nine alternating layers were fabricated. The cationic polymer used was a

high-molecular-weight chitosan and a high-deacetylated, low-molecular-weight chitosan

oligosaccharide lactate (Mw < 5000, 90% deacetylation) from Aldrich. Alginate was the anionic

polymer used. A chitosan solution containing CaC12 or alginate solution was added to

crosslinked-alginate particles dispersed in deionized water. Different batches were made

differing in the concentrations of chitosan and alginate solutions.

After allowing the particles to coat for 30 minutes, they were collected by centrifugation

and washed three times with deionized water to ensure that all free polyelectrolytes were

removed. Following the washes, particles were ready for the next coating or stored at 40C until









use. Table 3-3 shows all the formulations used to make the different batches of chitosan-alginate

microcapsules.

Decomposition of the Particles' Core

Decomposition of the particles' core was achieved in a case-specific manner by treating

particles with sequestrant agents specific to each crosslinking ion. The agents used to remove

calcium and copper ions were phosphate (Cellgro) and bovine serum albumin (Cellgro)

respectively. An initial concentration of 5% (w/v) chelating media with or without supplemental

ions was prepared. Multiple-layer particles, at a 5% (w/v) concentration, were incubated at room

temperature in the chelating medium specific to each crosslinking ion. After close observation,

old medium was replaced by same amount of fresh medium if core was not dissolved. Medium

was replaced until the core was decomposed. Gel-core particles were then stored at 40C until

further use.

Measurement of Particles' Deformability

The micropipette aspiration technique was applied to assess particles' deformability. In

this technique, a portion of a single microcapsule is drawn into a pipette by applying a pressure

difference between the inside of the pipette and the chamber containing the particle suspension

[94, 95]. The Plexiglas, home-built chamber open at one side, allowing for the pipette entry, was

prepared by fixing with vacuum grease a glass coverslip to the top and the bottom of the

chamber. Micropipettes were pulled from glass capillaries of 1 mm diameter using a pipette

puller. To attain the desired diameters, micropipettes were forged subsequently. Micropipette

tips with inside diameters approximately 20% smaller than the particles' diameter were

fabricated. Previous to their use, pipettes were backfilled with 0.9% NaCl solution using a

plastic syringe with a 97-mm long, 28G backfiller (Microfil, World Precision Instruments,

Sarasota, FL).









Before particles were used for deformability studies, they were incubated in sequestrant-

containing media until the core was decomposed. The prepared chamber was then filled with

microparticles suspended in sequestering media and mounted on the microscope stage. An initial

negative pressure of 5 cm H20 (approximately -4.9x103 dynes/cm2) was applied and pressure

was augmented in increments of five. After a section of the particle had been aspirated, the

negative pressure was decreased and the particle was unloaded. The whole aspiration process

was monitored and recorded using an inverted light microscope (Axiovert 100, Zeiss), a Carl

Zeiss video camera system, and a VCR. The recorded images were converted into digital images

using the Matrox software. Since the main goal of the study was to attain particles capable to

deform under micropipette suction pressures used to aspirate erythrocytes [96], particles were

considered deformable only when using pressures up to -40x103 dynes/cm2 (-41 cm H20).

Scanning Electron Microscopy (SEM) Analysis

Surface morphology of the microcapsules was examined by SEM after gold-palladium

coating of microcapsule samples on an aluminum stub. Samples for scanning electron

micrographs were obtained after 0, 1, 5, and 9 coatings. Droplets of microsphere solutions were

mounted on aluminum stubs, let air dried and sputter-coated with gold and palladium particles.

The stubs were mounted in a scanning electron microscope at 10.0 kV and imaged at x500,

xl000, and x5000.

Results and Discussion

PLA-Pluronic Particles

Microspheres made with PLA and Pluronics were not deformable when using the

micropipette technique, except for particles made with PLA and P105. Pluronics used had

different molecular weights and HLB ratios ranging from the most hydrophilic to the most

hydrophobic copolymer. The first type of Pluronic used was F127 which is a hydrophilic









copolymer very commonly used for drug delivery systems. F127 has a high PPO molecular

weight (4000 Da) and a PEO content of approximately 70 % (w). Pluronics were dissolved in

phosphate buffered saline (PBS, pH 7.4) at different concentrations ranging from 1% to 20%

(w/v). At concentrations below 20%, Pluronic solutions did not gel at room temperature. In

addition, particles obtained at lower concentrations were not uniform; therefore, 20% was the

Pluronic solution concentration used for all further experiments.

Besides variation of Pluronic solution concentration, particles with formulations of

different PLA/Pluronic ratios were synthesized. The ratios used were the following: 1:1, 1:2,

and 1:3. Microspheres were not obtained with the 1:1 formulation; thus we only continued using

the last two formulations for the preparation of following batches. Even though, F 127-PLA

microspheres presented a morphology with a thick shell and an empty core (Figure 3-4A), none

of the batches obtained resulted in deformable microspheres. Negative aspiration pressures of up

to 60 cm H20 (-58.8 x103 dynes/cm2) were applied to the particles and no signs of deformation

were observed. As a result, we decided to use L101 which is another high-PPO molecular

weight Pluronic (3300 Da), but in the hydrophobic range with a 10 % (w) PEO content. All

the L101-PLA formulations yielded microspheres with morphology similar to the F127-PLA

particles' (Figure 3-4B) and deformability was not achieved either.

Since the idea of using Pluronics was to increase the water content in the PLA

microsphere system so that particles become softer, we decided to use a surfactant with higher

hydrophile content and lower lipohilic content. The next Pluronic used was F68 containing 80

% (w) PEO and a very low PPO molecular weight (1800 Da). F68-PLA microspheres presented

a different morphology compared to the F127- and L101-PLA particles'. F68-PLA particles

presented a capsular morphology with a large PLA core and a very thin Pluronic shell (Figure









3-4C). After applying an aspiration pressure of -45 cm H20 (-44.1 x103 dynes/cm2), it was

found that the thin shell was deformable but not the core. Pressures of up to -60 cm H20 were

applied with no change in the deformable capability of the particles. All the F68-PLA

formulations showed similar morphology and deformability properties; therefore, we decided to

keep testing other Pluronics.

P105 with a 3300-Da PPO molecular weight and approximately 50% hydrophile content

was used next. P105-PLA microspheres presented a defined core-shell structure (Figure 3-5).

The PLA core was very porous as well as the P105 shell (Figure 3-6). The difference in

morphology seen in F68- and P105-PLA systems is attributed to the molecular weight of the

PPO parts. Previous studies have shown that a high molecular weight of the PPO content is

necessary to provide enough anchoring of the copolymer to the particle surface [36, 64].

Particles made with F68 displayed a very thin shell due to poor attachment of the surfactant to

the PLA core.

In addition to a different surface morphology, P105-PLA particles displayed a higher

deformable capability compared to the particles made with PLA and all the previous Pluronics.

During micropipette experiments, it was clearly seen that the shell was very deformable although

the core was not. A portion of the shell was drawn into the pipette forming a "tongue" when

applying a negative aspiration pressure of 25 cm H20 (-24.5 x103 dynes/cm2). However, no

additional change in deformability of the particle was observed even after increasing the pressure

to -60 cm H20. All batches made with P105 were very consistent.

Pluronics P105, P103, P104, and F108 were used next; all with a fixed propylene oxide

content (3300 Da) and variations of the ethylene oxide content: 50, 30, 40, and 80% respectively.

It was hypothesized that the difference in ethylene oxide content would have an influence on the









shell thickness; longer PEO units would yield particles with thicker shells. The goal was to

attain particles with a very thick, deformable shell and a small core. However, only P105-PLA

microspheres presented the core-shell structure, indicating that a 50 % PEO content is ideal for

the synthesis of capsular systems. Since we were not able to modify the shell and core size, we

decided to focus on other materials. Table 3-4 summarizes the deformability properties of

Pluronic-PLA particles.

Chitosan-Alginate Particles

Crosslinked alginate particles were made using the water-in-oil emulsion crosslinking

gelation technique. No particles were obtained when using a 0.5% (w/v) sodium alginate

solution. The best morphology was obtained when using a concentration of sodium alginate

solution at 3% (w/v) and a CaCl2 concentration of 5% (w/v). To coat the particles, different

concentrations of high-molecular-weight chitosan and chitosan oligosaccharide (COS) solutions

containing 0.1% (w/v) CaCl2 were used. Crosslinked-alginate particles coated with high

molecular weight chitosan did not yield good results; therefore, only chitosan oligosaccharide

was used for all further coatings. Also, different concentrations of alginate solutions were used

for the polyanionic coating; although only 0.1% (w/v) solution seemed to work. We were unable

to collect the particles when higher alginate concentrations were used.

Figure 3-7 shows optical and scanning electron micrographs of polyelectrolyte multiple

layer Ca-crosslinked alginate particles. The core-shell structure of the particles can be seen from

the SEM pictures. The evident invagination in the center of the particles (Figure 3-7B) is due to

loss of water from the gelatinous Ca-alginate core. After water removal, alginate chains

collapsed as opposed to P105-PLA microcapsules which kept their spherical shape (Figure 3-6).

Uncoated and coated calcium-alginate particles were not deformable when using the









micropipette technique before removal of crosslinking ions. Furthermore, uncoated, mono-layer

and multiple-layer coated particles were not stable after removing the calcium ions.

Alginate particles crosslinked with copper ions presented smaller diameters than Ca-

crosslinked particles (Figure 3-8). This could be explained by a faster crosslinking reaction,

resulting in less time for destabilization of the emulsion to take place and less coalition of small

particles into large ones. After water was removed from the Cu-alginate core, capsules collapsed

similar to Ca-alginate particles (Figure 3-8B).

Coated Cu-alginate particles were not deformable when using the micropipette as well.

However, when copper ions were removed, particles presented deformable properties. Particles

deformed and were completely aspirated into the micropipette tip after applying a 20-cm-H20

negative pressure (-19.6x103 dynes/cm2). While uncoated Cu-alginate particles were not stable

after removal of crosslinking ions; mono-layer particles were stable mostly in static conditions.

Some of the mono-layer particles ruptured under pressures of -20 cm H20. Best results were

obtained with Cu-alginate particles coated five times. Penta-layer particles were able to deform

under a 20-cm-H20 negative aspiration pressure several times while recovering their shape.

Removal of crosslinking ions was not as efficient for the nine-layer particles as for the lower-

layer ones under the conditions tested. As a result, only portions of the particles were

deformable under a 20-cm-H20 negative pressure. Particles were completely aspirated into the

micropipette only after applying negative pressures of 30-cm-H20 (-29.4 x103 dynes/cm2).

Table 3-5 summarizes the deformability properties of PEC chitosan-alginate capsules.

The ion exchange properties of alginate gels have been studied by many research groups.

This ion exchange process depends on several factors such as the amount of crosslinking ions

forming the gel, physical properties of the gel, competing ion concentration, pH and ionic









strength of the solution [97, 98, 99]. In addition, the order affinity of alginate for divalent ions

has been found to be Pb2+>Cu2+>Cd2+>Ba2+>Ni2+>Ca2+>Zn2+>C2+>Mn2+>Sr2 [85, 100, 101,

102]. After removal of calcium ions from the particles by phosphate ions present in the cell

culture medium, the polyelectrolyte complex system was not stable anymore. The calcium PEC

system collapsed regardless of the number of PEC layers because alginate had no affinity for the

other electrolytes present in the medium. On the contrary, when copper ions were chelated by

albumin, the alginate binding sites left were occupied by calcium ions present in the cell culture

medium. As a result, the ion exchange process along with the PEC layers kept the pre-

crosslinked Cu-alginate particles stable.

Conclusion

Only P105-PLA microspheres presented a well-defined core-shell structure with a highly

porous PLA core and a P105 shell. In addition to a different surface morphology, P105-PLA

particles displayed higher deformable capability compared to the particles made with PLA and

all the other Pluronics. During micropipette experiments, it was clearly seen that the shell

was very deformable although the core was not. The differences in morphology and

deformability properties are associated to the copolymer composition, indicating that a 50 %

PEO content as well as a high PPO molecular weight are ideal for the synthesis of capsular

systems.

Regarding PEC chitosan-alginate systems, only multiple-layer coated copper-alginate

particles presented deformability properties under micropipette aspiration after removal of

crosslinking ions. An ion exchange process along with the PEC layers kept the pre-crosslinked

Cu-alginate particles stable. Instability of Ca-alginate particles, subsequent to chelation of

calcium ions, was independent of the number of PEC layers. Particle strength after ion removal

can be enhanced by multiple coatings with low molecular weight (high percent deacetylation)









chitosan oligosaccharide. Further studies need to be conducted to compare the effects of various

divalent cations on the theological properties of PEC chitosan-alginate particles. In addition,

new techniques for the synthesis of particles need to be explored so that a size range of 5-10 tm

is attained. Finally, cell studies are needed to assess the particles' immunogenic effects for their

application in drug delivery.













-CH
--OCH -C


Figure 3-1. PLA molecular structure.


Table 3-1. List of commercially available Pluronics


Chemical Name
Poloxamer 123
Poloxamer 124
Poloxamer 181
Poloxamer 182
Poloxamer 182 LF
Poloxamer 183
Poloxamer 184
Poloxamer 185
Poloxamer 188
Poloxamer 212
Poloxamer 215
Poloxamer 217
Poloxamer 231
Poloxamer 234
Poloxamer 235
Poloxamer 237
Poloxamer 238
Poloxamer 282
Poloxamer 284
Poloxamer 288
Poloxamer 331
Poloxamer 333
Poloxamer 334
Poloxamer 335
Poloxamer 338
Poloxamer 401
Poloxamer 402
Poloxamer 403
Poloxamer 407


Structure
HO(C2H30)7(C3H50)21(C2H30)7H
HO(C2H30)1 1(C3H50)21 (C2H30)1 1H
HO(C2H30)3(C3H50)30(C2H30)3H
HO(C2H30)8(C3H50)3o(C2H30)sH
HO(C2H30)8(C3H50)3o(C2H30)sH
HO(C2H30)10(C3H50)30(C2H30)10H
HO(C2H30)13(C3H50)3o(C2H30)13H
HO(C2H30)19(C3H50)3o(C2H30)19H
HO(C2H30)75(C3H50)30(C2H30)75H
HO(C2H30)8(C3H50)35(C2H30)sH
HO(C2H30)24(C3H50)35(C2H30)24H
HO(C2H30)52(C3H50)35(C2H30)52H
HO(C2H30)6(C3H50)39(C2H30)6H
HO(C2H30)22(C3H50)39(C2H30)22H
HO(C2H30)27(C3H50)39(C2H30)27H
HO(C2H30)62(C3H50)39(C2H30)62H
HO(C2H30)97(C3H50)39(C2H30)97H
HO(C2H30)10(C3H50)47(C2H30)1oH
HO(C2H30)21(C3H50)47(C2H30)21H
HO(C2H30)122(C3H50)47(C2H30)122H
HO(C2H30)7(C3H50)54(C2H30)7H
HO(C2H30)20(C3H50)54(C2H30)20H
HO(C2H30)31(C3H50)54(C2H30)31H
HO(C2H30)38(C3H50)54(C2H30)3sH
HO(C2H30)128(C3H5O)54(C2H30)128H
HO(C2H30)6(C3H50)67(C2H30)6H
HO(C2H30)13(C3H50)67(C2H30)13H
HO(C2H30)21(C3H50)67(C2H30)21H
HO(C2H30)98(C3H50)67(C2H30)98H


BASF (USA) Trade Name
Pluronic L43
Pluronic L44
Pluronic L61
Pluronic L62
Pluronic L62 LF
Pluronic L63
Pluronic L64
Pluronic P65
Pluronic F68
Pluronic L72
Pluronic P75
Pluronic F77
Pluronic L81
Pluronic P84
Pluronic P85
Pluronic F87
Pluronic F88
Pluronic L92
Pluronic P94
Pluronic F98
Pluronic L101
Pluronic P103
Pluronic P104
Pluronic P105
Pluronic F108
Pluronic L121
Pluronic L122
Pluronic P123
Pluronic F127


















Figure 3-2. Alginate molecular structure.





HOCH2 HO NH2


HO 1(1 O ,,0 H
S\ / /n<15

HO NH2 CH2OH


Figure 3-3. Chitosan molecular structure.










Table 3-2. List of Pluronic-PLA particles made.
Pluronic-PLA Ratio (Pluronic/PLA)
F 127-PLA 1:3
F 127-PLA 1:2
F 127-PLA 1:3
F 127-PLA 1:2
F 127-PLA 1:3
F 127-PLA 1:3
F 127-PLA 1:2
F 127-PLA 1:1
L101-PLA 1:2
L101-PLA 1:3
F68-PLA 1:2
P105-PLA 1:2
P103-PLA 1:2
P104-PLA 1:2
F108-PLA 1:2


Pluronic Solution Concentration
1.0 % w/v
1.0 %w/v
5.0 % w/v
10 % w/v
10 % w/v
20 % w/v
20 % w/v
5.0 % w/v
20 % w/v
20 % w/v
20 % w/v
20 % w/v
20 % w/v
20 % w/v
20 % w/v










Table 3-3. List of chitosan-alginate particles made
CaC12 or Number of
Sodium Alginate
Soiu int Cu(N03)2 Coatings
Solution
Solution
0.5 %w/v 0.5 %w/v
0.5 % w/v 1.0 % w/v
0.5 % w/v 5.0 % w/v
1.0 % w/v 0.5 %w/v
1.0 % w/v 1.0 % w/v
1.0 % w/v 5.0 % w/v
3.0 % w/v 0.5 % w/v
3.0 % w/v 1.0 % w/v
3.0 % w/v 5.0 % w/v 0
3.0 % w/v 5.0 % w/v 1
3.0 % w/v 5.0 % w/v 5
3.0 % w/v 5.0 % w/v 9
3.0 % w/v 5.0 % w/v 0
3.0 % w/v 5.0 % w/v 1
3.0 % w/v 5.0 % w/v 5
3.0 % w/v 5.0 % w/v 9
3.0 % w/v 5.0 % w/v 0
3.0 % w/v 5.0 % w/v 1
3.0 % w/v 5.0 % w/v 5
3.0 % w/v 5.0 % w/v 9
3.0 % w/v 0.5M 0
3.0 % w/v 0.5M 1
3.0 % w/v 0.5M 5
3.0 % w/v 0.5M 9
3.0 % w/v 0.5M 0
3.0 % w/v 0.5M 1
3.0 % w/v 0.5M 5
3.0 % w/v 0.5M 9
3.0 % w/v 0.5M 0
3.0 % w/v 0.5M 1
3.0 % w/v 0.5M 5
3.0 % w/v 0.5M 9


Chitosan or COS Alginate Solution
Solution


0.3 % w/v
0.3 % w/v
0.3 % w/v

1.0 % w/v
1.0 % w/v
1.0 % w/v

3.0 % w/v
3.0 % w/v
3.0 % w/v

0.3 % w/v
0.3 % w/v
0.3 % w/v

1.0 % w/v
1.0 % w/v
1.0 % w/v

3.0 % w/v
3.0 % w/v
3.0 % w/v


0.3 % w/v
0.1 %w/v


0.1 %w/v
0.1 %w/v


0.1 %w/v
0.1 %w/v


0.1 % w/v
0.1 % w/v


0.1 % w/v
0.1 % w/v


0.1 % w/v
0.1 % w/v


















A B C


Figure 3-4. Optical micrograph of 20% Pluronic-PLA (1:2) microspheres A) F127-PLA, B)
L101-PLA, and C) F68-PLA. Original magnification = 100X; bars denote 50 [m.





















Figure 3-5. Optical micrograph of 20% P105-PLA (1:2) microspheres. Original magnification
100X.


A B C


Figure 3-6. Scanning electron micrograph of 20% P105-PLA Particles. A) 400X, bar denotes
100 atm; B) 1000X, bar denotes 50 atm; and C) 800X, bar denotes 50 tm.


Table 3-4. Composition and deformability properties of Pluronic-PLA microspheres
> *\ Max Negative Suction
Pluromnic-PLA (Ratio) Max Negative2 Deformation
PlonicPLA ( ) Pressure x 10' (dynes/cm) Deformation
F127-PLA (1:3) -58.8 No
F127-PLA (1:2) -58.8 No
L101-PLA (1:2) -58.8 No
L101-PLA (1:3) -58.8 No


F68-PLA (1:2)
P105-PLA (1:2)
P103-PLA (1:2)
P104-PLA (1:2)
F108-PLA (1:2)


-44.1
-24.5
-58.8
-58.8
-58.8


Only the shell
Only the shell
No
No
No
































Figure 3-7. Optical and scanning electron micrographs of 0.3% COS Ca-crosslinked alginate
particles: A) 400x, 5 coatings; B) 1000x, 1 coating; C) 1000x, 5 coatings; and D)
1000x,9 coatings. Bars denote 50 [m.


A B


Figure 3-8. Optical and scanning electron micrograph of 3% COS Cu-crosslinked alginate
particles, 5 coatings: A) 400x and B) 1000x. Bars denote 50am.










Table 3-5. Composition and deformability properties of PEC chitosan-alginate microcapsules.
Code Max Negative Suction
Code 3 2 Deformation
Pressure x 103 (dynes/cm2) Defor
Ca-0 0 Disintegrated
Ca-1 0 Disintegrated
Ca-5 0 Disintegrated
Ca-9 0 Disintegrated
Cu-0 0 Disintegrated
Cu-1 -19.6 Yes. Some particles ruptured
Cu-5 -19.6 Yes. Particles completely aspirated
Cu-9 -19.6 Yes. Portions of the particles aspirated









CHAPTER 4
SYNTHESIS AND CHARACTERIZATION OF PHYSICO-CHEMICAL AND RHEOLOGY
PROPERTIES OF POLYELECTROLYTE CHITOSAN-ALGINATE MICROPARTICLES
CROSSLINKED WITH CALCIUM, ZINC, OR COPPER IONS

Introduction

Microencapsulation techniques have been widely used to develop controlled-release drug

delivery systems, masquerade tastes and odors, reduce toxicity, and protect cells from the host

immune response in the absence of immunosuppression drugs [45, 46]. Some of these

techniques include emulsion solvent evaporation, solvent extraction, coacervation, spray-drying,

interfacial complexation, coating, and hot melt coating [46, 51]. Each method has both

advantages and disadvantages in the elaboration of polymeric microparticles.

The most frequently used method for the fabrication of alginate particles is the air-spray

crosslinking technique. Briefly, this method consists of spraying an alginate solution into a

collecting bath containing the crosslinking agent. As soon as the alginate droplets are in contact

with the crosslinking solution, particles are formed by gelation. The system for the development

of alginate microspheres can vary from very simple to more complex setups. The spray

approach can be attained by using a spray bottle or a more complicated micro-fluid device. The

latter is based on the extrusion of the alginate solution through an inner lumen while air or a

secondary solution is flowing on the outer lumen of the system. Alginate drops forming at the

ending tip of the inner lumen are detached by the air shear forces exerted on them.

The other method used to prepare alginate particulates is based on an emulsion-gelation

technique. The approach involves mixing of the alginate solution with an organic solvent to

create a water-in-oil (W/O) emulsion, followed by the addition of the crosslinking solution. This

technique can be varied by mixing the corsslinking solution with the organic phase before

creating the W/O emulsion. Even though both methods are fairly simple, cost effective, and up-









scalable for mass production to some degree; they also have drawbacks which limit their

application.

This study is focused on the synthesis and characterization of polyelectrolyte complex

(PEC) chitosan-alginate microparticles. Deformability, particle size, and toxicity are the main

goals of the study. Several variables, i.e. cross-linking agents, solvents and manufacturing

techniques, were studied to determine the set of parameters that would yield the most promising

red blood cell analog, in terms of deformable capability. In addition to deformability, the aim

was to obtain particles with a size range between 5 and 10 micron so that they are small to

deform and pass through capillaries as well as spleen fenestrations, while having a great drug

loading potential compared to the existing nano-size drug delivery particulates.

In this study we examined the application of two frequently used techniques in the

development of alginate particles. The air-spray crosslinking technique and an emulsion gelation

technique were applied for the synthesis of particles. Besides discerning the most appropriate

manufacturing process, the effect of different crosslinkers in the development of PEC

microparticulates was also studied. Calcium, zinc, and copper were the divalent cations used as

crosslinking agents and their effect on particles' morphology, deformability, stability, and

toxicity was analyzed.

Materials and Methods

Preparation of PEC Particles

Air-spray crosslinking technique

Natural polymeric microspheres were prepared by an air-sprayed crosslinking technique

[93]. The polymer used was alginate from Keltone (LV, food grade). From preliminary

experimentation, an initial concentration of 0.5M calcium chloride, zinc chloride and copper

nitrate solutions were used as the crosslinking baths. Sodium alginate was dissolved in









deionized water at a 3% (w/v) concentration. A double-lumen device was used to air-spray the

alginate solution into the stirred gelling bath. To control the size of the droplets, the following

parameters were studied: needle gauge, alginate solution flow rate, air pressure, and distance

between the needle tip and the gelling bath. While studying one parameter, all the others

remained constant in addition to alginate and gelling ion concentrations.

The alginate solution was extruded through a syringe needle (gauge range: 16G 30G) at a

constant rate of 1.2, 6, or 12ml/h using a Harvard dual-syringe pump (Harvard Apparatus,

Holliston, Massachusetts). Air was infused through the outer lumen at constant pressures

ranging from 40 to 60 psi, in 10-psi increments. Finally, the drop fall distance was varied from 5

to 15 cm, in increments of five. After alginate droplets were extruded into the crosslinking bath,

they were stirred at room temperature for 10 minutes and cured overnight. Alginate

microcapsules were then collected by centrifugation. Particles were washed three times with

deionized water (dH20) and stored at 40C for future coating.

W/O emulsion gelation technique

Sodium-alginate was dissolved in dH20 at a 3% (w/v) concentration and emulsified in an

organic phase containing soybean oil or cyclohexane (Aldrich, St. Louis, MO) and surfactant at

different concentrations (0.5 and 5.0 %, w/w). Pluronics L61, L121, L101, L64, L43 (kindly

donated by BASF) and Tween 80 were the surfactants tested. The first emulsion was obtained

by ultrasonication of the two phases at 60W for 1 min in an ice bath. A second aqueous solution

containing 0.5M of the crosslinking agent was added to the emulsion by air-spray (40 psi,

20ml/hr) at a 4-cm dropping distance while stirring the whole medium slowly with a magnetic

stirrer. Particles were allowed to cure for ten minutes under continuous stirring. Then medium

was allowed to rest for 24 hr so that particles would drop to the bottom of the container while the

oil phase was left at the top. After separation of the two phases, particles were collected by









filtration through a 45-[tm mesh and washed copiously to remove the organic solvent. When

soybean oil was used, the bead slurry was collected by centrifugation; washed ten times with

10% ethanol, once with 0.1% Triton solution and three times with dH20. When cyclohexane

was used, the particle solution was warmed at 370C to induce solvent evaporation. After

removal of the organic phase, particles were stored at room temperature for future coating.

Coating of alginate particles

Ca-, Zn, and Cu-alginate particles were coated applying the layer-by-layer (LbL)

absorption technique. The coating always started with the cationic polymer and ended with the

anionic polymer. Microcapsules with zero, two, six, and ten alternating layers were fabricated.

The cationic polymer used was a high-deacetylated, low-molecular-weight chitosan

oligosaccharide lactate (Mw < 5000, 90% deacetylation) from Aldrich. Alginate was the anionic

polymer used. A 3 % (w/v) chitosan solution containing CaC12 or a 0.1 % (w/v) alginate solution

was added to crosslinked-alginate particles dispersed in deionized water. After allowing the

particles to coat for 30 minutes, they were collected by centrifugation and washed three times

with deionized water to ensure that all free polyelectrolytes were removed. Following the

washes, particles were ready for the next coating or stored at 40C until use. Table 4-1 shows the

different batches made; each batch was repeated three times.

Decomposition of the particles' core

Decomposition of the particles' core was achieved by incubating particles in cell culture

medium, DMEM/F 12 50:50 or RPMI 1640, (Cellgro, Herndon, VA) supplemented with 5%

(w/v) bovine serum albumin, BSA (Sigma, St. Louis, MO). Multiple-layer particles, at a 5%

(w/v) concentration, were incubated at room temperature in the two types of chelating media,

differing in their electrolyte concentrations. After close observation, old medium was replaced

by same amount of fresh medium if core was not dissolved. Medium was replaced until









decomposition of the core was achieved. Gel-core particles were then stored at 40C until further

use.

Characterization of PEC Particles

To characterize the efficacy of the two manufacturing processes used, analyses of the

particles' surface morphology and size distribution were carried out. To characterize the effects

of using calcium, zinc, and copper as crosslinker agents in the development of alginate particles,

more tests were performed, including: surface morphology analysis, particle size distribution,

deformability assessment, particle stability and viscosity, coating adsorption, ion exchange, and

biocompatibility.

Surface morphology analysis

Surface morphology and stability of the microcapsules were examined by light microscopy

and scanning electron microscopy (SEM). Samples for scanning electron micrographs were

obtained after 0, 2, 6, and 10 coatings. Droplets of microsphere solutions were mounted on

aluminum stubs, let air dried and sputter-coated with gold and palladium particles. The stubs

were mounted in a scanning electron microscope at 10.0 kV and imaged at x500, x1000, and

x5000.

Particle size distribution

Size of microcapsules was analyzed by dispersing particles in deionized water at a

concentration of 0.1% (w/v). Measurements were carried out by a Beckman LS13320 Particle

Characterization Coulter (Beckman Instruments, Fullerton, CA). Calculation of the particle sizes

was carried out using the standard modus of the LS13320 Particle Size Analyzer software

(Beckman Instruments, Fullerton, CA). Percentage of particle diameters was used to describe

particle size. Each sample was measured in triplicate.









Particle stability analysis

To determine the particles' stability, different approaches were used. Initially, particles

were suspended at a concentration of 5% (w/v) in different types of media: dH20 (negative

control), regular RPMI 1640 and DMEM/F12 50:50 cell culture media, or 5 % BSA

supplemented RPMI 1640 and DMEM/F12 50:50 media (Table 4-2) [12]. Particles were

incubated at 250C or 37C under continuous orbital rotation to ensure constant mixing. At 0,

0.5, 1, 2, 4, 8, 12, 24, and 48 hours; samples were collected, visually inspected and subjected to a

partial vacuum pressure. Briefly, a 5-mL air-displacement pipetter and tip (Eppendorf,

Westbury, NY) set at the maximum volume were used to aspirate the particles. For this study,

large particles with a diameter ranging between one and two millimeters were used to facilitate

optical inspection. The total number of intact particles was represented as the percentage of the

total number of particles inspected.

Another approach used to determine stability of the particles involved centrifugation and

optical inspection. The study was carried out using the Eppendorf MiniSpin Microcentrifuge

(Eppendorf, Westbury, NY) which provides speeds of up to 14,000g. Particles' weight was

obtained to determine the centrifugal force exerted on them. For this study large particles and

microparticles were incubated at 250C in dH20, DMEM/F12 50:50, or 5 % BSA DMEM/F12

50:50 cell culture medium. After 0.5, 1 and 7 days, the 5 % (w/v) particle suspensions were

centrifuged for 20 sec or 5 min at centrifugal speeds ranging from 1000 13,000 rpm (67 -

11,337g). Particles were inspected prior and post centrifugation; changes in morphology and

ability to re-disperse were reported.

The other method used to verify particles' stability involved using the Wells-Brookfiled

Cone/Plate Digital Viscometer System (Brookfield, Stoughton, MA) with a CP-52 conical

spindle. Microparticles suspended in 5 % BSA DMEM/F12 50:50 cell culture medium at a 12 %









(w/v) density were subjected to a shear rate of 200 sec-1 for one minute. At the end of each run,

samples were collected for light microscopic visualization. All analyses were conducted in

triplicate.

Measurement of particles' deformability

The micropipette aspiration technique was applied to assess particles' deformability.

Micropipette tips with inside diameters approximately 20% smaller than the particles' diameter

were fabricated. Previous to their use, pipettes were backfilled with 0.9% NaCl solution using a

plastic syringe with a 97-mm long, 28G backfiller (Microfil, World Precision Instruments,

Sarasota, FL). Before particles were used for deformability studies, they were incubated in

sequestrant-containing media until core was decomposed. The prepared chamber was then filled

with microparticles suspended in sequestering media and mounted on the microscope stage.

An initial negative pressure of 5 cm H20 (approximately -4.9x103 dynes/cm2) was applied

and pressure was augmented in increments of five. After a section of the particle had been

aspirated, the negative pressure was decreased and the particle was unloaded. The whole

aspiration process was monitored and recorded using an inverted light microscope (Axiovert

100, Zeiss), a Carl Zeiss video camera system, and a VCR. The recorded images were converted

into digital images using the Matrox software. Since the main goal of the study was to attain

particles capable to deform under micropipette suction pressures used to aspirate erythrocytes

[96], particles were considered deformable only when using pressures up to -40x103 dynes/cm2.

Assessment of particles' viscosity

The viscosity of the particles in simulating plasma medium (DMEM/F12 50:50

supplemented with 5% BSA) was measured using the Wells-Brookfiled Cone/Plate Digital

Viscometer System (Brookfield, Stoughton, MA) with a CP-52 conical spindle. The principle of

the system is based on the rotation of the conical spindle at an accurate speed and detection of









the torque needed to overcome the viscous resistance caused by the sample fluid between the

cone and a fixed flat plate [103]. Since the cone/plate viscometer exerts a shear rate on the

particle suspension, this test in combination with light microscopy was also used to determine

particle stability.

Particles were suspended in simulating plasma media at a 12% (w/v) density and incubated

at room temperature for 2 hours prior to obtaining the theological measurements. After filling

the chamber with the required 0.5-mL sample volume, measurements were taken at a fixed

spindle speed of 100 rpm. At the end of each reading, samples were collected for visualization

of the particles' stability. Viscosities were calculated using the formulas and ranges found in the

instruction manual:

Factor = Range/100
Viscosity = Display Reading x Factor

where, the range is specific to the cone and speed used. For the cone CP-52 rotating at a

constant speed of 100 rpm, the range is 983 cps (983 mPa-sec) while the shear rate exerted on

the fluid sample is 200 sec-1

Coating adsorption analysis

Adsorption of each polyelectrolyte layer onto microcapsules was examined by measuring

the particle surface charge changes. Samples were dispersed in dH20 at a concentration of 0.1%

(w/v) and measurements were carried out by the Brookhaven ZetaPlus Analyzer (Brookhaven

Instruments Corp.,USA). The zeta potential was calculated from the solution conditions and the

measured electrophoretic mobility. Each sample was measured in triplicate and the values

reported were the mean value for the three replicate samples.









Ion exchange analysis

This analysis was only performed for Cu-crosslinked particles due to the toxicity concerns

regarding this transitional metal. To determine the amount of copper ions removed by albumin,

a modified bicinchoninic acid (BCA) protein assay was used. Cell culture media, RPMI 1640,

supplemented with 1% (v/v) penicillin/streptomycin and 5% (w/v) bovine serum albumin (BSA)

was used. Microparticles were suspended in supplemented media at a concentration of 5%

(w/v). Particles were incubated at 250C under continuous orbital rotation to ensure constant

mixing. At 24, 48, and 72 hours, samples were removed from the incubator and centrifuged at

3,000 rpm for 5 min. Supernatant was then removed and stored at 40C for future analysis.

The removed solution was replaced with an equal volume of fresh supplemented media.

Sample tubes containing microspheres were returned to the rotating incubator at previous

temperatures until the next time point. Analysis of the stored supernatant was conducted using

the bicinchoninic acid reagent from a BCA kit (Pierce, Rockford, IL). A purple-colored reaction

product was yielded by this assay with a strong absorbance at 562 nm. Concentrations of

albumin-reduced copper cations were determined by comparison to a standard curve. All

analyses were conducted in triplicate.

Assessment of particle cytotoxicity

Human dermal fibroblasts (HDF, passage 10) from ATCC (Manassas, VA) were used to

determine the level of toxicity of uncoated and coated calcium-, zinc-, and copper-crosslinked

particles. The type of test performed was indirect meaning extracts from the particles were

added to the cells instead of adding particles directly. To determine cell survival and recovery,

the MTT cell proliferation assay (Promega, Madison, WI) was used which measures the ability

of cells to convert the tetrazolium compound MTT by the action of succinate dehydrogenase to

water-insoluble formazan crystals. Two time points were studied: a short-term test to









demonstrate particles' toxic effects on cells and a long-term test to demonstrate survival, the

retention of cell regenerative capacity.

Previously made particles (all batches from Table 4-1) were dried so that a predetermined

particle concentration could be obtained. Different concentrations of copper powder were used

as positive controls. Particles and controls were incubated in DMEM (Cellgro, Herndon, VA)

cell culture medium supplemented with 5 % fetal bovine serum (FBS) and 1 %

antibiotic/antimycotic solution at 370C under continuous orbital rotation for 24 hr. Samples

were then removed from the incubator and centrifuged at 12,225g for 10 min. The supernatant

was collected, sterilized by membrane filtration (0.2 atm; Whatman), and stored at -800C for

future analysis.

HDF cells were seeded onto 96-well plates (Costar, Corning, NY) at a 1,000 cells/well

density. Initially cells were incubated in DMEM cell culture medium supplemented with 10 %

FBS and 1 % antibiotic/antimycotic (Ab/Am) solution. Cells were incubated at 370C, in a 95 %

02/5 % CO2 atmosphere for 24 hr before adding any treatment.

Extracts from the particles and positive controls were thawed and diluted in 2 % FBS and 1

% Ab/Am cell culture medium to desired concentrations. The particle and control concentrations

used were: 0.1, 1.0, 10, 100, andl000 atg/ml (w/v). Cells incubated in 10 % and 2 % FBS culture

medium and exposed to zero treatment were used as negative controls. Old cell medium was

aspirated and replaced by 100 atl of cell culture medium containing treatments. Cells were

exposed to the different conditions for 24 hr. When exposure time was over, cells were either

collected for MTT assay (short-term test) or culture medium containing treatments was replaced

by 10 % FBS medium (long-term test). For the long-term test, cells were allowed to recover for

2 days before performing the MTT assay. After completion of the cell viability/proliferation









assay, optical densities of each well were measured by a microplate reader set at 490 nm. All

conditions were done in replicates of six.

Data Interpretation

Data were expressed as mean values + standard error of the mean (SEM). To describe

statistical differences, one-way analysis of variance (ANOVA) and Tukey-Kramer multiple

comparison post test were used. Statistical significance was defined asp< 0.05.

Results and Discussion

The standard method of forming alginate beads by extruding alginate drops into a

crosslinking bath for gelation generates large particles with a diameter range between two and

five millimeters. To synthesize smaller alginate particles, two methods were employed: the air-

spray crosslinking technique and the water-in-oil (W/O) emulsion gelation technique. A 3 %

(w/v) alginate concentration and crosslinking ion solutions of 0.5M concentrations were used.

To control the size of the particles in the air-spray crosslinking technique, the following

parameters were studied: needle gauge, alginate solution flow rate, air pressure, and distance

between the needle tip and the gelling bath. For the emulsion-gelation technique, the oil phase

was composed of soybean oil or cyclohexane and diverse surfactants; also different stirring rates

were used to determine the optimal particle size.

Particles made with the air-spray crosslinking technique presented a monodispersed size

distribution depending on the parameter studied. The size of particles diminished significantly

when the inner diameter of the needle decreased (Figures 4-1 and 4-3) as expected. Likewise,

reducing the speed of alginate extrusion or increasing the pressure of air infused through the

outer lumen caused a significant reduction in the particle's diameter (Figures 4-2, 4-4, and 4-4).

The particle-forming properties were not altered significantly with changes in the distance

between the needle tip and the gelling bath tested in this study. These results are consistent with









many studies from other research groups since it is well-known that the particle size in the air-

spray method is not only determined by the needle gauge but more importantly by the air flow

rate [104, 105, 106].

Tables 4-3, 4-4, and 4-5 illustrate the particle sizes and standard error of the means under

various conditions using different needle gauges, alginate flow rates, and air pressures,

respectively. The smallest particles were obtained when alginate was extruded through a 30G

needle while the air pressure in the outer lumen was increased to 60 psi. Under these conditions,

the smallest average particle size formed was 44 3 [tm in diameter. Although, we were able to

decrease the particle sizes, the capability to produce monodispersed batches was lost at the air

pressure of 60 psi, in agreement with optical observations (Figure 4-5). The lost of

monodispersity due to increased outer lumen air pressures could be explained by drop coalition

at higher pressures. This observation was also reported by Haas mainly for the production of

smaller particles [107].

Another parameter tested in the formation of alginate particles was the usage of different

divalent cations as crosslinkers. Figure 4-6 shows a histogram of particle diameters as a function

of crosslinking ions obtained with the 30G needle, a 50-psi outer lumen air pressure, and a 1.2-

mL/h alginate extrusion rate. Using different crosslinkers altered significantly the size of

particles formed (Table 4-6). For alginate particles crosslinked with copper or zinc, the average

particle size obtained was 48 4 and 71 4 [tm, respectively. This difference in the mean size

of the particles between copper- and zinc-crosslinked alginate beads could be explained by the

polymer cation affinity. Even though the crosslinking mechanisms of copper and zinc ions are

not understood, it is well-known that the affinity of alginates for divalent cations depends mostly

on the electronic structure of the cation, with a higher affinity for copper ions than for zinc ions









[85, 100, 101, 102]. Copper is able to crosslink more densely with sodium alginate, resulting in

smaller particles with reduced water content as opposed to zinc-crosslinked alginate particles.

Particles made with the emulsion-gelation technique presented a poly-dispersed size

distribution, with diameters ranging from 2 to 60 [tm (Figure 4-7B). The best results were

obtained when using the Pluronic L61 concentration of 0.5% and a stirring speed of 2000 rpm.

Pluronic L61 is considered a hydrophobic surfactant with a 10 % (w) PEO content and a PPO

molecular weight of 1800 Da. Big agglomerates were obtained when using other hydrophobic

Pluronics: L101 and L121, both with a 10% (w) PEO content and PPO molecular weights of

3300 and 4000 Da, respectively. Indeed, the difference in morphology is attributed to the

molecular weight of the PPO parts. Hydrophobic, low-molecular-weight PPO Pluronics are

preferred for the production of alginate particles as opposed to Pluronic-polyester systems

where a high molecular weight of the PPO content is necessary to provide enough anchoring of

the copolymer to the particle surface [36, 64].

Variations in the stirring speed during the gelation process altered the morphology of the

particles. An increment in the average of particles' diameters was observed as the stirring speed

was augmented. This observation could be explained by an increased coalition of particles prior

to a complete matrix gelation. While the use of soybean oil as the organic phase was very

appealing due to its mild condition, complete oil removal was extremely hard to achieve.

Although samples were copiously washed, oil was still present in the particles. Microparticles

made using cyclohexane as the organic phase presented a narrower size distribution with the

majority of particles' diameters ranging from 5-15 tm. Solvent removal was successfully

achieved by evaporation.









The particles-forming properties in the emulsion-gelation technique were significantly

altered with changes in the crosslinking ions. Particles crosslinked with copper presented a tear-

drop shape as oppose to the spherical shape of particles crosslinked with zinc and calcium

(Figure 4-8). In addition to variations in the particles' shape, the size of the particles was also

affected by the different crosslinking ions. Table 4-7 illustrates the fractions of particles within

different size ranges. Also, size distribution histograms of Ca-, Zn-, and Cu-crosslinked particles

are shown in Figure 4-9. Although the majority of all the microspheres made were 0 20 [tm in

diameter, the smallest uniform microspheres were obtained with calcium and zinc. They had the

largest fractions, 69.7 9.77% and 77.2 0.50% respectively, in the size range of 5 10 atm;

while copper-crosslinked microparticles had the largest fraction (46.6 0.96%) in the size range

of 15 20 tm.

The difference in particles' morphology and size was not only dependent on the crosslinker

but also on the solvent used as the organic phase of the emulsion. The sizes of Ca-, Zn-, and Cu-

crosslinked particles were larger when soybean oil was used. However, Cu-crosslinked alginate

particles presented the smallest diameters (data from Chapter 3) and a spherical morphology.

The contradictory results between the use of soybean and cyclohexane could be explained by

differences in the solvents' viscosities and in the alginate's affinity for the cations. Since

alginate has the highest affinity for copper ions, the crosslinking reaction occurs a lot faster than

for the other ions. With high viscous solvents, such as soybean oil, a fast crosslinking reaction

prevents destabilization of the emulsion and, therefore, less coalition of small particles into large

ones. In addition, phase separation takes more time in highly viscous solvents, which slows

down particles' settlement into the aqueous phase resulting in spherical particles. On the









contrary, in low viscous solvents (i.e. cyclohexane) copper-crosslinked particles form rapidly and

settle quickly into the aqueous phase, generating the tear-drop shape.

Ca-, Zn, and Cu-alginate particles were coated applying the layer-by-layer (LbL)

absorption technique. The coating always started with the cationic polymer and ended with the

anionic polymer. Microcapsules with zero, two, six, and ten alternating layers were fabricated.

Figure 4-10 shows photographs of alginate beads made with the conventional dripping

mechanism: by extruding alginate as drops into a crosslinking solution for gelation. These 2 3

mm beads presented a spherical shape and size that were independent of the type of crosslinker

used. However, beads' mean diameter decreased with multiple coatings and their shape changed

from spherical to biconcave. Zinc-crosslinked particles showed a more pronounced biconcave

configuration as opposed to calcium- and copper-crosslinked particles. These observations could

be explained also by the affinity of alginate for divalent ions in the order of

Pb2+>CU2+>Cd2+>Ba2+>Ni2+>Ca2+>Zn2+>C2>Mn2>Sr [85, 100, 101, 102]. The new shape of

the coated particles was very similar to the biconcave shape of red blood cells, which represents

a great advance in the drug delivery field. It is well-known that the extraordinary flexibility and

respiratory functions of red blood cells are attributed to their structural characteristics [11, 9, 10].

The biconcave disk shape of our particles provides a greater surface area ideal for gas exchange

as opposed to the traditional spherical shape of the existing drug delivery particulates.

The surface morphology of microparticles made with the gelation-emulsion technique was

determined by SEM analysis for uncoated and bilayer PEC alginate particles. A difference in

morphological structure can be observed among alginate particles formed with calcium, zinc or

copper cations. As shown in Figure 4-11 (C,D,E,F), the surface morphology of the Zn- and Cu-

microparticles looked smoother before any coating was applied. On the contrary, the calcium-









alginate microspheres looked very rough on the surface even prior to any coating (Figure 4-11

(A,B). This sponge-like structure of the uncoated calcium-alginate microspheres could be an

artifact created by collapsing of the pore walls due to dehydration. Although the order affinity of

alginates for cations has been established as Cu2+>Ca2+>Zn2+, binding sites for these cations are

different [108]. Zinc is able to crosslink less selectively than calcium and hence produces more

extensive crosslinking of alginate [109, 108]. As a result, calcium cations generate a more

permeable alginate matrix with a high water content that is more susceptible to morphological

changes after dehydration. Bilayer PEC microparticles presented a very rough surface

morphology with some aggregations which is attributed to the PEC coatings.

Size distribution histograms of Ca-, Zn-, and Cu-crosslinked uncoated and coated

microparticles are shown in Figure 4-12. Although the majority of all the microspheres made

were 0 20 [tm in diameter, the size distributions changed significantly with the number of

coatings. Uncoated calcium- and zinc-alginate microspheres had a very narrow size distribution

with more than 70% of the particles in the 5 10 [tm range; while coated particles presented a

more polydispersed size distribution. Differences in size distribution between uncoated and

multiple coated microparticles could be a consequence of particle agglomeration which increased

with the number of coatings.

Stability of the particles was determined by the three different methods previously

described. Initially, particles were suspended at a concentration of 5% (w/v) in dH20, RPMI

1640 and DMEM/F 12 50:50 cell culture media supplemented with different concentrations of

BSA. Particles were incubated at 250C or 370C under continuous orbital rotation and at

predetermined intervals samples were collected, visually inspected and subjected to a partial

vacuum pressure. All particles incubated in dH20 looked intact independent of the number of









coatings or crosslinkers at the end of the study. Uncoated Ca-, Zn-, and Cu-alginate particles

dissolved after one hour incubation in both 5% BSA cell culture media at 370C (Figure 4-13).

By day 1, all calcium-alginate uncoated and coated particles were dissolved regardless of the cell

culture medium (Figure 4-13 A, B).

It is apparent from the histograms shown in Figure 4-13 (C,D,E,F) that the stability of

zinc- and copper- alginate particles depended on the number of coatings and the incubation

medium. Stability of these particles was directly proportional to the number of coatings. Also,

particles were more stable in DMEM/F 12 50:50 cell culture medium than in RPMI 1640. These

results were expected due to the electrolyte content difference of each media. DMEM/F12 50:50

electrolyte content is very similar to human plasma (Table 4-2); however, RPMI 1640 has

approximately a 6-fold reduction of Ca2+ and a 10-fold excess of phosphate ions. This difference

in ions is detrimental for the stability of PEC alginate particles. The ion exchange properties of

alginate gels have been studied by many research groups and the process depends on several

factors such as: the amount of crosslinking ions forming the gel, physical properties of the gel,

competing ion concentration, pH and ionic strength of the solution [97, 98, 99].

After removal of crosslinking ions from the particles by albumin and phosphate ions

present in the RPMI 1640 cell culture medium, the polyelectrolyte complex system was not

stable anymore. The calcium PEC system collapsed regardless of the number of PEC layers

because alginate had no affinity for the other electrolytes present in the medium. The zinc and

copper PEC systems were stable slightly longer because of the alginate affinity for the cations

and the more extensive crosslinking of these systems. However, the lack of calcium ions that

would replace the alginate binding sites left by the previous crosslinkers accelerated the collapse

of the systems. On the contrary, when zinc and copper ions were chelated by albumin in









DMEM/F12 medium, the alginate binding sites left were occupied by calcium ions present in the

cell culture medium. As a result, the ion exchange process along with the PEC layers kept the

pre-crosslinked Zn- and Cu-alginate particles stable.

Figures 4-14 and 4-15 illustrate how other factors alter the ion exchange process in alginate

gels. Stability of the particles is directly proportional to the concentration of albumin present in

the medium (Figure 4-14), as expected. Temperature is another important factor influencing the

ion exchange process. The kinetics of ion exchange in our chitosan-alginate PEC system was

slower at 250C than at 370C. Although our results were expected, we cannot determine the

equilibrium and kinetics of the ion exchange process taking place in the PEC coated zinc-and

copper-alginate microparticles at this time. Previous studies on the kinetics of metal ion uptake

by alginates showed that uptake begins with a rapid phase (minutes to hours) followed by a

relative slow phase (up to one day)[97]. There are many models that describe the equilibrium

and kinetics of ion exchange in alginate gels, such as the Langmuir, Freundlich, and the surface

complex formation models [97]. Although each of the models has advantages and

disadvantages, it may be possible to use one of them to determine the equilibrium of our system.

The other two approaches used to determine stability of the particles involved

centrifugation and shear followed by optical inspection. Zn- and Cu-alginate particles with zero,

two, six, or ten PEC coatings were incubated in 5% BSA DMEM/F12 50:50 medium at a 5%

(w/v) density overnight. Particles were then exposed to centrifugal speeds ranging from 1000 -

13,000 rpm (67 11,337g) for 20 seconds or 5 minutes. Uncoated Zn-alginate particles failed

when centrifuged at 1000 rpm for 20 sec; while uncoated Cu-alginate particles failed at 4000

rpm. All the coated zinc- and copper-alginate particles sustained all the centrifugal speeds, even

at the high speed of 13,000 rpm for 5 min. Figures 4-16 and 4-17 illustrate optical micrographs









of zinc- and copper-alginate PEC microparticles exposed to a shear rate of 200 sec-1 for 1 min.

Microparticles had been suspended in 5 % BSA DMEM/F12 50:50 medium at a 12 % (w/v)

density prior to testing. Uncoated zinc particles were completely destroyed (Figure 4-16A);

while pieces of very swollen uncoated copper particles could still be seen (Figure 4-17A).

Differences between the two, six, and ten multilayer particles were not apparent from Figures 4-

16 (B,C,D) and 4-17 (B,C,D). Although some of the bilayer zinc particles looked slightly

swollen (Figure 4-16B), the difference was not significant. The observations from the

centrifugation and shear tests coincide with our previous results, showing again the strength of

copper-alginate gels over the zinc-alginate ones. In addition, it is evident the stabilizing role of

PEC coatings in both systems.

The micropipette aspiration technique was used as previously described to determine

deformability of the calcium-, zinc-, and copper- alginate PEC particles. Concurring with our

previous results, all coated alginate particles suspended in dH20 were not deformable when

using the micropipette. However, when crosslinking ions were removed, zinc and copper

particles presented deformable properties, while calcium particles disintegrated regardless of the

number of PEC coatings. As Table 4-8 and Figure 4-18 illustrate, zinc and copper particles

deformed and were completely aspirated into the micropipette tip after applying a 20-cm-H20

negative pressure (-19.6x103 dynes/cm2).

While uncoated particles were not stable after removal of crosslinking ions; bi-layer

particles were stable mostly in static conditions. Less pressure (-15 cm H20; -14.7x103

dynes/cm2) was necessary to aspirate PEC bilayer zinc-crosslinked microcapsules. Some of the

bi-layer zinc and copper particles ruptured under pressures of-15 and -20 cm H20, respectively.

Best results were obtained with microcapsules coated six times. Hexa-layer zinc and copper









microparticles were able to deform under a 20-cm-H20 negative aspiration pressure several

times while recovering their shape. Removal of crosslinking ions was not as efficient for the ten-

layer copper microparticles as for the lower-layer ones under the conditions tested. As a result,

only portions of the deca-layer Cu-crosslinked microparticles were deformable under a 20-cm-

H20 negative pressure as oppose to the 10-PEC-layer zinc microcapsules which were completely

aspirated under the same pressure. These observations agree with our previous results,

demonstrating once more the difference in strength of copper-, zinc-, and calcium-alginate gels

along with the stabilizing role of the PEC coatings.

Viscosity of the particles in simulating plasma medium (DMEM/F12 50:50 supplemented

with 5% BSA) was measured using the Wells-Brookfiled Cone/Plate Digital Viscometer System

with a CP-52 conical spindle. Zinc and copper microparticles were suspended in simulating

plasma medium at a 12% (w/v) density and incubated at room temperature for 2 hours prior to

obtaining the theological measurements. Measurements were taken at a fixed spindle speed of

100 rpm and a shear rate of 200 sec -. Viscosity values for blood, plasma, and serum were

obtained from studies done in healthy adults by Rosenson et al [24]. A histogram of the

samples' viscosity is given in Figure 4-19 and Table 4-9 demonstrates the viscosity values and

standard error of the means obtained under the conditions tested.

It is apparent from the data that viscosity values of uncoated zinc and copper microparticle

suspensions were significantly higher compared to the viscosity of blood. The uncoated Zn and

Cu particle suspensions presented increments in viscosity of seven and six folds, respectively. It

can be seen from Figures 4-16A and 4-17A that uncoated zinc particles were completely

destroyed while there were only remaining pieces of very swollen uncoated copper particles.

Rupture of particles followed by a release of alginate molecules into the salt-containing medium









caused an increment of the suspensions' viscosities. Solutions of bilayer PEC zinc and copper

particles had viscosities of 4.92 and 5.90 mPa*s, respectively. These values were not

significantly different from the viscosity values of human blood and Figures 4-16B and 4-17B

demonstrate the presence of particles in the solutions.

While the viscosity average for the solutions containing copper microcapsules with six

PEC coatings was 4.92 mPa*s, six-layer zinc-alginate microparticle solutions had a high

viscosity value of 9.83 mPa-s for the initial 30 seconds. The viscosity for the latter sample

decreased to 4.92 mPa*s at 60 seconds. The difference in viscosity values with respect to time in

the six-PEC-layer zinc microparticle solutions could be explained by agglomeration of the

particles. Figure 4-16C shows some aggregation of hexa-layer Zn particles. In static conditions,

particles' agglomeration was even higher, causing an initial higher resistance to the rotating

cone. As the sample fluid was exposed to shear rates of 200 sec-1, particles' agglomeration was

broken up, decreasing the solution viscosity.

The viscosity mean value of the ten-layer copper microparticle solution was 8.19 + 0.87,

representing a significant increment compared to the blood viscosity. This result correlates with

our previous stability and deformability studies. The ten multilayer copper-alginate systems

formed the strongest microparticles due to the alginate high affinity for copper ions and the

increased stability provided by the high number of coatings. Regarding the ten-layer zinc-

alginate microspheres, suspensions containing these particles showed an extremely high

viscosity value of 26.2 mPa*s for the initial 30 seconds. Although, viscosity decreased with time

to 8.85 mPa*s at 60 seconds, it was still about 2.7-fold higher than blood viscosity. Once more,

particles' agglomeration seemed be the cause of the changes in viscosity with respect to time.

Figure 4-16D shows large aggregates of deca-layer Zn particles.









Adsorption of each polyelectrolyte layer onto microcapsules was examined by measuring

the particle surface charge changes. We first examined uncoated alginate particles dispersed in

dH20. There was no significant difference between the zeta potentials of uncoated zinc- and

copper-alginate particles (Table 4-10). It is apparent from Figure 4-20 that the zeta potential of

microparticles with the chitosan surface layer (odd numbers) is slightly less negative than

microparticles with the alginate surface layer (even numbers). Although the data showed the

presence of each polyelectrolyte layer, the difference between uncoated and chitosan-surface

coated microparticles was not significant until the second chitosan coating for zinc particles and

the third coating for copper particles. Also, Figure 4-20A illustrates that the zeta potential of

zinc microspheres with a total of eight and ten PEC layers ending in alginate differed

significantly compared to uncoated microspheres.

It is evident that microbeads retained a negative surface charge throughout the

modification of the surface layer. Thickness of the polymeric films on colloidal templates is

considered to be in the monomolecular-layer range [110]. Since the low molecular weight (<

5000Da) chitosan oligosaccharide was used as the cationic polymer, layer thickness was very

thin and the polyelectrolyte molecules failed to overcharge the microbead surface. Therefore,

reversing of the microparticles' surface charge did not occur. As mentioned earlier,

agglomeration of coated zinc-alginate microspheres increased with the number of layers.

Eventually, the formation of aggregates resulted in an irregular coating of the multilayer zinc

particles.

Since the toxicity of copper ions was a concern, we decided to analyze the amount of

copper ions released from microparticles into the chelating medium. To determine the amount of

copper ions sequestered by albumin a modified bicinchoninic acid (BCA) protein assay was









used. Microparticles were suspended in RPMI 1640 medium, supplemented with 1%

penicillin/streptomycin and 5% BSA, at a concentration of 5% (w/v). Particles were incubated at

250C under continuous orbital rotation and supernatants were collected at predetermined

intervals. The removed solution was replaced with an equal volume of fresh supplemented

medium and samples were returned to the rotating incubator until the next time point.

Concentrations of albumin-reduced copper cations were determined by comparison to a standard

curve.

Figure 4-21 illustrates the amount of copper released in a period of three days. Only

concentrations of copper released from uncoated particles were detected. It was determined that

a 5% (w/v) suspension of uncoated particles released a total copper concentration of

approximately 3 mM in 24 hours (Table 4-11). Although no released copper from uncoated

particles was observed on the second day, some amount was detected on the third day. The

results obtained for the coated samples were in agreement with visual observations. At day 1,

microparticles with two coatings looked very swollen as opposed to six-layer particles, while

ten-layer beads looked unchanged. At day 2, some of the two- and six-layer particles were

disintegrated, whereas microparticles with ten coatings looked swollen. Since PEC coatings

improved the stability of particles, the ion exchange process took longer as the number of PEC

layers increased. Therefore, small amounts of copper ions were released into the medium during

each time point, which were not detected by the modified BCA assay due to its sensitivity at the

microgram level. Since the dose limit of copper intake levels in adults is 10 mg/day (157

tmol/day) and the prescribed IV dose for patients with copper deficiency anemia is 3 mg/day

(47.2 [tmol/day) [111], the low amounts of copper released from the 5% (w/v) suspension of

coated particles are still below toxic levels.









The cytotoxicity of the uncoated and coated calcium-, zinc-, and copper-crosslinked

particles was assessed using an in vitro cell proliferation assay. Human dermal fibroblast cells

were exposed to extracts from the particles and the MTT assay was used to determine cell

survival and recovery. Figure 4-22 shows histograms of the cell survival and recovery

percentages as a function of extract concentration. Absorbance readings from cells incubated in

2% FBS medium were used to calculate the survival and recovery percentages. Cells exposed to

the extracts from calcium particles showed up to a 2.5-fold increment in cell growth independent

of the coating number or the extract concentration (Table 4-12; Figure 4-22A). Likewise, Figure

4-22B demonstrates that a 24-hour exposure of calcium particle extracts did not alter cell

proliferation in the long term. While cells exposed to supernatants from the coated zinc particles

showed also up to a 2.5-fold increase in cell proliferation, cells exposed to uncoated particles'

extracts at the 100 and 1000 [tg/mL concentrations did not survive (Figure 4-22C). Similar to

calcium, exposure to zinc particle supernatants did not affect cell proliferation in the long term,

except for cells in contact with extracts from uncoated zinc particles at the two highest

concentrations (Figure 4-22D).

When cells were incubated in copper particle supernatants, a 2.5-fold increase in

proliferation was seen only for cells exposed to the 0.1 tg/ml concentration of coated particles'

extracts (Figure 4-22E). Although Table 4-12 illustrates a higher proliferation for cells in

contact with supernatants from coated particles as opposed to the control, it is apparent that cell

proliferation decreases with the extract concentration. In addition, cells exposed to the highest

concentration of uncoated particle supernatants did not survive. In contrast to cells incubated at

high doses of uncoated zinc particle extracts, cells exposed to toxic levels of uncoated copper

supernatants were able to slowly recover within two days (Figure 4-22F). The long-term









histogram shown in Figure 4-22F also demonstrates that proliferation of cells exposed to all the

other extracts was not compromised.

These results are in agreement with our previous observations for the particle stability and

the ion exchange process of our PEC chitosan-alginate particles. It has been shown repetitively

that particle stability is enhanced by the number of PEC coatings, except for calcium-crosslinked

particles. We have also demonstrated that the copper-alginate matrix is the strongest and,

therefore, forms the most stable particles. As a result, PEC layers not only keep the zinc and

copper particles stable but also decrease the release rate of crosslinking ions into the medium,

which explains differences in cell toxicity between the uncoated and coated particles. In

addition, higher toxicity is observed in uncoated zinc particles due to less alginate affinity for

this cation as opposed to Cu2+. It is apparent from the recovery histograms that the percent

values are lower, which coincide with the values obtained for cells incubated in 10% FBS

medium, our negative control (Table 4-12). These observations could be explained by the cells

reaching confluency, resulting in inhibition of cell proliferation due to cell-to-cell contact. Since

the MTT assay can only detect mitotic cells, the correct cell number of non-dividing cells cannot

be determined through this assay.

The use of copper as a crosslinker for alginate gels has always been limited due to its

toxicity. However, our results coincide with other research groups which have found that small

amount of copper induces cells proliferation, differentiation, and migration in vitro [112, 113,

114]. Hu was able to demonstrate that human umbilical vein endothelial cells incubated in 10%

FBS medium supplemented with 500 tM copper were able to increase in cell number by 216%

compared to a 248% increment induced by 20 ng/ml bFGF [112]. Also, Rodriguez et al showed

that copper at the 50 tM concentration decreased the proliferation rate of human mesenchymal









stem cells while increasing cell differentiation into osteogenic and adipogenic cell lines. Besides

proliferation and differentiation, migration of keratinocytes was induced by exposure of cells to

zinc or copper concentrations of 1.8 or 2 [tg/mL, respectively [114].

Conclusion

Particles made with air spray crosslinking technique presented a mono-dispersed size

distribution; however, the smaller particle size obtained was 50 [tm. Although, particle size

decreased with increments in the outer lumen air pressure, monodispersity was lost. Particles

made with W/O emulsion-gelation technique using cyclohexane as the organic phase presented a

narrow size distribution; diameter of uncoated particles ranged from 5-15 tm. Microparticles

crosslinked with copper presented a tear-drop shape as oppose to the spherical shape of particles

crosslinked with zinc and calcium. Particle size decreased when increasing the number of

coatings. After the tenth coating, there was approximately a 50% size reduction in particles with

an initial diameter ranging between 2 3 mm. Besides a decreased in size with multiple

coatings, particles changed their shape from spherical to biconcave. Zinc-crosslinked particles

showed a more pronounced biconcave configuration as opposed to calcium- and copper-

crosslinked particles.

After removing cross-linkers of coated microspheres, zinc- and copper-alginate capsules

were deformable and remained stable using the micropipette technique under physiological

pressures. Best results were obtained with Zn- and Cu-crosslinked alginate particles coated six

times. In addition, these capsules were stable for a period of up to 10 days in sequestrant-

containing media and sustained shear rates of 200 sec-1. Stability of the zinc- and copper-

alginate microparticles was improved by the number of polyelectrolyte coatings. The ten

multilayer copper-alginate systems formed the strongest particles due to the alginate high affinity

for copper ions and the increased stability provided by the high number of coatings. Calcium-









crosslinked particles were not stable after removal of crosslinking ions; even multiple

polyelectrolyte coatings did not improve stability.

Viscosity values for solutions containing copper microcapsules with two and six PEC

coatings were close to blood viscosity values; while only suspensions of two-layer zinc particles

showed viscosity values similar to blood's. Coating microbeads with chitosan oligosaccharide

did not affect the particles surface charge as they retained a negative surface charge throughout

the modification of the surface layer. From the ion exchange studies, it was determined that a

5% (w/v) suspension of uncoated copper particles released a total copper concentration of

approximately 3 mM in 24 hours. Copper released from the coated particles could not be

determined under the conditions tested. Results from the MTT assay demonstrated that multiple

coatings decreased toxicity of heavy-metal crosslinked particles. The highest level of toxicity

was shown by uncoated copper- and zinc-crosslinked particles at 1000 [tg/ml. The rest of the

samples seemed not to affect cells under the conditions studied.









Table 4-1. The PEC chitosan-alginate particle samples made
Batch # Crosslinker
1 Ca
2 Ca
3 Ca
4 Ca
5 Zn
6 Zn
7 Zn
8 Zn
9 Cu
10 Cu
11 Cu
12 Cu


# of Coatings
0
2
6
10
0
2
6
10
0
2
6
10


Table 4-2. Concentrations of major electrolytes (mmol/L) in human plasma and cell culture
media
Electrolyte Plasma RPMI 1640 DMEM/F12 50:50
Bicarbonate 24 30 23.81 29.02
Calcium 2.00 2.75 0.42 2.0
Chloride 100 110 108.03 127.96
Magnesium 0.8 1.1 0.4 0.7
Phosphate 0.53 0.90 5.64 0.95
Potassium 4.0 5.6 5.36 4.18
Sodium 130- 155 113.95 150.25
Sulfate 0.35 0.75 0.4 0.703
Nitrate --- 0.84 0.00037












800
700
E600
500
.2 400
300
200
100
0-
30G 22G 18G 16G
***p <0.001 Needle Gauge


Figure 4-1. Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of the needle gauge. Particles were made
with an alginate extrusion rate of 12 mL/h, a 10-cm dropping distance, and the outer
lumen air pressure set at 50 psi. Significant differences between particles made with
the 30G needle and all the other needle gauges are reported.






Table 4-3. Particle size diameter of copper-crosslinked alginate particles as a function of the
needle gauge
Particle Diameter (tm)
Needle Needle Inner
Gauge Diameter (tm) Mean SEM
30G 330 85.37 4.39
22G 711 132.31 8.19
18G 1270 482.96 42.78
16G 1651 688.55 125.51













100 -

80 -
N
c 60

40
C=
20 -

0
1.2 6 12
p <0.05 Alginate Flow Rate (mL/h)
***p < 0.001

Figure 4-2. Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of alginate flow rate. Particles were made
using a needle gauge of 30G (330 [tm), a 10-cm dropping distance, and the outer
lumen air pressure set at 50 psi. Significant differences between particles made with
1.2 and 12 mL/h and between 6 and 12 mL/h are reported.






Table 4-4. Particle size diameter of copper-crosslinked alginate particles as a function of
alginate flow rate
Particle Diameter ([tm)
Alginate Flow Rate
(mL/h) Mean SEM
1.2 85.83 3.22
6.0 93.22 3.17
12.0 105.32 3.61























A B


Figure 4-3. Optical photomicrographs of copper-crosslinked alginate particles made with the air-
spray crosslinking technique using A) a 1.65-mm (16G) syringe needle and extruding
flow rate of 12 ml/hr, and B) a 0.33-mm (30G) syringe needle and extruding flow rate
of 1.2 ml/hr. The dropping distance set at 10 cm and the outer lumen air pressure set
at 50 psi. Original magnification = x50, bar denotes 200 rm.











100
90
80
E 70 *
60 .1.2 ml/hr
S50 .12 ml/hr
40 40
a30
20
10
0
p <0.05 40 psi 60 psi
p <0.05
*** p <0.001 Air Pressure

Figure 4-4. Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of air pressure. Particles were made using
a needle gauge of 30G (330 [tm) and a 10-cm dropping distance. While testing air
pressure or alginate flow rate, all other parameters remained constant.








Table 4-5. Particle size diameter of copper-crosslinked alginate particles as a function of
alginate flow rate
Particle Diameter (tm)
Air Pressure Alginate Flow Rate
(psi) (mL/h) Mean SEM
40 1.2 63.58 4.78
40 12.0 81.14 6.16
60 1.2 47.63 4.04
60 12.0 44.50 3.40












.79


Si

i*1*


U


A B
















C D


Figure 4-5. Optical photomicrographs of copper-crosslinked alginate particles made with the air-
spray crosslinking technique: A) air pressure = 40 psi and alginate flow rate = 1.2
mL/h; B) air pressure = 40 psi and alginate flow rate = 12 mL/h; C) air pressure = 60
psi and alginate flow rate = 1.2 mL/h; and D) air pressure = 60 psi and alginate flow
rate = 12 mL/h. Original magnification = x50, bar denotes 200 im.


,c. ,













70

60
E
S50
N
c 40
.2 30
.2 20

10
0 -

***p < 0.001


Copper


Zinc


Figure 4-6. Particle size diameter of copper-crosslinked alginate particles made with the air-
spray crosslinking technique as a function of the crosslinking ion. Particles were
made using a needle gauge of 30G (330 [tm), a 10-cm dropping distance, the outer
lumen air pressure set at 50 psi, and an alginate extrusion rate of 1.2 mL/h.






Table 4-6. Particle size diameter of copper-crosslinked alginate particles as a function of the
crosslinking ion


Crosslinking Ion
Copper
Zinc


Particle Diameter ([tm)
Mean SEM
47.63 4.04
71.33 4.41
























A B


Figure 4-7. Optical photomicrographs of calcium-crosslinked alginate particles made with the
air-spray crosslinking technique A) and with the W/O emulsion gelation technique
B). A) Original magnification = x50, bar denotes 500 im. B) Original magnification
= xl00, bar denotes 100 im.










































Figure 4-8. Optical micrographs of alginate microspheres made with the W/O emulsion gelation
technique. Alginate particles were crosslinked with (A, B) calcium, (C, D) zinc, or
(E, F) copper ions. The left row original magnification = xl00, bars denote 100 rim;
right row magnification = x200, bars denote 50 im.












100
90
80
i70
> 60
C 50
S40
1 30
20
10
0


0 5 10 15 20 25 30 35 40
Particle Diameter (um)


0 5 10 15 20 25 30 35 40
Particle Diameter (um)


100
90
80
70
> 60
C 50
S40
L 30
20
10
0


-Ca_0_1
-Ca_0_2
-Ca_0_3






A









-Zn_0_1
-Zn_0_2
-Zn_0_3






B









-Cu 0_1
-Cu_0_2


0 5 10 15 20 25 30 35 40
Particle Diameter (um) C



Figure 4-9. Particle size distribution. Percentage of total number of uncoated microspheres
versus particle diameter: A) calcium-, B) zinc-, and C) copper-crosslinked alginate
particles. Measurements carried out by a Beckman LS Particle Characterization
Coulter.




Full Text

PAGE 1

1 DEFORMABLE MICROPARTI CLES WITH MULTIPLE FUNCTIONS FOR DRUG DELIVERY AND DEVICE TESTING By TAILI T. THULA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Taili T. Thula

PAGE 3

3 To my parents for their unconditional love

PAGE 4

4 ACKNOWLEDGMENTS I thank God for giving me the opportunity to accomplish my goals a nd to meet wonderful people who supported and helped me. I thank my parents for their support and endless love. I thank Asdrubal for always being there for me, es pecially when I most needed support. I thank my little brother, Jeancho, for all his help counting endless batches of particles. Also, I thank my two other musketeers: Jompo and John. You guys made it a fun journey and I could not have done without you. Finally, I thank my professors (Dr. Christopher Batich, Dr. Gregory Schultz, Dr. Roger Tran-Son-Tay and Dr. Malisa Sarntinora nont) for all of their su pport throughout this lengthy project. I thank them for sharing thei r knowledge and for having faith in me.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............13 CHAPTER 1. INTRODUCTION..................................................................................................................15 2. BACKGROUND....................................................................................................................19 Blood: Composition, Properties, and Functions.....................................................................19 Composition....................................................................................................................19 Properties..................................................................................................................... ....20 Functions...................................................................................................................... ...20 Red Blood Cell: Morphology, Functions, Rheology, and Hematopoiesis.............................21 Morphology and Functions..............................................................................................21 Rheological Properties.....................................................................................................23 Hematopoiesis.................................................................................................................23 Red Blood Cell as a Model for Drug Delivery................................................................25 Encapsulation Methods...................................................................................................25 Advantages and Disadvantages of Red Blood Cells in Drug Delivery...........................26 Drug Delivery Systems.......................................................................................................... .27 Microspheres and Microcapsules....................................................................................28 Microencapsulation Techniques......................................................................................29 Immune System Response to Foreign Elements.............................................................32 3. CHARACTERIZATION OF THE DEFORMABLE BEHAVIOR OF PLURONICPLA AND POLYELECTROLYTE CHITOS AN-ALGINATE MICROPARTICLES.........40 Introduction................................................................................................................... ..........40 Poly(Lactic Acid)............................................................................................................41 Pluronics..................................................................................................................... ..42 Alginate....................................................................................................................... ....43 Chitosan....................................................................................................................... ....45 Materials and Methods.......................................................................................................... .46 Preparation of PLA-Pluronic Particles.........................................................................46 Preparation of Algi nate Particles.....................................................................................46 Coating of Alginate Particles...........................................................................................47 Decomposition of the Particles Core..............................................................................48 Measurement of Particles Deformability.......................................................................48

PAGE 6

6 Scanning Electron Microscopy (SEM) Analysis.............................................................49 Results and Discussion......................................................................................................... ..49 PLA-Pluronic Particles.................................................................................................49 Chitosan-Alginate Particles.............................................................................................52 Conclusion..................................................................................................................... .........54 4. SYNTHESIS AND CHARACTERIZATION OF PHYSICO-CHEMICAL AND RHEOLOGY PROPERTIES OF POLY ELECTROLYTE CHIT OSAN-ALGINATE MICROPARTICLES CROSSLINKED WITH CALCIUM, ZINC, OR COPPER IONS.....64 Introduction................................................................................................................... ..........64 Materials and Methods.......................................................................................................... .65 Preparation of PEC Particles...........................................................................................65 Air-spray crosslinking technique.............................................................................65 W/O emulsion gelation technique............................................................................66 Coating of alginate particles.....................................................................................67 Decomposition of the particles core.......................................................................67 Characterization of PEC Particles...................................................................................68 Surface morphology analysis...................................................................................68 Particle size distribution...........................................................................................68 Particle stability analysis..........................................................................................69 Measurement of particles deformability.................................................................70 Assessment of particles viscosity............................................................................70 Coating adsorption analysis.....................................................................................71 Ion exchange analysis...............................................................................................72 Assessment of particle cytotoxicity..........................................................................72 Data Interpretation...........................................................................................................74 Results and Discussion......................................................................................................... ..74 Conclusion..................................................................................................................... .........89 5. ENCAPSULATION A ND RELEASE OF BOVINE SERUM ALBUMIN FROM POLYELECTROLYTE CHITOSAN -ALGINATE MICROPARTICLES CROSSLINKED WITH ZI NC OR COPPER IONS............................................................120 Introduction................................................................................................................... ........120 Materials and Methods.........................................................................................................121 Preparation of FITC-Labeled BSA (F-BSA).................................................................121 Micro-Core Preparation.................................................................................................122 Coating of Alginate Particles.........................................................................................122 Characterization of F-BSA-Loaded Microcapsules......................................................123 Surface morphology analysis.................................................................................123 Particle size distribution.........................................................................................123 Coating adsorption analysis...................................................................................124 Determination of protein loading efficiency..........................................................124 In vitro release kinetics of F-BSA microcapsules..................................................125 Bovine serum albumin ELISA...............................................................................125 Data Interpretation.........................................................................................................126

PAGE 7

7 Results and Discussion.........................................................................................................126 Conclusion..................................................................................................................... .......132 6. CONCLUSION AND FUTURE WORK.............................................................................144 LIST OF REFERENCES.............................................................................................................151 BIOGRAPHICAL SKETCH.......................................................................................................161

PAGE 8

8 LIST OF TABLES Table page 2-1 Plasma protein con centrations (mg/100 mL).....................................................................35 2-2 Physicochemical properties of the major plasma proteins.................................................36 2-3 Concentrations of major electrolyt es (mEq/L) in whole blood and plasma......................37 2-4 Concentrations of various organic compounds (mg/100mL) in whole blood and plasma......................................................................................................................... .......37 2-5 Controlled release drug delivery systems..........................................................................39 2-6 Examples of commercialized PLGA copolymers microspheres.......................................39 3-1 List of commercially available Pluronics.......................................................................56 3-2 List of Pluronic-PLA particles made..............................................................................58 3-3 List of chitosan-al ginate particles made............................................................................59 3-4 Composition and deformability properti es of Pluronic-PLA microspheres...................61 3-5 Composition and deformability properties of PEC chitosan-alginate microcapsules.......63 4-1 The PEC chitosan-alginate particle samples made............................................................91 4-2 Concentrations of major electrolytes (mmol/L) in human plasma and cell culture media.......................................................................................................................... ........91 4-3 Particle size diameter of copper-crossli nked alginate particles as a function of the needle gauge................................................................................................................... ....92 4-4 Particle size diameter of copper-cross linked alginate particles as a function of alginate flow rate............................................................................................................. ...93 4-5 Particle size diameter of copper-cross linked alginate particles as a function of alginate flow rate............................................................................................................. ...95 4-6 Particle size diameter of copper-crossli nked alginate particles as a function of the crosslinking ion............................................................................................................... ...97 4-7 Particle size distribution of uncoated Ca-, Zn-, and Cu-cro sslinked alginate particles...101 4-8 Composition and deformability properties of PEC chitosan-alginate microcapsules.....110 4-9 Viscosity values of PEC ch itosan-alginate microcapsules..............................................113

PAGE 9

9 4-10 Zeta potential values of PEC chitosan-alginate microcapsules.......................................115 4-11 Concentration values of copper ions released from PEC chitosan-alginate microcapsules.................................................................................................................. .116 4-12 Fibroblast cell survival and recovery percentages as a function of extract concentrations from PEC microparticles.........................................................................118 5-1 The PEC chitosan-alginate particle samples made..........................................................133 5-2 Zeta potential values of PEC F-BSAloaded chitosan-alginate microcapsules...............140 5-3 Encapsulation efficiency of F-BSA loaded microparticles prepared from zincand copper-crosslinked alginate gels......................................................................................142 5-4 Release kinetic profile of F-BSA from uncoated and multiple PEC coated zincand copper-alginate microspheres during a 48-hour study.....................................................143

PAGE 10

10 LIST OF FIGURES Figure page 2-1 Scanning electron micrograph of red blood cell aggregates, rouleaux..............................38 2-2 Shear rate dependence of normal human blood viscoelasticity.........................................38 3-1 PLA molecular structure....................................................................................................56 3-2 Alginate molecular structure..............................................................................................57 3-3 Chitosan molecular structure.............................................................................................. 57 3-4 Optical micrograph of 20% Pluronic-PLA (1:2) microspheres......................................60 3-5 Optical micrograph of 20% P105-PLA (1:2) microspheres..............................................61 3-6 Scanning electron micrograph of 20% P105-PLA Particles.............................................61 3-7 Optical and scanning elec tron micrographs of 0.3% COS Ca-crosslinked alginate particles...................................................................................................................... ........62 3-8 Optical and scanning el ectron micrograph of 3% COS Cu-crosslinked alginate particles, 5 coatings.......................................................................................................... ..62 4-1 Particle size diameter of copper-crossli nked alginate particles made with the airspray crosslinking technique as a function of the needle gauge........................................92 4-2 Particle size diameter of copper-crossli nked alginate particles made with the airspray crosslinking technique as a f unction of alginate flow rate.......................................93 4-3 Optical photomicrographs of copper-crossli nked alginate particle s made with the airspray crosslinking technique..............................................................................................94 4-4 Particle size diameter of copper-crossli nked alginate particles made with the airspray crosslinking technique as a function of air pressure................................................95 4-5 Optical photomicrographs of copper-crossli nked alginate particle s made with the airspray crosslinking technique..............................................................................................96 4-6 Particle size diameter of copper-crossli nked alginate particles made with the airspray crosslinking technique as a f unction of the crosslinking ion...................................97 4-7 Optical photomicrographs of calcium-cross linked alginate particles made with the air-spray crosslinking technique........................................................................................98 4-8 Optical micrographs of alginate micros pheres made with the W/O emulsion gelation technique...................................................................................................................... ......99

PAGE 11

11 4-9 Particle size distributi on. Percentage of total num ber of uncoated microspheres versus particle diameter...................................................................................................100 4-10 Optical photographs of Ca-, Zn-, and Cu-c rosslinked alginate particles made with the air-spray crosslinking technique......................................................................................102 4-11 Scanning electron micrographs of algi nate microspheres made with the W/O emulsion gelation technique............................................................................................103 4-12 Particle size distribution of uncoa ted and multiple-layer PEC particles..........................104 4-13 Stability studies of PEC microparticles...........................................................................105 4-14 Optical photomicrographs of bilayer coppe r-crosslinked alginate particles incubated 24 hours....................................................................................................................... .....106 4-15 Optical photographs of uncoated zi nc-crosslinked algi nate particles..............................107 4-16 Optical micrographs of zinc-crosslinked PEC microparticles exposed to a shear rate of 200-sec-1 for one minute..............................................................................................108 4-17 Optical micrographs of copper-crosslinke d PEC microparticles exposed to a shear rate of 200-sec-1 for one minute.......................................................................................109 4-18 Optical photomicrograph of hexa-layer al ginate-chitosan capsule during micropipette aspiration..................................................................................................................... .....111 4-19 Viscosity values of zincand copper-crosslinked PEC microparticles...........................112 4-20 Zeta potential values of A) zincan d B) copper-crosslinked multiple layer PEC particles...................................................................................................................... ......114 4-21 Concentration of copper ions released from chitosan-alginate PEC microparticles........116 4-22 Survival and recovery percenta ges of human dermal fibroblasts....................................117 5-1 Optical micrographs of blank zinc-alg inate microparticles made with the W/O emulsion gelation technique............................................................................................134 5-2 Optical micrographs of blank copper-al ginate microspheres made with the W/O emulsion gelation technique............................................................................................135 5-3 Optical micrographs of F-BSA-loaded al ginate microparticles coated with ten PEC layers......................................................................................................................... .......136 5-4 Scanning electron micrographs of zinc-cro sslinked alginate microparticles made with the W/O emulsion gelation technique..............................................................................137

PAGE 12

12 5-5 Scanning electron micrographs of copper-c rosslinked alginate microparticles made with the W/O emulsion gelation technique......................................................................138 5-6 Zeta potential values of A) zincan d B) copper-crosslinked multiple layer PEC particles encapsulating F-BSA.........................................................................................139 5-7 Encapsulation efficiency of zincand c oppercrosslinked alginate microparticles........141 5-8 In vitro release of F-BSA from uncoated and multiple PEC coated zincand copperalginate microspheres over a 48-hour period...................................................................142

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEFORMABLE MICROPARTI CLES WITH MULTIPLE FUNCTIONS FOR DRUG DELIVERY AND DEVICE TESTING By Taili T. Thula May 2007 Chair: Christopher Batich Major: Biomedical Engineering Since the HIV epidemic of the 1990s, researchers have attempted to develop a red blood cell analog. Even though some of these substitutes are now in Phas e III of clinical trials, their use is limited by side effects and short half-life in the human body. As a result, there is still a need for an effective erythrocyte analog with mi nimum immunogenic and side effects, so that it can be used for multiple applications. Finding new approaches to develop more efficient blood substitutes will not only bring valuable advances in the clinical approach, but also in the area of in vitro testing of medical devices. We examined the feasibility of creating a deformable multi-functional, biodegradable, biocompatible particle for applications in drug delivery and device testing. As a preliminary evaluation, we synthesized different types of microcapsules using natural and synthetic polymers, various cross-linking agents, and diverse manufacturing techniques. After fully characterizing of each system, we determined the most promising red blood cell analog in terms of deformability, stability and toxicity. We also examined the encapsulation and release of bovine serum albumin (BSA) within these deformable particles. After removal of cross-linkers, zincand copper-alginate microparticles surrounded by multiple polyelectrolyte layers of chitosan ol igosaccharide and alginate were deformable and

PAGE 14

14 remained stable under physiological pressures applied by the micropipe tte technique. In addition, multiple coatings decreased toxicity of heavy-metal crosslinked particles. BSA encapsulation and release from chitosan-algi nate microspheres were contingent on the crosslinker and number of polyel ectrolyte coatings, respectively. Further rheological studies are needed to determine how closely these particles simulate the behavior of erythrocytes. Also, studies on the encapsulation and re lease of different proteins, in cluding hemoglobin, are needed to establish the desired controlled release of bi oactive agents for the proposed delivery system.

PAGE 15

15 CHAPTER 1 INTRODUCTION For the past century there has been an intere st in blood substitutes. However, it was not until the mid 1980s that efforts to find a commer cially available product scaled up. Numerous groups tried to develop a red bl ood cell analog due to the HIV epidemic of the time [1, 2]. Another reason that drove many researchers in to the area was US Army concern about blood supplies in the battlefield [1, 3]. Large investme nts were made by the military in search for the ideal red blood cell analog. This ideal blood analog woul d carry oxygen, provide rapid expansion of the blood space, eliminate infec tious risks of transfusion, have universal compatibility, and be available for immediate us e without need for special storage [4]. However, about 10 years, all efforts and s upport were dropped since the developed analogs did not meet the desired requireme nts [1]. Nowadays, some of t hose developed blood substitutes are commercially available or in Phase III of clinical trials. Although many of them can fulfill essential functions of transfused blood, providing expansion of the plasma volume and carrying oxygen; their use is still limited due to side e ffects and short half-life time within the human body [4]. The blood substitutes that have been studied more extensively can be grouped into two classes: modified hemoglobin solutions and perfluorocarbon emulsions. The native human hemoglobin molecule has a number of advantages as an oxygen carrier, including high capacity for oxygen, the lack of antigens af ter purification, a prolonged sh elf life, and the ability to withstand harsh purification proced ures [2, 4]. However, it needs to be modified in order to decrease its oxygen affinity and to pr event rapid dissociation of the native 22 tetramer into yperbo which are very toxic [ 4, 2 ]. Modifications of the hem oglobin molecule have included interand intra-molecular cross-linking, conjugati on to polymers, and more recently lipid or

PAGE 16

16 polymeric encapsulation. Even though modificat ion of hemoglobin prevents toxicity due to breakdown of the molecule, other side effects ha ve been observed with all modified hemoglobin solutions. The main side effect, due to nitric oxide scavenging, is vasoconstriction resulting in an increase in the mean arterial blood pressu re and decrease in the cardiac index [4]. Perfluorocarbons have a molecular structure ve ry similar to hydrocar bons, differing in the replacement of hydrogen by fluorine atoms and their electron affinity [5]. Perfluorocarbons are excellent carriers of oxygen and carbon dioxide; in addition, they can be mass-produced fairly easily and source-independently. However, since th ey are not miscible in aqueous systems, they have to be prepared as emulsions which leads to unwanted side effects due to surfactants [5, 6]. Complement activation is the major problem with perfluorocarbon-based products. Patients present flu-like symptoms, including increased body temperatures and decr eased platelet counts [ 6, 5 ]. In addition to side effect s, perfluorocarbon emulsions ha ve shown poor efficacy and very short half-life within the bloodstream [6]. New strategies to develop more efficient bl ood substitutes are necessary since all current approaches lack effectiveness and patients norma lly depend on donated blood as the only source for transfusion. Upcoming new potential blood-borne pathogens are always a concern regarding blood transfusions; even though t oday that blood supply is much safer in developed countries due to improved screenings [3, 5, 7]. Another ongoing problem is shortage of blood reserve due to an annual increase of blood demand at a much faster pace than blood donations [5, 8]. In addition to aging, worldwide population has been affected in the recent years by an increased number of natural disasters, acts of terrorism, and civil and internationa l conflicts, leading to a steep increment in blood demand. Indeed, accord ing to the American Association of Blood Banks, a shortage of packed red blood cells is estimated by 2030 [3]. Other limitations to blood

PAGE 17

17 collected for transfusion incl ude: recipient immune suppressi on, short-term storage, and decreased oxygenation capabilities [7]. The use of artificial erythroc ytes could solve all the problems related to traditional transfused blood. In addition to carrying oxygen, the new generation of erythrocyte analogs could also be used as controlled-release delive ry systems for the intravenous administration of therapeutic agents. This new cl ass of therapeutics should meet the following criteria: be nonimmunogenic, have long-term storag e stability and an intravascular half life of weeks to months, have no significant side effects, and be simple to manufacture and steril ize at permissible widescale production costs. Finding new approaches for the development of more efficient blood substitutes will not only bring valuable advances in the clin ical approach, but al so in the area of in vitro testing of medical devices and products. Certainly, ma ny medical equipment and products have been engineered with the purpose to interact directly w ith blood, e.g., dialysis machines, ventricular assist devices, heart-lung mach ines, heart valves, catheters, graf ts, stents, among others. Since blood will be in direct contact and, sometimes, even subject to mechanical stress with these types of devices and products, it is necessa ry to test their effect on blood in vitro before moving on to clinical trials. However, in vitro testing with real blood carries a whole set of problems such as cost, increment in rules and re gulations, availability, variability among donors, and inconsistency with storage time. As a result, particles with deformability properties and dispersed in fluids with viscosities similar to blood will represent an invaluable innovation for the testing of devices and products. This study proposes to examine the feas ibility of creating a multi-functional, biodegradable, biocompatible particle with deformable properties for applications in in-vitro

PAGE 18

18 testing of medical devices and products, and in drug delivery. The proposed method is based on the hypothesis that: (1) polymeric microspheres with a highly porous inner structure or microcapsules with a dissolvable gel core can be designed and synthesized so that they attain appropriate deformability properties, and (2) th at the ability of encapsulating and releasing proteins within these microparticles, while stil l keeping some deformability capability, will allow practical use. The aim of this dissertation is to conduct feasibility stud ies on the synthesis and preliminary testing of different types of microcapsules using natural and synthetic polymers, various cross-linking agents and solvents, and diverse manufacturing techniques. The aim will also include discerning the most promising red blood cell analog, in terms of deformable capability, to eventually be tested in an in vivo study. The deformability property of the microcapsules will be based on the well-establi shed micropipette aspiration technique. In addition to finding microcapsules with corr ect deformability properties, another aim is to develop a novel drug delivery system by using thes e deformable particles. The focus of this second aim is on examining the delivery system f easibility: encapsulation, st orage, and release of proteins from microcapsules wit hout affecting the protein bioactivity and the particle elastic property. All components of the system will be designed to be non-toxic, and use materials already approved for human contact. They will be combined in a novel way to test the hypothesis through two specific aims. Each aim will have defined thresholds for measuring acceptable outcomes, and if successful, will demons trate the feasibility of developing a simple technology for the manufacturing of a red blood cell analog for in vitro testing of medical devices and products, and as a drug delivery system.

PAGE 19

19 CHAPTER 2 BACKGROUND Blood: Composition, Properties, and Functions Composition Blood is a specialized type of connective tissue composed of a suspension of living cells in an aqueous solution of electrolytes and non-electrolytes called plasma [ 9, 10 ]. The cellular portion of blood is composed of erythrocytes, leu kocytes, and platelets. Erythrocytes normally composed about 45 percent of the total blood volu me; this percentage is known as hematocrit [ 10 ]. In adults, the hematocrit ranges from 0.38 to 0.54: in females is slightly lower compared to males, ranging between 0.38 0.46 and 0.42 0.54 respectively [ 11 ]. Conversely, not even one percent of the total blood volume is constituted of leukocytes and platelets, while the remaining volume (about 55%) is made up by plasma [ 10 ]. Blood plasma is a straw-colored fl uid composed of about 90% (w) H2O, 8% (w/v) proteins, and 2% (w/v) inorga nic and organic substances [ 9, 10 ]. Plasma makes approximately 4% of an individuals total body weight (40 to 45ml/Kg) [ 11 ]. The majority of proteins found in plasma are produced in the liver and this makes about 7% of total plasma solutes (6.5 to 8g/dl) [ 11 ]. The three main plasma proteins include: albumin, accounting for 60%; globulins; and fibrinogen (Tables 2-1 and 2-2) [ 10, 11, 12 ]. Most of the blood func tions are actually performed by plasma proteins. The rest of plasma solutes include nutrients, gases, hormones, various cell metabolic waste products, and ions [ 10 ]. Table 2-3 shows a list of all the electrolytes present in plasma whose normal concentration is essentia l for nerve conduction, muscle contraction, blood clotting, fluid balance an d acid-base regulation [ 11, 12 ]. Other organic compounds found in plasma are listed in Table 2-4 [ 12 ].

PAGE 20

20 Properties Although, the total amount of blood varies wi th age, sex, weight, and body build among other factors; on average, blood accounts for a pproximately eight perc ent of the total body weight. Also, the normal blood volume in healthy adu lt males is 5 6 liters whereas it is slightly less for adult females (4 5 liters) [ 11 ]. The pH of blood ranges between 7.35 and 7.45 and its temperature of 38C (100.4F) is a li ttle higher than th e body temperature [ 10 ]. Blood is denser than water and has a viscosity approximately five times higher [ 10 ]. Although plasma is considered a Newtonian viscous fluid, having a li near relationship between the shear stress and strain rate; whole bl ood is considered non-Newtonian, showing both viscous and elastic properties [ 9, 13, 14 ]. Thus, the non-Newtonian property of blood is acquired from the cellular portion, and more specifically from the red blood cel ls or erythrocytes. The viscoelastic property of blood is affected by blood compositional parame ters, such as hematocrit and certain plasma proteins. Other factors affecti ng blood viscoelasticity include changes in osmotic pressure, pH and temperature; as well as administration of blood volume expanders a nd pathologies such as myocardial infarction, peripheral vascul ar disease, cancer and diabetes [ 9, 15, 16, 17 ]. Functions Blood performs several functions which can all be classified in three categories: distribution, regulation, and prot ection. Blood transports oxygen and nutrients to all cells, metabolic waste products to elimination si tes, and hormones to target organs [ 10, 3, 11 ]. Diffusion and partial pressure are fundamental processes involved in the transport and exchange of oxygen and carbon dioxide [ 11 ]. While the oxygen partial pressure (pO2) in the alveoli is approximately 100 mm Hg pulmonary capillaries have a pO2 of 40 mm Hg [ 10, 11 ]; resulting in a 60-mm-Hg diffusion gr adient in favor of pulmonary capillaries. On the contrary, systemic capillaries have a pO2 of 100 mm Hg as oppos ed to 40 mm Hg pO2 in tissues. In this

PAGE 21

21 case, the 60-mm-Hg diffusion gradient is in favo r of tissues, resulting in oxygen diffusion from systemic capillaries to tissues. In addition to distribution of substances, bl ood is in charge of maintaining appropriate body temperatures, pH, osmotic pressure and fluid volume in the circulatory system [ 11, 10 ]. Blood disseminates heat throughout the body and deals with the excess by transporting it to the skin surface. Regarding normal pH maintenan ce, many blood proteins act as buffers providing an alkaline reserve of bicarbonate atoms [ 10 ]. Besides prevention of sudden pH changes, blood proteins in conjunction with platelets also help stopping ex cessive fluid loss from the bloodstream, maintaining an optimal fluid volume in the circ ulatory system. The protecting action of blood includes prevention of blood lo ss and infections. Antibodies, complement proteins, and white blood cells circulating in the blood help defend the body against foreign aggressors [ 10 ]. Red Blood Cell: Morphology, Functi ons, Rheology, and Hematopoiesis Morphology and Functions Red blood cells (RBCs), or eryt hrocytes, are fully differen tiated, anucleated, non-dividing cells present in normal blood at high concentrat ions, with a hematocrit ratio (cell volume/blood volume) of approximately 0.45 in large vessels and 0.25 in small arterioles or venules [ 9, 10 ]. Matured erythrocytes are bound by plasma membrane and have essentially no organelles. They are composed of approximately 97% hemoglobin, not including water, and they have other proteins whose mainly function is to maintain the plasma membrane or support changes in the RBC shape [ 11, 10 ]. The biconcave shape of erythrocyt es is sustained by a net of fibrous proteins (mainly spectrin) which is deformable, giving erythrocytes enough flexibility to change shape as needed [ 10, 9 ].

PAGE 22

22 The red cells structural characteristics cont ribute to its extraordinary flexibility and respiratory functions. Its fibrous-protein memb rane is unstressed in the normal biconcave configuration and large deforma tions of the cell can occur wit hout stretching of the membrane and without any change in the pressure differe ntial between th e interior and the exterior [ 10, 11, 9 ]. In addition to the RBC deformable membrane, its small size and biconcave disk shape, with a diameter of approximately 7.5 m and a thickness of 2 m, provide a vast surface area ideal for gas exchange [ 11, 9, 10 ]. Due to the lack of mitochondria red blood cells generate ATP by an anaerobic mechanism; as a result, they do not use any of the oxyge n they transport, making them very efficient oxygen carriers [ 10 ]. As mentioned earlier, most of the erythroc ytes content is com posed of hemoglobin. Hemoglobin (Hb) exists in a tetramer ic form, consisting of two alpha ( ) and two beta ( ) polypeptide chains each bound to a ringlike heme group [ 18, 11 ]. Each heme group carries an atom of iron which can combine reversibly wi th one molecule of oxygen. Since a single red blood cell contains about 250 million hemoglobin mo lecules, each cell can carry about 1 billion molecules of oxygen [ 18 ]. Erythrocytes provide a protective envir onment for hemoglobin, preventing it from breaking down into yperbo which adversely a ffects the kidneys, blood viscosity, and osmotic pressure [ 18, 11 ]. In addition, erythrocytes protec t the Hb molecule from undergoing an unhindered oxidation process of its ir on center, resulting in the transi tion of Hb from the ferrous (HbFe2+) functional to the ferric nonfunctional form (HbFe3+) [ 19 ]. Along with the oxidative process, damaging and toxic species can form, including the ferryl protein (HbFe4+) which can peroxidize lipids, degrade carbohydra tes, and cross-link proteins [ 19 ]. Indeed, the red blood cell

PAGE 23

23 offers a reductase system rich in the enzymes catalase and superoxide dismutase which takes care of the spontaneous oxida tion of the ferrous iron [ 11, 19, 18, 1 ]. Rheological Properties Erythrocytes are the major factor contributi ng to blood viscosity. G. B. Thurston was the first to show that human blood exhi bits elasticity and viscosity [ 20 ]. While blood viscoelasticity depends on the elastic behavior of erythrocytes blood rheology is governed by cell aggregation, flow-induced cell organization and deformability [ 21 ]. One of the defining characteristic of RBCs is that they form aggregates known as rouleaux (Figure 2-1) [ 22 ], depending on the presence of certain plasma prot eins and on blood flow rates [ 9 ]. These aggregates are space efficient since, at normal hematocrit levels, the available plasma space is extremely limited for free cell motion without deformation. At very sl ow blood flows, shear rates on the cells are very small and human blood becomes a big aggregate with the properties of a viscoelastic solid [ 9 ]. Although increasing blood flows will break up th e rouleaux and reduce blood viscosity, further rearrangement of erythrocytes is needed to optimize the plasma space for cell motion. RBC deformation becomes important in reducin g blood viscosity even further at shear rates greater than 100 s-1 (Figure 2-2) [ 9, 23 ]. From a mechanical vi ewpoint, the red blood cell is composed of an elastic membrane surrounding an incompressible Newtonian viscous fluid [ 24 ]. Since it lacks a nucleus and organe lles, the intracellular matrix is considered as a protein rich, low viscous solution which facilitates deformation [ 25 ]. As a result, the cell can undergo an unlimited number of large deformations wit hout changing its volume, surface area, and stretching or tearing of the membrane [ 24 ]. Hematopoiesis Hematopoiesis refers to the process of bl ood cell formation which occurs in the bone marrow [ 10 ]. This process is specific to each t ype of blood cell and depends on the body needs

PAGE 24

24 and different regulatory factors [ 10 ]. All of the cellular elements in blood arise from the pleuripotent hematopoietic stem cell or hemocytoblast [ 11 ]. Once a cell is committed to a specific blood cell type by the appearance of membrane surface receptors, its maturation pathway is unique to the cell type [ 10, 11 ]. As cells mature, they migrate through the thin walls of the sinusoids to enter the blood, resulting in an average of one ounce of new blood produced daily [ 10 ]. Erythrocyte production, known as erythropoiesis, involves three distinct phases: preparation for hemoglobin production, hemoglob in synthesis and accumulation, and nucleus ejection [ 10 ]. During the first two phases, ce lls produce large amounts of ribosomes (proerythroblast to erythrobl ast) and divide many times (e rythroblast to normoblast) [ 10 ]. After the cell reaches a hemoglobin concentration of a pproximately 35%, its nuclear function ends and its nucleus degenerates and is ejected, causing the cell to co llapse inward and adopt the biconcave shape (reticulocyte) [ 10 ]. Reticulocytes still cont ain a small amount of clumped ribosomes and rough endoplasmic reticulum which are degraded later by intracellular enzymes [ 10 ]. The entire process from hemocytoblast to erythrocyte takes approximately five to seven days [ 10, 11 ]. In healthy people, new erythrocytes are produced at a rapid rate of more that 2 million per second [ 11 ]. Erythropoiesis is controlled hormona lly and depends on appr opriate supplies of iron and certain B vitamins; however, its direct stimulus is provided by erythropoietin (EPO) [ 11, 10 ]. EPO stimulates stem cells in the bone marrow to produce red cells blood and the kidneys play the major role in EPO synthesis [ 10, 11 ]. When there is a drop in normal blood oxygen levels due to reduced number of red blood cells decreased oxygen availa bility or increased tissue demands for oxygen; kidneys accelerate their EPO release [ 11, 10 ]. Kidneys are so

PAGE 25

25 important in the red blood cell formation process that a renal pathology reflects directly in the erythrocyte production. To c ite an example, dialysis patients do not produce enough erythropoietin to support normal erythropoiesis, resu lting in RBC counts less than half that of healthy individuals [ 10 ]. Matured erythrocytes are unabl e to synthesize proteins, grow and divide and they lose their flexibility, becoming increasingly rigid and fragile as they age [ 10 ]. Red blood cells have a useful life span of approximately 120 days. Ma ture erythrocytes swel l and are engulfed and destroyed by macrophages in the spleen. The he me of their hemoglobin is separated from the globin and it is degraded to bilirubin while the globin is me tabolized and broken down into amino acids which are released back to the circulation [ 10 ]. At the same time, the iron core is recycled, bound to protein (a s ferritin or hemosiderin) and stored for reuse [ 10, 11 ]. Red Blood Cell as a Model for Drug Delivery Erythrocytes are major candidates for drug de livery applications due to their abundance and some unique characteristics such as long life -span in circulation, excellent biocompatibility and biodegradability, and non-immunogenecity [ 26, 27 ]. As a result, they have been explored extensively for two potential applications: (1) the sustained delivery of therapeutic agents in the blood stream for a relative long term; and (2) the continuous and targeted delivery of drugs or enzymes to organs of the reti culendothelial system (RES) [ 26, 28 ]. Certainly, us ing erythrocytes as biological carriers offers an alternative to othe r carrier systems such as liposomes or polymeric microand nano-partic les that have been used for encapsulation of various drugs, enzymes and peptides with therapeutic activity [ 29 ]. Encapsulation Methods In general, the steps to prepare carrier er ythrocytes include blood collection, erythrocyte separation, drug encapsulation, resealing of the RBC carrier, and finally re-injection to the

PAGE 26

26 organism [ 26 ]. Different techniques have been suggested to accomplish drug encapsulation within erythrocytes along with a proper delivery. The osmotic methods, which are based on the encapsulation under reduced osmotic pressure conditions, are the most widely used. Some of these methods include hypotonic dilution, hypot onic pre-swelling, osmotic pulse, hypotonic hemolysis and hypotonic dialysis [ 29, 26, 28 ]. Other techniques used for drug encapsulation within erythrocytes consist of endocytosis and chemical and electrical processes [ 26, 29 ]. Advantages and Disadvantages of Re d Blood Cells in Drug Delivery As mentioned, erythrocytes present unique ch aracteristics that made them a desirable system for drug delivery. In addition to bei ng a natural, biocompatible, biodegradable, and nonimmunogenic system; carrier RBCs offer the chance of loading a fairly high amount of drug in a small volume, assuring dose sufficiency usi ng a limited volume of erythrocyte samples [ 27, 30 ]. Other advantages in using erythrocytes as dr ug delivery systems include: their abundance, size, morphology and inert intracellular environmen t; providing drug protection from endogenous factors and cell metabolic activities, as well as the organism protection ag ainst toxic drug effects [ 26, 27 ]. However, the two main advantages of carri er erythrocytes are that they act as a true drug delivery system by modifying the drugs pharmacokinetic and pharmacodynamic parameters, and their selective distribution to the RES organs [ 26, 28, 29 ]. The latter property of the carrier RBCs is of great therapeutic importance in drugs such as antibiotics, enzymes or antiHIV peptides, among others [ 28, 26 ]. The clinical application of carri er erythrocytes has been lim ited by two main factors. The lack of reliable and appropriate in-vitro stor age methods for maintenance of cell survival and drug content has become a major limiting factor [ 26 ]. Furthermore, autologous applications might be limited depending on the disease state since the RBC morphology is directly affected by certain pathologies. Although, the use of allocarriers could so lve the problem, a whole set of

PAGE 27

27 complications arises from th is solution such as loss of the non-immunogenic property and insufficient blood donors. Drug Delivery Systems The Food and Drug Administration (FDA) has approved a number of proteins, including monoclonal antibodies, growth fact ors, cytokines, soluble recept ors, and hormones, to treat a variety of diseases [ 31, 32, 33 ]. However, conventional oral and intravenous (IV) delivery of these drugs is usually not effective because of the inherent instability of many proteins [ 33, 31, 32, 34, 35 ]. Proteins have a very short in vivo half-life, are incapab le of diffusing through biological membranes and are unstable in the body environment [ 33, 36, 37 ]. Although intravenous protein administra tion is most effective, daily injections and high protein concentrations are required to achieve an eff ective local concentration for a prolonged time [ 33, 38 ]. Frequent systemic doses increase treatmen t cost, patient discomfort, and side effects. To improve delivery of proteins, many contro lled-release delivery systems composed of polymeric biomaterials have been developed [ 39, 40, 41 ]. The main goal of developing these systems is to control the release of drugs so that a th erapeutic level is achie ved for long periods of time. Besides the therapeutic advantage, ther e is a business aspect fo r the great interest in controlled-delivery systems. Due to increa sing FDA regulations, pharmaceutical companies need to invest more than $800 million for introducing a new drug in the market in addition to spending more than 10 years of research and development work [ 42 ]. Therefore, creating new devices or systems that deliver the same drug in a controlled manne r is an economical strategy of extending the license of the same drugs [ 42 ]. Controlled-release delivery systems are classi fied depending on the mechanism controlling the drug release (Table 2-5) [ 42 ]. The most promising delivery a pproach is the encapsulation of protein drugs within biodegradabl e polymers processed in a form that facilitates administration

PAGE 28

28 through a syringe needle (p articulate systems) [ 33, 41, 43 ]. Currently, three injectable polymer configurations are used: nanoor microspheres, which are sp herical matrix particles with the drug uniformly distributed in the matrix; nanoor microcapsules which are conformed of a welldefined core containing the therapeutic agent and a polymer membrane surrounding the core; and cylindrical implants of approximately 0.8-1.5 mm in diameter [ 42, 43 ]. Microspheres and microcapsules have several advantages over cylin drical implants, including less painful and a more simplified administration [ 43 ]. Besides decreasing cost and frequency of inject ions, encapsulation of proteins and peptides within biodegradable polymeric particulates ha s three key advantages over conventional drug delivery systems [ 44 ]: localization of the drug at the site of action, continued and prolonged rele ase of the therapeutic drug, and protection of proteins and peptides ag ainst chemical or enzymatic degradation from the physiological environment. Microspheres and Microcapsules Microencapsulation has been widely used not only to develop controlled-release drug delivery systems; but to masquerade tastes and o dors, reduce toxicity, and protect cells from the host immune response in the abse nce of immunos uppression drugs [ 45, 46 ]. Among all the microencapsulation systems, biodegradable polyme ric nanoand micro-sized particulates are the most promising ones for controlled delivery of different drugs, either hydrophobic or hydrophilic ones. Microspheres are considered polymeric matr ices with no superficial membrane. In this system, the drug is relatively distributed hom ogeneously through the entire polymer matrix, resulting in a release kinetic governed by erosion and diffusion [ 42 ]. Although microspheres have been made with different kinds of polym ers such as polyesters, polymethacrylates, and

PAGE 29

29 celluloses; poly(lactide) or poly(lactide-co-glic olide)-based microspheres have been the most studied systems due to the excellent biocompatib ility and biodegradability properties of the polymers [ 42, 47 ]. Table 2-6 shows a number of co mmercially available poly(lactide-coglycolide) (PLGA) microsphe res used for drug delivery [ 42 ]. On the contrary, microcapsules present a well -defined core-shell structure where the core is loaded with the therapeutic agent and th e shell provides the pharm acokinetic-release limiting factor [ 42, 48 ]. The release kinetics for this type of system is governed by diffusion from the core through the degrading shell [ 48 ]. To date a diversity of materials has been employed as shell components (i.e. synthetic and natura l polyions, proteins, nucleic acids, lipids, nanoparticles, etc) while biologi cal cells, latex and inorganic particles, oil dispersion, and organic crystals have been used as core templates [ 49, 48 ]. Microencapsulation Techniques Several microencapsulation approaches w ith biodegradable polymers have been developed, which are currently used in numerous applications in industr y, agriculture, medicine, pharmacy, and biotechnology [ 50 ]. Some of these methods include emulsion solvent evaporation, solvent extracti on, coacervation, spray-drying, inte rfacial complexation, coating, and hot melt coating [ 46, 51 ]. Although, each method has both a dvantages and disadvantages in the elaboration of polymeric mi croparticles, a common problem w ith all the approaches is the preparation of truly efficient sustai ned-release delivery systems. The most commonly used methods of prepari ng protein-loaded micr ospheres are the waterin-oil-in-water (W/O/W) double emulsion and the oil-in-water (O/W) si ngle emulsion solvent evaporation techniques [ 47, 52 ]. Both methods involve the polymer dissolution in an organic solution; mixing (shake, vortex, or sonication) of the drug, either in the powder or liquid form,

PAGE 30

30 with the organic phase; and dispersion of the em ulsion/solution into a co ntinuous aqueous phase followed by solvent evaporation or extraction. Although these techni ques are fairly simple, cost effective, and easily up-scalable for mass production; they lack effectiveness in producing a linear controlled-re lease system. The main drawback s of these methodologies include denaturing of some encapsulated proteins due to the manufacturing process conditions (such as intimate contact with organic solvent, heating, and mechanical forces), poor encapsulation efficiency for hydrophilic drugs, and the polydispe rsed size of prepared microspheres generally ranging from 10 to 100 m. Although some of these technological problems have been addressed by the use of milder organic solvents and surfactants, the process still needs improvement. In addition to the emulsion solvent evaporat ion methods, the polymer phase separation and spray-drying techniques have also being used for the preparation of protein-loaded microspheres. The advantages of the polymer phase separati on method are that no aqueous phase is involved, eliminating the loss of protein th rough the aqueous phase as in th e emulsion techniques, and that the whole process takes place at room temperature, which avoids heat-induced denature of the protein [ 47 ]. However, elevated concentrations of residual solvents have been found in the microspheres [ 47 ], making them highly toxic and immunoge netic. Spray-drying has also being used to encapsulate hydrophilic and hydr ophobic drugs within several polymers [ 53, 54 ]. The major advantages of this techni que are the one-step process and eas e of parameter control as well as of scale-up [ 54 ]. On the other hand, the preparation of microand nano-sized capsules involves a wide variety of manufacturing techniques. For this kind of system, fabr ication of the core is not as important as the shell preparation since the syst ems defining properties will depend on the latter.

PAGE 31

31 While any of the techniques mentioned earlier coul d be applied to manufacture the core, the most common approach used to fabricate the sh ell involves self-ass embly strategies [ 55 ]. This method, initially applied to planar surf aces, employs oppositely charged nanocomposite multilayer films assembled onto templates of charged colloidal particles [ 55, 56 ]. The layer-by-layer (LbL) absorption technique, originally introduced by Decher [ 57 ], is the method currently used to fabri cate the core-shell particle sy stem known as polyelectrolyte complex (PEC). The LbL method is based on the electrostatic in teraction between polyanions and polycations that are consecutively absorbed on a charged planar or spherical surface [ 49, 58, 55 ]. The LbL method provides an effective and simple approach to manufacture PEC systems with customized chemical and physical properties significantly different to those of the colloidal template [ 51, 58 ]. In addition to customization of part icle properties, the LbL approach permits one to control the thickness of PE C layers with nanometer precision [ 55, 56, 48, 59, 7 ]. An important feature of PEC particles is the successive dissolution of the colloidal template, resulting in free-standing, polyelect rolyte-shell capsules w ith the shape and size determined by the template [ 7, 56, 58, 48 ]. The core removal is attained in a case-specific manner since it depends on the templates chemical nature [ 51 ]. The major disadvantage of polyelectrolyte capsules is their instabilit y, depending on pH, temperature, and salt concentrations [ 51, 60 ]. However, this disadvantage can be used to develop controlled-release drug delivery capsules. In addition, capsules w ith prolonged release prop erties can be obtained by increasing the number of polyelectrolyte layers [ 58, 48 ]. Another significant property of these capsules is the selective permeability of their polyelectrolyte shell: permeable for small molecules and ions and impermeable fo r higher molecular weight compounds [ 56, 58, 61 ]. Certainly, PEC capsules offer very attractive properties for the encapsulation of proteins,

PAGE 32

32 peptides, oligonucleotides, genes, and cells fo r many applications in biotechnology, medicine, pharmaceutics, cosmetics, and the food industry. Immune System Response to Foreign Elements The immune system is a complex network co mposed of many proteins, cells, and a few well-defined organs [ 62 ]. Its main function is to protec t us against pathogenic agents and diseases by recognizing bacteria, fu ngi, viruses, parasites, cancerous cells, and foreign elements [ 42, 62 ]. The immune system actually recognizes macromolecules (such as proteins or polysaccharides) of foreign elements and the degree of immunogenecity of these structures will depend on their foreignness, molecular size, chemical composition and complexity, and the ability to be processed and presented with a major histological complex molecule [ 62 ]. Depending on the foreign agents location in the body, different organs and cel ls will be involved in the immune response. The immune system response represents the ma jor limiting factor in the efficacy of longterm, controlledrelease delivery particulates. Intravenously admi nistered particulate carriers are rapidly recognized by cells of the reticuloendothelial system (RES) and, consequently, removed from the systemic circulation within minutes [ 62, 63, 64, 65 ]. The efficient elimination of particulate systems by the RES is known for a long time. Depending on the surface chemistry, charge and hydrophilicity, the removal process is initiated by opsonization. A set of plasma proteins, known as opsonins, absorbs onto th e surface of particulates and make them recognizable to the RES [ 64 ]. Classical opsonin mol ecules include immunoglobulins, complement proteins (such as C1q, C3b, and iC3b), apolipoproteins, von Willebrand factor, thrombospondin, fibronectin, a nd mannose-binding protein [ 63, 64 ]. Opsonized particulates are then phagocytosed by hepatic midzona l and periportal Kupffer cells [ 63, 66 ]. The spleen and bone marrow might also participate in the part icle clearance process from the bloodstream

PAGE 33

33 depending on the pathophysiological conditions a nd the physicochemical characteristics of the particulate carriers [ 66 ]. When particulate carriers are small and have a neutral-ch arged surface, they are not efficiently opsonized and, therefore, they are poorly recognize d by Kupffer cells [ 63 ]. However, they might still go through a cl earance process from the vascul ature by fenestrations in the hepatic sinusoidal endothelium, the spleen or bone marrow [ 63, 66 ]. Particles with diameters of less than 100 nm get trapped by ex trusion through endothelial fenest rations in the space of Disse and the hepatic parenchyma [ 63 ]. The size, deformability, and sphericity of drug delivery particulates also play a cr ucial role in their removal by the sinusoidal spleen [ 63 ]. Particles larger than 200 nm and their aggregates can be physically trapped in th e spleen fenestrations, unless they are deformable as in the case of erythrocytes [ 63, 65 ]. On the other hand, the particle elimination mechanism by the bone marrow is more complex and species-dependent, capable of removing particles from the circulation by transcellular and intercellular paths [ 66 ]. Certainly, physicochemical factors of drug de livery particulates are critical for their recognition and removal from the bloodstream by the RES. For the past 30 years, it has been known that hydrophilic particles re mained in the circulation l onger than hydrophobic ones due to rapid opsonization of the latter [ 67 ]. The particles surface charge is another factor influencing the clearance process due to elect rostatic interactions with blo od components and cell surfaces. Although, there are conflicting viewpo ints regarding the surface charge it is believed that neutral charged particulates have a longe r half-life in the bloodstream [ 67, 64 ]. The other major factor determining the RES removal of particulate syst ems is the size of the particles. A narrow particle diameter range, between 100 and 200 nm, is preferred to avoid particle entrapment in hepatic and splenic fenestrations [ 66 ]; unless, the particulates are deformable in which case

PAGE 34

34 larger diameter particles could be administered, increasing the dr ug loading potential. Indeed, a fuller understanding of the physicoc hemical properties of drug delivery particulates and their effects on the immune response will make it possibl e to design particle carriers with reduced affinity to the cells of the RES.

PAGE 35

35 Table 2-1. Plasma protein concentrations (mg/100 mL) Protein Plasma Total 6500-8000 Albumin 4000-4800 1-globulins 380-870 2-globulins 570-940 -globulins 730-1380 -globulins 590-1450 Fibrinogen 200-400

PAGE 36

36 Table 2-2. Physicochemical proper ties of the major plasma proteins Species Concentration (mg/ml) Mol weight (Da) pI Sedimentation const in water at 20C (10-3 cm/dyn*s) Diffusion coeff in water at 20C (10-7 cm2/s) Partial specific volume of protein at 20C (ml/g) Prealbumin 10-40 5.5x104 4.7 4.2 ----Albumin 35-45 6.6x104 4.9 4.6 6.1 0.733 1-seromucoid 0.5-1.5 4.4x104 2.7 3.1 5.3 0.675 1-antitrypsin 2.0-4.0 5.4x104 4.0 3.5 5.2 0.646 2-macroglobulin 1.5-4.5 7.2x105 5.4 19.6 2.4 0.735 2-haptoglobin Type 1.1 1.0-2.2 1.0x105 4.1 4.4 4.7 0.766 Type 2.1 1.6-3.0 2.0x105 4.1 4.3-6.5 ----Type 2.2 1.2-2.6 4.0x105 --7.5 ----2-Ceruloplasmin 0.15-0.60 1.6x105 4.4 7.08 3.76 0.713 Transferrin 2.0-3.2 7.6x104 5.9 5.5 5.0 0.758 Lipoproteins LDL ( <1.019) 1.5-2.3 1.5x107 -->12 5.4 --LDL ( <1.063) 2.8-4.4 3.2x106 --0-12 ----HDL2 ( =1.093) .37-1.17 4.4x105 --4-8 ----HDL3 ( =1.149) 2.17-2.70 2.0x105 --2-4 ----IgA (monomer) 1.4-4.2 1.6x105 --7 3.4 0.725 IgG 6-17 1.5x105 6.8 6.5-7.0 4.0 0.739 IgM 0.5-1.9 9.5x105 --18-20 2.6 0.724 C1q 0.1-0.25 4.0x105 --11.1 ----C3 1.5-1.7 1.8x105 6.4 9.55 4.5 --C4 0.2-0.5 2.1x105 --10.1 ----Fibrinogen 2.0-4.0 3.4x105 5.5 7.6 1.97 0.723

PAGE 37

37 Table 2-3. Concentrations of major electr olytes (mEq/L) in whole blood and plasma Electrolyte Whole blood Plasma Bicarbonate 19-23 24-30 Calcium 4.8 4.0-5.5 Chloride 77-86 100-110 Magnesium 3.0-3.8 1.6-2.2 Phosphate 0.76-1.1 1.6-2.7 Potassium 40-60 4.0-5.6 Sodium 79-91 130-155 Sulfate 0.1-0.2 0.7-1.5 Table 2-4. Concentrations of various orga nic compounds (mg/100mL) in whole blood and plasma Species Whole blood Plasma Amino acids 38-53 35-65 Bilirubin 0.2-1.4 0.2-1.4 Cholesterol 115-225 120-200 Creatine 2.9-4.9 2.5-3.0 Creatinine 1-2 0.6-1.2 Fat, neutral 85-235 25-260 Fatty acids 250-390 150-500 Glucose 80-100 60-130 Lipids, total 445-610 285-675 Nonprotein N 25-50 19-30 Phospholipids 225-285 150-250 Urea 20-40 20-30 Uric acid 0.6-4.9 2.0-6.0 Water 81-86g 93-95g

PAGE 38

38 Figure 2-1. Scanning electron micrograph of red blood cell aggregates, rouleaux. Figure 2-2. Shear rate dependence of normal hum an blood viscoelasticity at 2 Hz and 22 C.

PAGE 39

39 Table 2-5. Controlled re lease drug delivery systems Diffusion-controlled Reservoir and monolithic systems Water penetration-controlled Osmotic and swelling-controlled systems Chemically-controlled Biodegradable reservoir and monolithic systems Biodegradable polymer backbones with pendant drugs Responsive Physicallyand chemically-responsive systems (T, solvents, pH, ions) Mechanical, magnetic or ultrasound-responsive systems Biochemically-responsive; self-regulated systems Particulate Microparticulates Polymer-drug conjugates Polymeric micelle systems Liposome systems Table 2-6. Examples of commercia lized PLGA copolymers microspheres Trade Name Drug Decapeptyl Depot Triptorelin Enantone LP Leuprorelin Somatulin LP Lanreotide Parlodel LAR Bromocriptine Sandostatin-LAR Ocreotide Nutropin Recombinant Human Growth Factor Lupron Leuprolide Acetate

PAGE 40

40 CHAPTER 3 CHARACTERIZATION OF THE DEFORMAB LE BEHAVIOR OF PLURONIC-PLA AND POLYELECTROLYTE CHITOSAN -ALGINATE MICROPARTICLES Introduction Since the1980s numerous groups have tried to develop a red blood cell analog due to the HIV epidemic of the time [1, 2]. Even though some of these substitutes ar e now in phase III of clinical trials, their use is very limited due to side effects and short half-life time within the human body [ 4 ]. As a result, there is still a need for an effective erythrocyte analog with minimum immunogenic and side effects, so that it can be used for multiple applications. Besides the imperative need of a blood substitute for in vi vo use, there is also a need of it for in-vitro testing of medical devices and products. Many types of medical equipment and products have been engineered with the purpose to interact directly with blood. To illustrate, some of the these devices include dialysis machines, ventricular assist devices, h eart-lung machines, heart valves, catheters, grafts, stents, among others. Since blood will be in direct contact and, sometimes, ev en subject to mechanical stress with this type of devices, it is n ecessary to test their effect on blood in vitro before moving on to clinical trials. However, testi ng with real blood involves an array of complications such as cost, increment in rules and regulati ons, availability, vari ability among donors, and inconsistency with storage time. Particles with deformability properties and dispersed in fluids with viscosities similar to blood will represent a valuable adva nce for the testing of medical devices and products. In this study we investigated the use of synthetic and natural biode gradable polymers as possible materials for the development of biodegradabl e, biocompatible, and multi-functional particles with deformability properties fo r applications in device testing and drug delivery. The main goal

PAGE 41

41 is to attain particles capable to deform under micropipette suction pressu res used to aspirate erythrocytes. Poly(Lactic Acid) In the development of medical devices, especi ally for controlled-release delivery systems, synthetic biodegradable polymers ar e frequently used as carriers for protein drugs. Synthetic polymers are preferred over biological material s because of their biocompatibility, minimal immunogenecity, biodegradabilit y, and high manufactured reproduc ibility. The polymers most often used for the fabrication of drug delivery systems are poly(glycolic acid) (PGA), poly(lactic acid) (PLA), and the copolymer pol y(lactide-co-glycolide) (PLGA) [ 33, 68 ] due to their good biocompatibility, variable mechanical processa bility, and a wide ra nge of biodegradable properties [ 33, 41 ]. Among these three polyesters, PLA (F igure 3-1) with the chemical formula (C3H4O2)n is the most hydrolytically stable in addition to its long history of safe and biodegradable use as resorb able suture materials [ 69, 43, 70, 71 ]. PLA degrades by a well-known erosion process in to natural lactic-aci d metabolites that are easily eliminated by the body. Hydrolysis of the ester bond initiates the erosion process; followed by a reduction in molecular weight and an increase in the acidic environment which, in turns, accelerates degradation [ 42 ]. The PLA erosion rate is c ontrolled by varying the molecular weight and type of polylactide monomer used [ 71, 68, 72, 42 ]. These factors determine the hydrophilicity and crystallinity, which govern the rate of water penetration [ 68, 44 ]. The higher the molecular weight, the longer the polymer reta ins its structural inte grity, the slower its degradation rate [ 68 ]. In this study, DL-PLA was the pol ymer used since it is completely amorphous and its lack of crystallinity causes it to degrade faster than L-PL A. In addition, it has a lower tensile strength which makes an at tractive property when trying to manufacture deformable microspheres.

PAGE 42

42 Pluronics Surfactants are an important constituent for many colloidal suspensions, governing rheological properties and stabil ity against phase separation of many commercial dispersions in the cosmetic, food, and medical field [ 73 ]. Poloxamers, also known by the commercial name Pluronics (BASF, Wyandote, USA) are commonly used as stabil izers in the preparation of emulsions and colloids. Pluronics are tr iblock ABA-type copolymers composed of poly(ethylene oxide)-poly(propyl eneoxide)-poly(ethylene oxide ) (PEO-PPO-PEO) blocks arranged in a basic structure. Table 3-1 shows a list of the commercially available Pluronics which can be classified as hydrophobic or hydrophilic depending on the EO/PO ratio [ 74 ]. Pluronics are soluble in wate r and polar solvents. In a queous solutions, they exhibit temperature-dependent rheological propertie s due to their amphiphilic structures [ 74 ]. Gelation of Pluronics solutions will o ccur above a certain temperatur e and it will also depend on the polymer concentration and EO/PO ratio [ 74 ]. Moreover, Pluronics in aqueous solutions are capable of self-assembling into multi-molecular aggregates (micelles), which makes them attractive for drug delivery applications [ 75 ]. Micelle formation will also depend on block copolymer concentrations, EO and PO unit length s as well as temperature and type of solvent used [ 76, 77 ]. In addition to unique gelation and micelle fo rmation properties, some of the Pluronics exhibit minimal toxicities in vivo allowing their clinical use [ 78 ]. Furthermore, modification of polymeric surfaces with PEO to reduce nonspecific protein adsorption and undesired bioadhesion in biological envi ronments has been proven eff ective for the past 30 years [ 64, 36 ]. The valuable properties of PEO surface modifications are associated to the distinctive structure of PEO molecules which show greatly hydrated, non-ionic, and mobile chains in an aqueous environment [ 64 ]. However, using Pluronic copolymers instead of the PEO homopolymer, for

PAGE 43

43 surface modification is preferred due to the improved stabiliza tion of the system, provided by anchoring of the PPO units to the solid surface [ 36, 64 ]. In this study, conjugations of DL-PLA and Pluronics with different compositions and mo lecular weights were used to investigate the systems feasibility in the development of deformable particles. Alginate Alginates are random, anionic, linear, polysaccharides derived from brown algae. These natural block copolymers are conformed by va rying ratios of unbranched chain 1,4-linked -Dmannuronic acid (M) and -L-guluronic acid (G) residues (Figure 3-2) [ 79, 80, 81, 82 ]. Frequency and distribution of these monomers along the polymer chain are irregular and dependent on the source of origin [ 83 ]. Although alginates have been isolated from bacteria, three species of brown algae (Laminaria yperborean, Ascophyllum nod osum, and Macrocystis pyrifera) are the primary source for commercially available alginates [ 84 ]. Approximately 40% of the algae dry weight is alginate which is found in the intracellular matrix as a mixed salt of diverse cations from the sea water (i.e. Mg2+, Ca2+, Sr2+, Ba2+, and Na+) [ 84 ]. Though alginates have several uni que properties, gela tion and swellability are the two most significant. Alginate forms a gel in the presence of divalent cations and calcium-crosslinked gels are the most abundant in nature [ 84 ]. The affinity of alginate s for divalent cations depends mostly on the electronic structur e of the cation, with the highest affinity for copper ions and the least for manganese ions [ 85 ]. Metal ions, especially calcium, are more commonly used to crosslink gel alginates since toxic ity has been a limitant factor in the use of transition metals. The sol-gel transformation begins with the exch ange of monovalent ions from the G residues of water-soluble alginate salts with divalent cations [ 80, 86 ]. Binding of divalent cations to G blocks is highly cooperative, formin g stacks of more than 20 monomers [ 84 ]. Martinsen et al. described the physical orientation of the stacks as the egg-box stru cture, being more abundant for

PAGE 44

44 high G alginates than for low G ones [ 83 ]. However, Wang et al. found that transition metal ions bind to both G and M residues w ith a binding distribution depe ndent on solvent conditions [ 85 ]. As a result, alginate composition as well as cati on affinity will play a significant role in the properties of the crosslinked gel matrix. The swelling property of crosslinked alginate gels depends on solutes present in the solution [ 85 ]. Sequestering agents form stable co mplexes with bound cations, resulting in a polymer chain relaxation and volume expansion [ 61 ]. Calcium-alginate gel matrices start swelling and are further destabili zed upon contact with solutions containing chelators such as phosphate, lactate, citrate, ethylenediamine te traacid (EDTA) or high concentrations of nongelling cations like sodium or magnesium ions [ 61, 84 ]. Swelling and degradation of alginate gels due to removal of crosslinker ions, time, pH or temperature are dete rmining features in the development of controlled-release systems. As a result, alginate matrices have been commonly used for a variety of delivery sy stems including gels, f ilms, beads, microparticles, and sponges. In addition to alginates unique properties, their chemistry a nd relatively mild crosslinking conditions, fairly easy processability, source abundant, low price, biocompatibility, and biodegradability have enabled th eir used in a wide variety of biomedical applications [ 80, 81, 82 ]. Some of these applications include immunop rotective containers in cell transplantation, cell scaffolds, controlled-release drug delivery system s, and surgical dressings for the treatment of adhesion problems in tissue repair, capi llary hemorrhage blockage, and burns [ 45, 61, 83, 84 ]. Certainly, a wide range of matr ices with different morphologies pore size, water content, and release rates can be manufactured by selecting determined alginate compositions, crosslinkers, additives, gelation conditions and coating agents.

PAGE 45

45 Chitosan Chitosan is a cationic, lin ear, naturally occurring polys accharide with structural characteristics similar to glycosaminoglycans [ 87, 88 ]. This polycationic biopolymer is fundamentally composed of (1-4)-linked D-glucos amine units with some percentage of N-acetylD-glucosamine units (Figure 3-3) [ 50 ]. Chitosan is commonly obtained by alkaline deacetylation of chitin since it is rarely found in nature [ 89 ]. Conversely, chitin is the second most abundant natural polysaccharide [ 90 ], present in the exoskelet on of crustaceans, mollusks, the cell walls of fungi, and the cuticle of insects [ 87 ]. Chitin is a homopolymer consisting of (1-4)-linked N-acetyl-D-glucosamine units, whic h can be specifically modified by controlled chemical reactions [ 50 ]. The degree of deacetylation of chitin will determ ine the functional properties of chitosan. A low-molecular-weight, highly-deacetylated chitosan, known as chitosan oligosaccharide (COS), is obtained when chito san is further hydrolyzed [ 89 ]. COS has very promising properties which can be utilized in a wide range of applications. While chito san is only biologically active in acidic environments due to poor solubility, COS is water soluble at neut ral pH as it maintains its cationic nature. This prope rty allows COS to electrostatically interact with polyanionic polymers and molecules in diverse aqueous envi ronments, forming polyelectrolyte materials optimal for drug delivery [ 83, 91 ]. Besides the physicochemical features, many usef ul biological properties of chitosan have been recognized, including biocom patibility; biodegradability; low toxicity; mucoadhesion; and antifungal, antimicrobial, anticoagulant, an titumoral and hipolipidemic activity [ 50, 87, 92 ]. Chitosan is metabolized and degraded into nontoxic products by enzymes, such as lysozyme, lipase and chitosanase [ 50 ]. While the latter is found in pl ants and insects, the other two are present in mammals [ 50 ]. All these interesting characteri stics have led to the recognition of

PAGE 46

46 chitosan and its derivatives as potential materials in numerous applications in agriculture, environment, food industry, medicine, pharmacy, a nd biotechnology. Some of the applications in the medicine and pharmaceutical fields incl ude: surgical sutures, sponges and bandages, matrices and coatings for drug delivery syst ems, orthopedic and den tistry materials, cell scaffolds, and immunoprotective barr iers for cell transplantation [ 50, 88 ]. To attain microcapsules with deformable proper ties, we studied differe nt concentrations of alginate and chitosan for the development a polye lectrolyte complex (PEC) system consisting of multiple-layer microcapsules with a dissolvable gel core. Materials and Methods Preparation of PLA-Pluronic Particles Polyester microspheres were prepared by the water-in-oil-inwater (W/O/W) doubleemulsion solvent extraction/evapor ation technique. The polymer used was D,L-poly lactic acid (DL-PLA) (Mw 350 kDa) from Bi rmingham Polymers. In addition to PLA, Pluronics with different HLB (kindly donated by BASF) were used as surfactants (Table 3-2). Pluronic was dissolved at different concentra tions in phosphate buffer saline (PBS), pH 7.4, and emulsified in methylene chloride containing 2% (w/w) DL-P LA. This first water-in-oil emulsion was generated by ultrasonication for 60 seconds in an ice water bath. This emulsion was added to 1.5% polyvinyl alcohol (PVA) solution at 4C unde r continuous stirring at 1500 rpm for 30 min. Microspheres were collecte d by filtration, through a 300m nylon mesh, and centrifugation. PLA-Pluronic microparticles were washed thre e times with deionized water, lyophilized, and stored at -20C until use. Preparation of Alginate Particles Natural polymeric microspheres were prepared by a water-in-oil emulsion crosslinkinggelation technique [ 93 ]. The polymer used was alginate from Keltone (LV, food grade).

PAGE 47

47 Calcium chloride and copper nitrate at different concentrations were used as crosslinking solutions (Table 3-3). The aqueous phase consis ted of sodium alginate dissolved in deionized water at different concentrations. The oil phase consisted of soyb ean oil. The first emulsion was obtained by ultrasonication of two phases at 60W for 1 min in an ice bath. A second aqueous solution containing different con centrations of crosslinking agen t was added to the emulsion by air-spray (40 psi, 20ml/hr) at a 4-cm dropping distance while stir ring the whole medium slowly with a magnetic stirrer. Partic les were allowed to cure for ten minutes under cont inuous stirring. Then medium was allowed to rest for 24 hr so that particles would dr op to the bottom of the container while the oil phase was left at the t op. After separation of th e two phases, particles were collected by filtration through a 45m mesh and washed copiously to remove the oil. Particles were stored at 4C until future use. Coating of Alginate Particles Caand Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption technique. The coating always started and ende d with the positive polymer. Microcapsules with zero, one, five, and nine alternating layers were fabricated. The cationic polymer used was a high-molecular-weight chitosan and a high-de acetylated, low-molecular-weight chitosan oligosaccharide lactate (Mw < 5000, 90% deacetylati on) from Aldrich. Algi nate was the anionic polymer used. A chitosan solution containing CaCl2 or alginate solution was added to crosslinked-alginate pa rticles dispersed in deionized wate r. Different ba tches were made differing in the concentrations of chitosan and alginate solutions. After allowing the particles to coat for 30 minutes, they were co llected by centrifugation and washed three times with deionized water to ensure that all free polyelectrolytes were removed. Following the washes, particles were r eady for the next coating or stored at 4C until

PAGE 48

48 use. Table 3-3 shows all the formulations used to make the different batches of chitosan-alginate microcapsules. Decomposition of the Particles Core Decomposition of the particle s core was achieved in a case-specific manner by treating particles with sequestrant agents specific to each cro sslinking ion. The agents used to remove calcium and copper ions were phosphate (Ce llgro) and bovine serum albumin (Cellgro) respectively. An initial concentr ation of 5% (w/v) chelating medi a with or without supplemental ions was prepared. Multiple-layer particles, at a 5% (w/v) concen tration, were incubated at room temperature in the chelating me dium specific to each crosslin king ion. After close observation, old medium was replaced by same amount of fres h medium if core was not dissolved. Medium was replaced until the core was decomposed. Gelcore particles were then stored at 4C until further use. Measurement of Particles Deformability The micropipette aspiration technique was applie d to assess particles deformability. In this technique, a portion of a single microcapsule is drawn into a pipette by applying a pressure difference between the inside of the pipette a nd the chamber containing the particle suspension [ 94, 95 ]. The Plexiglas, home-built chamber open at one side, allowing for the pipette entry, was prepared by fixing with vacuum grease a glass coverslip to the top and the bottom of the chamber. Micropipettes were pulled from glass capillaries of 1 mm diameter using a pipette puller. To attain the desired diameters, micropipettes were forg ed subsequently. Micropipette tips with inside diameters approximately 20% smaller than the particles diameter were fabricated. Previous to their use, pipettes were backfilled with 0.9% NaCl solution using a plastic syringe with a 97-mm long, 28G backf iller (Microfil, World Precision Instruments, Sarasota, FL).

PAGE 49

49 Before particles were used for deformability studies, they were incubated in sequestrantcontaining media until the core was decomposed. The prepared chamber was then filled with microparticles suspended in sequestering media and mounted on the microscope stage. An initial negative pressure of 5 cm H2O (approximately -4.9x103 dynes/cm2) was applied and pressure was augmented in increments of five. After a section of the particle had been aspirated, the negative pressure was decreased and the particle was unloaded. The whole aspiration process was monitored and recorded using an inverted light microscope (Axiovert 100, Zeiss), a Carl Zeiss video camera system, and a VCR. The record ed images were converted into digital images using the Matrox software. Since the main goal of the study was to attain particles capable to deform under micropipette suction pressure s used to aspirate erythrocytes [ 96 ], particles were considered deformable only when using pressures up to -40x103 dynes/cm2 (-41 cm H2O). Scanning Electron Microscopy (SEM) Analysis Surface morphology of the microcapsules was examined by SEM after gold-palladium coating of microcapsule samples on an al uminum stub. Samples for scanning electron micrographs were obtained after 0, 1, 5, and 9 coa tings. Droplets of microsphere solutions were mounted on aluminum stubs, let air dried and spu tter-coated with gold and palladium particles. The stubs were mounted in a scanning elec tron microscope at 10.0 kV and imaged at x500, x1000, and x5000. Results and Discussion PLA-Pluronic Particles Microspheres made with PLA and Pluronics were not deformable when using the micropipette technique, except for particles ma de with PLA and P105. Pluronics used had different molecular weights and HLB ratios ra nging from the most hydrophilic to the most hydrophobic copolymer. The first type of Plur onic used was F127 which is a hydrophilic

PAGE 50

50 copolymer very commonly used for drug delive ry systems. F127 has a high PPO molecular weight (4000 Da) and a PEO conten t of approximately 70 % (w). Pluronics were dissolved in phosphate buffered saline (PBS, pH 7.4) at differe nt concentrations ranging from 1% to 20% (w/v). At concentrations belo w 20%, Pluronic solutions did not gel at room temperature. In addition, particles obtained at lo wer concentrations were not uni form; therefore, 20% was the Pluronic solution concentration used for all further experiments. Besides variation of Pluronic solution con centration, particles w ith formulations of different PLA/Pluronic ratios we re synthesized. The ratios us ed were the following: 1:1, 1:2, and 1:3. Microspheres were not obt ained with the 1:1 formulation; thus we only continued using the last two formulations fo r the preparation of following batches. Even though, F127-PLA microspheres presented a morphology with a thick shell and an em pty core (Figure 3-4A), none of the batches obtained resulted in deformable microspheres. Negative as piration pressures of up to 60 cm H2O (-58.8 x103 dynes/cm2) were applied to the particles and no signs of deformation were observed. As a result, we decided to use L101 which is another high-PPO molecular weight Pluronic (3300 Da), but in the hydrophobic range with a 10 % (w) PEO content. All the L101-PLA formulations yielded microsphere s with morphology similar to the F127-PLA particles (Figure 3-4B) and deform ability was not achieved either. Since the idea of using Plur onics was to increase the water content in the PLA microsphere system so that particles become soft er, we decided to use a surfactant with higher hydrophile content and lower lipohilic content. The next Pluronic used was F68 containing 80 % (w) PEO and a very low PPO molecular weight (1800 Da). F68-PLA microspheres presented a different morphology compared to the F127and L101-PLA particles. F68-PLA particles presented a capsular morphology with a large PLA co re and a very thin Pl uronic shell (Figure

PAGE 51

51 3-4C). After applying an aspi ration pressure of -45 cm H2O (-44.1 x103 dynes/cm2), it was found that the thin shell was de formable but not the core. Pressures of up to -60 cm H2O were applied with no change in the deformable cap ability of the particles. All the F68-PLA formulations showed similar morphology and deform ability properties; therefore, we decided to keep testing other Pluronics. P105 with a 3300-Da PPO molecular weight and approximately 50% hydrophile content was used next. P105-PLA microspheres presented a defined core-shell structure (Figure 3-5). The PLA core was very porous as well as the P105 shell (Figure 3-6) The difference in morphology seen in F68and P105-PLA systems is attributed to the mol ecular weight of the PPO parts. Previous studies have shown that a high molecular weight of the PPO content is necessary to provide enough anchoring of the copolymer to the particle surface [ 36, 64 ]. Particles made with F68 displayed a very thin sh ell due to poor attachme nt of the surfactant to the PLA core. In addition to a different surface morphol ogy, P105-PLA particles displayed a higher deformable capability compared to the particles made with PLA and a ll the previous Pluronics. During micropipette experiments, it was clearly seen that the shell was very deformable although the core was not. A portion of the shell was drawn into the pipette forming a tongue when applying a negative aspirati on pressure of 25 cm H2O (-24.5 x103 dynes/cm2). However, no additional change in deformability of the particle was observed even after increasing the pressure to -60 cm H2O. All batches made with P105 were very consistent. Pluronics P105, P103, P104, and F108 were used next; all with a fixed propylene oxide content (3300 Da) and variations of the ethylene oxide content: 50, 30, 40, and 80% respectively. It was hypothesized that the differe nce in ethylene oxide content would have an influence on the

PAGE 52

52 shell thickness; longer PEO units would yield particles with th icker shells. The goal was to attain particles with a very thick, deformable shell and a small core. However, only P105-PLA microspheres presented the core-shell structure, indicating that a 50 % PEO content is ideal for the synthesis of capsular systems. Since we were not able to modify the shell and core size, we decided to focus on other materials. Table 3-4 summarizes the deformability properties of Pluronic-PLA particles. Chitosan-Alginate Particles Crosslinked alginate particles were made using the water-in-oil emulsion crosslinking gelation technique. No particles were obtaine d when using a 0.5% (w/v) sodium alginate solution. The best morphology was obtained when using a concentration of sodium alginate solution at 3% (w/v) and a CaCl2 concentration of 5% (w/v). To coat the particles, different concentrations of high-molecular-weight chitosan and chitosan oligosaccharide (COS) solutions containing 0.1% (w/v) CaCl2 were used. Crosslinked-algina te particles coated with high molecular weight chitosan did not yield good resu lts; therefore, only chit osan oligosaccharide was used for all further coatings. Also, different concentrations of alginate solutions were used for the polyanionic coating; alt hough only 0.1% (w/v) solution seemed to work. We were unable to collect the particles when higher al ginate concentrations were used. Figure 3-7 shows optical and scanning electron micrographs of polyelectrolyte multiple layer Ca-crosslinked alginate partic les. The core-shell structure of the particles can be seen from the SEM pictures. The evident i nvagination in the center of the pa rticles (Figure 3-7B) is due to loss of water from the gelatinous Ca-alginate co re. After water removal, alginate chains collapsed as opposed to P105-PLA microcapsules wh ich kept their spherica l shape (Figure 3-6). Uncoated and coated calcium-alginate part icles were not deformable when using the

PAGE 53

53 micropipette technique before removal of cross linking ions. Furthermore, uncoated, mono-layer and multiple-layer coated particles were not stable after removing the calcium ions. Alginate particles crosslinked with copper ions presented smaller diameters than Cacrosslinked particles (Figure 3-8) This could be explained by a faster crosslinking reaction, resulting in less time for destabilization of the em ulsion to take place and less coalition of small particles into large ones. After water was rem oved from the Cu-alginate core, capsules collapsed similar to Ca-alginate particles (Figure 3-8B). Coated Cu-alginate particles were not deform able when using the micropipette as well. However, when copper ions were removed, particle s presented deformable properties. Particles deformed and were completely aspirated into the micropipette tip af ter applying a 20-cm-H2O negative pressure (-19.6x103 dynes/cm2). While uncoated Cu-alginate particles were not stable after removal of crosslinking ions ; mono-layer particles were stable mostly in static conditions. Some of the mono-layer particles ru ptured under pressures of -20 cm H2O. Best results were obtained with Cu-alginate particles coated five tim es. Penta-layer particles were able to deform under a 20-cm-H2O negative aspiration pressure several times while recoveri ng their shape. Removal of crosslinking ions wa s not as efficient for the nine-layer particles as for the lowerlayer ones under the conditions tested. As a result, only por tions of the particles were deformable under a 20-cm-H2O negative pressure. Particles were completely aspirated into the micropipette only after applying negative pressures of 30-cm-H2O (-29.4 x103 dynes/cm2). Table 3-5 summarizes the deformability properties of PEC chitosan-alginate capsules. The ion exchange properties of alginate gels have been studied by many research groups. This ion exchange process depends on several fa ctors such as the amount of crosslinking ions forming the gel, physical properties of the ge l, competing ion concen tration, pH and ionic

PAGE 54

54 strength of the solution [ 97, 98, 99 ]. In addition, the order affinity of alginate for divalent ions has been found to be Pb2+>Cu2+>Cd2+>Ba2+>Ni2+>Ca2+>Zn2+>Co2+>Mn2+>Sr2+ [ 85, 100, 101, 102 ]. After removal of calcium ions from the particles by phosphate i ons present in the cell culture medium, the polyelectrolyte complex syst em was not stable anymore. The calcium PEC system collapsed regardless of the number of PEC layers because alginate had no affinity for the other electrolytes present in the medium. On the contrary, when copper ions were chelated by albumin, the alginate binding sites left were occu pied by calcium ions present in the cell culture medium. As a result, the i on exchange process along with the PEC layers kept the precrosslinked Cu-alginate particles stable. Conclusion Only P105-PLA microspheres presented a well-d efined core-shell structure with a highly porous PLA core and a P105 shell. In addi tion to a different surface morphology, P105-PLA particles displayed higher deform able capability compared to the particles made with PLA and all the other Pluronics. Du ring micropipette experiments, it wa s clearly seen that the shell was very deformable although the core wa s not. The differences in morphology and deformability properties are associated to th e copolymer composition, indicating that a 50 % PEO content as well as a high PPO molecular we ight are ideal for the synthesis of capsular systems. Regarding PEC chitosan-alginate systems, only multiple-layer coated copper-alginate particles presented deformability properties under micropipette aspiration after removal of crosslinking ions. An ion exch ange process along with the PEC layers kept the pre-crosslinked Cu-alginate particles stable. Instability of Ca-a lginate particles, subse quent to chelation of calcium ions, was independent of the number of PEC layers. Part icle strength after ion removal can be enhanced by multiple coatings with low molecular weight (high percent deacetylation)

PAGE 55

55 chitosan oligosaccharide. Further studies need to be conducted to compare the effects of various divalent cations on the rheological properties of PEC chitosan-alginate pa rticles. In addition, new techniques for the synthesis of particles need to be explored so that a size range of 5-10 m is attained. Finally, cell studies are needed to assess the particles immunogenic effects for their application in drug delivery.

PAGE 56

56 Figure 3-1. PLA mol ecular structure. Table 3-1. List of commercially available Pluronics Chemical Name Structure BASF (USA) Trade Name Poloxamer 123 HO(C2H3O)7(C3H5O)21(C2H3O)7H Pluronic L43 Poloxamer 124 HO(C2H3O)11(C3H5O)21(C2H3O)11H Pluronic L44 Poloxamer 181 HO(C2H3O)3(C3H5O)30(C2H3O)3H Pluronic L61 Poloxamer 182 HO(C2H3O)8(C3H5O)30(C2H3O)8H Pluronic L62 Poloxamer 182 LF HO(C2H3O)8(C3H5O)30(C2H3O)8H Pluronic L62 LF Poloxamer 183 HO(C2H3O)10(C3H5O)30(C2H3O)10H Pluronic L63 Poloxamer 184 HO(C2H3O)13(C3H5O)30(C2H3O)13H Pluronic L64 Poloxamer 185 HO(C2H3O)19(C3H5O)30(C2H3O)19H Pluronic P65 Poloxamer 188 HO(C2H3O)75(C3H5O)30(C2H3O)75H Pluronic F68 Poloxamer 212 HO(C2H3O)8(C3H5O)35(C2H3O)8H Pluronic L72 Poloxamer 215 HO(C2H3O)24(C3H5O)35(C2H3O)24H Pluronic P75 Poloxamer 217 HO(C2H3O)52(C3H5O)35(C2H3O)52H Pluronic F77 Poloxamer 231 HO(C2H3O)6(C3H5O)39(C2H3O)6H Pluronic L81 Poloxamer 234 HO(C2H3O)22(C3H5O)39(C2H3O)22H Pluronic P84 Poloxamer 235 HO(C2H3O)27(C3H5O)39(C2H3O)27H Pluronic P85 Poloxamer 237 HO(C2H3O)62(C3H5O)39(C2H3O)62H Pluronic F87 Poloxamer 238 HO(C2H3O)97(C3H5O)39(C2H3O)97H Pluronic F88 Poloxamer 282 HO(C2H3O)10(C3H5O)47(C2H3O)10H Pluronic L92 Poloxamer 284 HO(C2H3O)21(C3H5O)47(C2H3O)21H Pluronic P94 Poloxamer 288 HO(C2H3O)122(C3H5O)47(C2H3O)122H Pluronic F98 Poloxamer 331 HO(C2H3O)7(C3H5O)54(C2H3O)7H Pluronic L101 Poloxamer 333 HO(C2H3O)20(C3H5O)54(C2H3O)20H Pluronic P103 Poloxamer 334 HO(C2H3O)31(C3H5O)54(C2H3O)31H Pluronic P104 Poloxamer 335 HO(C2H3O)38(C3H5O)54(C2H3O)38H Pluronic P105 Poloxamer 338 HO(C2H3O)128(C3H5O)54(C2H3O)128H Pluronic F108 Poloxamer 401 HO(C2H3O)6(C3H5O)67(C2H3O)6H Pluronic L121 Poloxamer 402 HO(C2H3O)13(C3H5O)67(C2H3O)13H Pluronic L122 Poloxamer 403 HO(C2H3O)21(C3H5O)67(C2H3O)21H Pluronic P123 Poloxamer 407 HO(C2H3O)98(C3H5O)67(C2H3O)98H Pluronic F127

PAGE 57

57 Figure 3-2. Alginate molecular structure. Figure 3-3. Chitosan molecular structure.

PAGE 58

58 Table 3-2. List of Plur onic-PLA particles made. Pluronic-PLA Ratio (Plu ronic/PLA) Pluronic Solution Concentration F127-PLA 1:3 1.0 % w/v F127-PLA 1:2 1.0 % w/v F127-PLA 1:3 5.0 % w/v F127-PLA 1:2 10 % w/v F127-PLA 1:3 10 % w/v F127-PLA 1:3 20 % w/v F127-PLA 1:2 20 % w/v F127-PLA 1:1 5.0 % w/v L101-PLA 1:2 20 % w/v L101-PLA 1:3 20 % w/v F68-PLA 1:2 20 % w/v P105-PLA 1:2 20 % w/v P103-PLA 1:2 20 % w/v P104-PLA 1:2 20 % w/v F108-PLA 1:2 20 % w/v

PAGE 59

59 Table 3-3. List of chitosa n-alginate particles made Sodium Alginate Solution CaCl2 or Cu(NO3)2 Solution Number of Coatings Chitosan or COS Solution Alginate Solution 0.5 % w/v 0.5 % w/v 0.5 % w/v 1.0 % w/v 0.5 % w/v 5.0 % w/v 1.0 % w/v 0.5 %w/v 1.0 % w/v 1.0 % w/v 1.0 % w/v 5.0 % w/v 3.0 % w/v 0.5 % w/v 3.0 % w/v 1.0 % w/v 3.0 % w/v 5.0 % w/v 0 3.0 % w/v 5.0 % w/v 1 0.3 % w/v 3.0 % w/v 5.0 % w/v 5 0.3 % w/v 0.3 % w/v 3.0 % w/v 5.0 % w/v 9 0.3 % w/v 0.1 % w/v 3.0 % w/v 5.0 % w/v 0 3.0 % w/v 5.0 % w/v 1 1.0 % w/v 3.0 % w/v 5.0 % w/v 5 1.0 % w/v 0.1 % w/v 3.0 % w/v 5.0 % w/v 9 1.0 % w/v 0.1 % w/v 3.0 % w/v 5.0 % w/v 0 3.0 % w/v 5.0 % w/v 1 3.0 % w/v 3.0 % w/v 5.0 % w/v 5 3.0 % w/v 0.1 % w/v 3.0 % w/v 5.0 % w/v 9 3.0 % w/v 0.1 % w/v 3.0 % w/v 0.5M 0 3.0 % w/v 0.5M 1 0.3 % w/v 3.0 % w/v 0.5M 5 0.3 % w/v 0.1 % w/v 3.0 % w/v 0.5M 9 0.3 % w/v 0.1 % w/v 3.0 % w/v 0.5M 0 3.0 % w/v 0.5M 1 1.0 % w/v 3.0 % w/v 0.5M 5 1.0 % w/v 0.1 % w/v 3.0 % w/v 0.5M 9 1.0 % w/v 0.1 % w/v 3.0 % w/v 0.5M 0 3.0 % w/v 0.5M 1 3.0 % w/v 3.0 % w/v 0.5M 5 3.0 % w/v 0.1 % w/v 3.0 % w/v 0.5M 9 3.0 % w/v 0.1 % w/v

PAGE 60

60 Figure 3-4. Optical micrograph of 20% Pluronic-PLA (1:2) microspheres A) F127-PLA, B) L101-PLA, and C) F68-PLA. Original magnification = 100X; bars denote 50 m. A B C

PAGE 61

61 Figure 3-5. Optical micrograph of 20% P105-PLA (1:2) microspheres. Original magnification = 100X. Figure 3-6. Scanning electron micrograph of 20% P105-PLA Particles. A) 400X, bar denotes 100 m; B) 1000X, bar denotes 50 m; and C) 800X, bar denotes 50 m. Table 3-4. Composition and deformability pr operties of Pluronic-PLA microspheres Pluronic-PLA (Ratio) Max Negative Suction Pressure x 103 (dynes/cm2) Deformation F127-PLA (1:3) -58.8 No F127-PLA (1:2) -58.8 No L101-PLA (1:2) -58.8 No L101-PLA (1:3) -58.8 No F68-PLA (1:2) -44.1 Only the shell P105-PLA (1:2) -24.5 Only the shell P103-PLA (1:2) -58.8 No P104-PLA (1:2) -58.8 No F108-PLA (1:2) -58.8 No A B C

PAGE 62

62 Figure 3-7. Optical and scanni ng electron micrographs of 0.3% COS Ca-crosslinked alginate particles: A) 400x, 5 coatings; B) 1000x, 1 coating; C) 1000x, 5 coatings; and D) 1000x,9 coatings. Bars denote 50 m. Figure 3-8. Optical and scanni ng electron micrograph of 3% COS Cu-crosslinked alginate particles, 5 coatings: A) 400x and B) 1000x. Bars denote 50 m. A B D C A B

PAGE 63

63 Table 3-5. Composition and deformability properti es of PEC chitosan-alginate microcapsules. Code Max Negative Suction Pressure x 103 (dynes/cm2) Deformation Ca-0 0 Disintegrated Ca-1 0 Disintegrated Ca-5 0 Disintegrated Ca-9 0 Disintegrated Cu-0 0 Disintegrated Cu-1 -19.6 Yes. Some particles ruptured Cu-5 -19.6 Yes. Particles completely aspirated Cu-9 -19.6 Yes. Portions of the particles aspirated

PAGE 64

64 CHAPTER 4 SYNTHESIS AND CHARACTERIZATION OF PHYSICO-CHEMICAL AND RHEOLOGY PROPERTIES OF POLYELECTROLYTE CHITOSAN-ALGINATE MICROPARTICLES CROSSLINKED WITH CALCIUM, ZINC, OR COPPER IONS Introduction Microencapsulation techniques ha ve been widely used to de velop controlled-release drug delivery systems, masquerade tastes and odors, reduce toxicity, and protect cells from the host immune response in the abse nce of immunosuppression drugs [ 45, 46 ]. Some of these techniques include emulsion solv ent evaporation, so lvent extraction, coacervation, spray-drying, interfacial complexation, co ating, and hot melt coating [ 46, 51 ]. Each method has both advantages and disadvantages in the elaboration of pol ymeric microparticles. The most frequently used method for the fabric ation of alginate particles is the air-spray crosslinking technique. Briefl y, this method consists of sprayi ng an alginate solution into a collecting bath containing the cros slinking agent. As soon as the alginate droplets are in contact with the crosslinking solution, particles are formed by gelation. The system for the development of alginate microspheres can vary from very simple to more complex setups. The spray approach can be attained by using a spray bottle or a more complicated micro-fluid device. The latter is based on the extrusion of the alginate solu tion through an inner lumen while air or a secondary solution is flowing on th e outer lumen of the system. Alginate drops forming at the ending tip of the inner lumen are detached by the air shear forces exerted on them. The other method used to prepare alginate pa rticulates is based on an emulsion-gelation technique. The approach involve s mixing of the alginate soluti on with an organic solvent to create a water-in-oil (W/O) emulsion, followed by th e addition of the cross linking solution. This technique can be varied by mixi ng the corsslinking solution with the organic phase before creating the W/O emulsion. Even though both methods are fairly simple, cost effective, and up-

PAGE 65

65 scalable for mass production to some degree; they also have drawbacks which limit their application. This study is focused on the synthesis and ch aracterization of polye lectrolyte complex (PEC) chitosan-alginate micropartic les. Deformability, particle size, and toxicity are the main goals of the study. Several vari ables, i.e. cross-linking agen ts, solvents and manufacturing techniques, were studied to dete rmine the set of parameters that would yield the most promising red blood cell analog, in terms of deformable capab ility. In addition to deformability, the aim was to obtain particles with a size range between 5 and 10 micron so that they are small to deform and pass through capillaries as well as spleen fenestrations, wh ile having a great drug loading potential compared to the existi ng nano-size drug delivery particulates. In this study we examined the application of two frequently used techniques in the development of alginate particles. The air-spra y crosslinking technique and an emulsion gelation technique were applied for the synthesis of part icles. Besides discerning the most appropriate manufacturing process, the eff ect of different crosslinkers in the development of PEC microparticulates was also studied. Calcium, zinc and copper were the divalent cations used as crosslinking agents and their effect on pa rticles morphology, deform ability, stability, and toxicity was analyzed. Materials and Methods Preparation of PEC Particles Air-spray crosslinking technique Natural polymeric microspheres were prepared by an air-sprayed crosslinking technique [ 93 ]. The polymer used was alginate from Keltone (LV, food grade). From preliminary experimentation, an initial con centration of 0.5M calcium chlori de, zinc chloride and copper nitrate solutions were used as the crosslinki ng baths. Sodium alginate was dissolved in

PAGE 66

66 deionized water at a 3% (w/v) concentration. A double-lumen devi ce was used to air-spray the alginate solution into the stirred gelling bath. To control the si ze of the droplets, the following parameters were studied: needle gauge, alginate solution flow rate, air pressure, and distance between the needle tip and the gelling bath. While studying one parameter, all the others remained constant in addition to algi nate and gelling ion concentrations. The alginate solution was extruded through a syri nge needle (gauge range: 16G 30G) at a constant rate of 1.2, 6, or 12ml/h using a Ha rvard dual-syringe pump (Harvard Apparatus, Holliston, Massachusetts). Air was infused through the outer lumen at constant pressures ranging from 40 to 60 psi, in 10-psi increments. Finally, the drop fall distance was varied from 5 to 15 cm, in increments of five. After alginate droplets were extruded in to the crosslinking bath, they were stirred at room temperature fo r 10 minutes and cured overnight. Alginate microcapsules were then collect ed by centrifugation. Particles were washed three times with deionized water (dH2O) and stored at 4 C for future coating. W/O emulsion gelation technique Sodium-alginate was dissolved in dH2O at a 3% (w/v) concentra tion and emulsified in an organic phase containing soybean oil or cyclohexane (Aldrich, St Louis, MO) and surfactant at different concentrations (0.5 and 5.0 %, w/ w). Pluronics L61, L121, L101, L64, L43 (kindly donated by BASF) and Tween 80 were the surfacta nts tested. The first emulsion was obtained by ultrasonication of the two phases at 60W for 1 min in an ice bath. A second aqueous solution containing 0.5M of the crosslinking agent was added to the emulsion by air-spray (40 psi, 20ml/hr) at a 4-cm dropping distance while stir ring the whole medium sl owly with a magnetic stirrer. Particles were allowed to cure for ten minutes under continuous stirring. Then medium was allowed to rest for 24 hr so that particles would drop to the bottom of the container while the oil phase was left at the top. After separation of the two phases, particles were collected by

PAGE 67

67 filtration through a 45m mesh and washed copiously to remove the organic solvent. When soybean oil was used, the bead slurry was coll ected by centrifugation; wa shed ten times with 10% ethanol, once with 0.1% Triton solution and three times with dH2O. When cyclohexane was used, the particle solution was warmed at 37C to induce solvent evaporation. After removal of the organic phase, particles were st ored at room temperature for future coating. Coating of alginate particles Ca-, Zn, and Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption technique. The coating always starte d with the cationic polymer and ended with the anionic polymer. Microcapsules with zero, two, six, and ten altern ating layers were fabricated. The cationic polymer used was a high-deacetylated, low-molecular-weight chitosan oligosaccharide lactate (Mw < 5000, 90% deacetylati on) from Aldrich. Algi nate was the anionic polymer used. A 3 % (w/v) ch itosan solution containing CaCl2 or a 0.1 % (w/v) alginate solution was added to crosslinked-alginate particles disp ersed in deionized water. After allowing the particles to coat for 30 minutes, they were co llected by centrifugation and washed three times with deionized water to ensure that all free po lyelectrolytes were removed. Following the washes, particles were ready for the next coating or stored at 4C until use. Table 4-1 shows the different batches made; each ba tch was repeated three times. Decomposition of the particles core Decomposition of the particles core was achie ved by incubating particles in cell culture medium, DMEM/F12 50:50 or RPMI 1640, (Cellg ro, Herndon, VA) supplemented with 5% (w/v) bovine serum albumin, BSA (Sigma, St. Louis, MO). Multiple-layer particles, at a 5% (w/v) concentration, were incubated at room temp erature in the two type s of chelating media, differing in their electrolyte concentrations. Af ter close observation, old medium was replaced by same amount of fresh medium if core wa s not dissolved. Medium was replaced until

PAGE 68

68 decomposition of the core was achieved. Gel-core particles were then stored at 4C until further use. Characterization of PEC Particles To characterize the efficacy of the two manuf acturing processes used, analyses of the particles surface morphology and size distribution were carried out. To characterize the effects of using calcium, zinc, and copper as crosslinker ag ents in the development of alginate particles, more tests were performed, including: surface morphology analysis, particle size distribution, deformability assessment, particle stability and viscosity, coating adsorption, ion exchange, and biocompatibility. Surface morphology analysis Surface morphology and stability of the micro capsules were examined by light microscopy and scanning electron microscopy (SEM). Sa mples for scanning electron micrographs were obtained after 0, 2, 6, and 10 coatings. Droplet s of microsphere solutions were mounted on aluminum stubs, let air dried a nd sputter-coated with gold and palladium particles. The stubs were mounted in a scanning electron micros cope at 10.0 kV and imaged at x500, x1000, and x5000. Particle size distribution Size of microcapsules was analyzed by disper sing particles in deionized water at a concentration of 0.1% (w/v). Measurements were carried out by a B eckman LS13320 Particle Characterization Coulter (Beckman Instruments, Fulle rton, CA). Calculation of the particle sizes was carried out using the sta ndard modus of the LS13320 Particle Size Analyzer software (Beckman Instruments, Fullerton, CA). Percentage of particle diameters was used to describe particle size. Each sample was measured in triplicate.

PAGE 69

69 Particle stability analysis To determine the particles stability, different approaches were used. Initially, particles were suspended at a concentration of 5% (w/v) in different types of media: dH2O (negative control), regular RPMI 1640 and DMEM/F 12 50:50 cell culture media, or 5 % BSA supplemented RPMI 1640 and DMEM/F12 50:50 media (Table 4-2) [ 12 ]. Particles were incubated at 25C or 37C under c ontinuous orbital rotation to ensure constant mixing. At 0, 0.5, 1, 2, 4, 8, 12, 24, and 48 hours; samples were collect ed, visually inspected and subjected to a partial vacuum pressure. Briefly, a 5-mL air-displacement pipetter and tip (Eppendorf, Westbury, NY) set at the maximum volume were used to aspirate the particles. For this study, large particles with a diameter ranging between one and two millimeters were used to facilitate optical inspection. The total number of intact part icles was represented as the percentage of the total number of particles inspected. Another approach used to determine stabilit y of the particles involved centrifugation and optical inspection. The study was carried out us ing the Eppendorf Mini Spin Microcentrifuge (Eppendorf, Westbury, NY) which provides speeds of up to 14,000g. Particles weight was obtained to determine the centrifugal force exerte d on them. For this study large particles and microparticles were incubated at 25C in dH2O, DMEM/F12 50:50, or 5 % BSA DMEM/F12 50:50 cell culture medium. After 0.5, 1 and 7 days, the 5 % (w/v) particle suspensions were centrifuged for 20 sec or 5 min at centrif ugal speeds ranging from 1000 13,000 rpm (67 11,337g). Particles were inspected prior and post centrifugation; cha nges in morphology and ability to re-disperse were reported. The other method used to verify particles stability involved using the Wells-Brookfiled Cone/Plate Digital Viscomet er System (Brookfield, Stought on, MA) with a CP-52 conical spindle. Microparticles suspended in 5 % BS A DMEM/F12 50:50 cell culture medium at a 12 %

PAGE 70

70 (w/v) density were subjected to a shear rate of 200 sec-1 for one minute. At the end of each run, samples were collected for light microscopic vi sualization. All analyses were conducted in triplicate. Measurement of particles deformability The micropipette aspiration technique was applied to assess particles deformability. Micropipette tips with inside diameters approxim ately 20% smaller than the particles diameter were fabricated. Previous to their use, pipettes were backfilled with 0.9% NaCl solution using a plastic syringe with a 97-mm long, 28G backf iller (Microfil, World Precision Instruments, Sarasota, FL). Before particles were used for deformability studies, they were incubated in sequestrant-containing media until core was decompos ed. The prepared chamber was then filled with microparticles suspended in sequestering me dia and mounted on the microscope stage. An initial negative pressure of 5 cm H2O (approximately -4.9x103 dynes/cm2) was applied and pressure was augmented in increments of fi ve. After a section of the particle had been aspirated, the negative pressu re was decreased and the par ticle was unloaded. The whole aspiration process was monitored and recorded us ing an inverted light microscope (Axiovert 100, Zeiss), a Carl Zeiss video camera system, and a VCR. The recorded images were converted into digital images using the Matrox software. Since the main goal of the study was to attain particles capable to deform under micropipette suction pressures us ed to aspirate erythrocytes [ 96 ], particles were considered deformable only when using pressures up to -40x103 dynes/cm2. Assessment of particles viscosity The viscosity of the particles in si mulating plasma medium (DMEM/F12 50:50 supplemented with 5% BSA) was measured us ing the Wells-Brookfiled Cone/Plate Digital Viscometer System (Brookfield, Stoughton, MA) with a CP-52 conical spindle. The principle of the system is based on the rotation of the conica l spindle at an accurate speed and detection of

PAGE 71

71 the torque needed to overcome the viscous resi stance caused by the sample fluid between the cone and a fixed flat plate [ 103 ]. Since the cone/plate viscomet er exerts a shear rate on the particle suspension, this test in combination with light microscopy was also used to determine particle stability. Particles were suspended in simulating plasma media at a 12% (w/v) density and incubated at room temperature for 2 hours pr ior to obtaining the rheological measurements. After filling the chamber with the required 0.5-mL sample volume, measurements were taken at a fixed spindle speed of 100 rpm. At the end of each r eading, samples were collected for visualization of the particles stability. Viscosities were cal culated using the formulas and ranges found in the instruction manual: Factor = Range/100 Viscosity = Display Reading x Factor where, the range is specific to the cone and speed used. For the cone CP-52 rotating at a constant speed of 100 rpm, the range is 983 cps (983 mPasec) while the shear rate exerted on the fluid sample is 200 sec-1. Coating adsorption analysis Adsorption of each polyelectrolyte layer ont o microcapsules was examined by measuring the particle surface charge changes. Samples were dispersed in dH2O at a concentration of 0.1% (w/v) and measurements were carried out by the Brookhaven ZetaPlus Analyzer (Brookhaven Instruments Corp.,USA). The zeta potential wa s calculated from the solution conditions and the measured electrophoretic mobility. Each sample was measured in triplicate and the values reported were the mean value for the three replicate samples.

PAGE 72

72 Ion exchange analysis This analysis was only performed for Cu-crosslin ked particles due to the toxicity concerns regarding this transitional metal. To determine the amount of copper ions removed by albumin, a modified bicinchoninic acid (BCA) protein a ssay was used. Cell culture media, RPMI 1640, supplemented with 1% (v/v) penicillin/strepto mycin and 5% (w/v) bovine serum albumin (BSA) was used. Microparticles were suspended in su pplemented media at a concentration of 5% (w/v). Particles were incubated at 25C under continuous orbital rotatio n to ensure constant mixing. At 24, 48, and 72 hours, samples were re moved from the incubator and centrifuged at 3,000 rpm for 5 min. Supernatant was then removed and stored at 4C for future analysis. The removed solution was replaced with an e qual volume of fresh supplemented media. Sample tubes containing microspheres were retu rned to the rotating incubator at previous temperatures until the next time point. Analysis of the stored supernatant was conducted using the bicinchoninic acid reagent from a BCA kit (Pie rce, Rockford, IL). A purple-colored reaction product was yielded by this assay with a str ong absorbance at 562 nm. Concentrations of albumin-reduced copper cations were determined by comparison to a standard curve. All analyses were conducted in triplicate. Assessment of particle cytotoxicity Human dermal fibroblasts (HDF, passage 10) from ATCC (Manassas, VA) were used to determine the level of toxicity of uncoated and coated calcium, zinc-, and copper-crosslinked particles. The type of test performed was indi rect meaning extracts fr om the particles were added to the cells instead of adding particles dir ectly. To determine cell survival and recovery, the MTT cell proliferation assay (Promega, Madison, WI) was used which measures the ability of cells to convert the tetr azolium compound MTT by the action of succinate dehydrogenase to water-insoluble formazan crystals. Two time po ints were studied: a short-term test to

PAGE 73

73 demonstrate particles toxic effects on cells a nd a long-term test to demonstrate survival, the retention of cell re generative capacity. Previously made particles (all batches from Table 4-1) were dried so that a predetermined particle concentration could be obtained. Different concentrati ons of copper powder were used as positive controls. Particle s and controls were incubated in DMEM (Cellgro, Herndon, VA) cell culture medium supplemented with 5 % fetal bovine serum (FBS) and 1 % antibiotic/antimycotic solution at 37C under continuous orbi tal rotation for 24 hr. Samples were then removed from the incubator and centrifuged at 12,225g for 10 min. The supernatant was collected, sterilized by membrane filtration (0.2 m; Whatman), and stored at 80C for future analysis. HDF cells were seeded onto 96-well plates (Costar, Corning, NY) at a 1,000 cells/well density. Initially cells were incubated in DM EM cell culture medium supplemented with 10 % FBS and 1 % antibiotic/antimyco tic (Ab/Am) solution. Cells were incubated at 37C, in a 95 % O2/5 % CO2 atmosphere for 24 hr before adding any treatment. Extracts from the particles and positive contro ls were thawed and d iluted in 2 % FBS and 1 % Ab/Am cell culture medium to desired concentrat ions. The particle and control concentrations used were: 0.1, 1.0, 10, 100, and1000 g/ml (w/v). Cells incubate d in 10 % and 2 % FBS culture medium and exposed to zero treatment were used as negative controls. Old cell medium was aspirated and replaced by 100 l of cell culture medium contai ning treatments. Cells were exposed to the different conditi ons for 24 hr. When exposure time was over, cells were either collected for MTT assay (short-term test) or cult ure medium containing treatments was replaced by 10 % FBS medium (long-term test). For the longterm test, cells were allowed to recover for 2 days before performing the MTT assay. Afte r completion of the cell viability/proliferation

PAGE 74

74 assay, optical densities of each well were measur ed by a microplate reader set at 490 nm. All conditions were done in replicates of six. Data Interpretation Data were expressed as mean values standa rd error of the mean (SEM). To describe statistical differences, one-way analysis of variance (ANOVA) and Tukey-Kramer multiple comparison post test were used. St atistical significance was defined as p 0.05. Results and Discussion The standard method of forming alginate b eads by extruding alginate drops into a crosslinking bath for gelation generates large pa rticles with a diameter range between two and five millimeters. To synthesize smaller alginate particles, two methods were employed: the airspray crosslinking technique and the water-in-o il (W/O) emulsion gelatio n technique. A 3 % (w/v) alginate concentration and crosslinking ion solutions of 0.5M concentrations were used. To control the size of the part icles in the air-spray crossl inking technique, the following parameters were studied: needle gauge, alginate solution flow rate, air pressure, and distance between the needle tip and the gelling bath. For the emulsion-gelation technique, the oil phase was composed of soybean oil or cyclohexane and diverse surfactants; also different stirring rates were used to determine the optimal particle size. Particles made with the air-spray crosslin king technique presente d a monodispersed size distribution depending on the para meter studied. The size of particles diminished significantly when the inner diameter of the needle decreased (Figures 4-1 and 4-3) as expected. Likewise, reducing the speed of alginate ex trusion or increasing the pressu re of air infused through the outer lumen caused a significant reduction in the pa rticles diameter (Figures 4-2, 4-4, and 4-4). The particle-forming properties were not altere d significantly with changes in the distance between the needle tip and the gel ling bath tested in this study. Th ese results are consistent with

PAGE 75

75 many studies from other research groups since it is well-known that the pa rticle size in the airspray method is not only determined by the need le gauge but more importantly by the air flow rate [ 104, 105, 106 ]. Tables 4-3, 4-4, and 4-5 illustrate the particle sizes and standard error of the means under various conditions using different needle gauges alginate flow rates, and air pressures, respectively. The smallest particles were obt ained when alginate was extruded through a 30G needle while the air pressure in the outer lumen was increased to 60 psi. Under these conditions, the smallest average particle size formed was 44 3 m in diameter. Although, we were able to decrease the particle sizes, th e capability to produce monodisperse d batches was lost at the air pressure of 60 psi, in agreement with optic al observations (Figure 4-5). The lost of monodispersity due to in creased outer lumen air pressures co uld be explained by drop coalition at higher pressures. This observation was also reported by Haas mainly for the production of smaller particles [ 107 ]. Another parameter tested in the formation of alginate particles was the usage of different divalent cations as crosslinkers. Figure 4-6 show s a histogram of particle diameters as a function of crosslinking ions obtained with the 30G need le, a 50-psi outer lumen air pressure, and a 1.2mL/h alginate extrusion rate. Using different cross linkers altered signif icantly the size of particles formed (Table 4-6). For alginate partic les crosslinked with copper or zinc, the average particle size obtained was 48 4 and 71 4 m, respectively. This di fference in the mean size of the particles between copperand zinc-crosslinked alginate b eads could be explained by the polymer cation affinity. Even though the crosslin king mechanisms of copper and zinc ions are not understood, it is well-known that the affinity of alginates for di valent cations depends mostly on the electronic structure of the cation, with a higher affinity for copper ions than for zinc ions

PAGE 76

76 [ 85, 100, 101, 102 ]. Copper is able to crosslink more de nsely with sodium alginate, resulting in smaller particles with reduced water content as opposed to zinc-crosslinked alginate particles. Particles made with the emulsion-gelation technique presented a poly-dispersed size distribution, with diameter s ranging from 2 to 60 m (Figure 4-7B). The best results were obtained when using the Pluronic L61 concentra tion of 0.5% and a stirring speed of 2000 rpm. Pluronic L61 is considered a hydrophobic surfac tant with a 10 % (w) PEO content and a PPO molecular weight of 1800 Da. Big agglomerat es were obtained when using other hydrophobic Pluronics: L101 and L121, both with a 10% (w) PEO content and PPO molecular weights of 3300 and 4000 Da, respectively. Indeed, the diff erence in morphology is attributed to the molecular weight of the PPO parts. Hydrophobi c, low-molecular-weight PPO Pluronics are preferred for the production of alginate partic les as opposed to Pluroni c-polyester systems where a high molecular weight of the PPO cont ent is necessary to pr ovide enough anchoring of the copolymer to the part icle surface [36, 64]. Variations in the stirring speed during the gelation proce ss altered the morphology of the particles. An increment in the average of par ticles diameters was observed as the stirring speed was augmented. This observation could be explai ned by an increased coali tion of particles prior to a complete matrix gelation. While the use of soybean oil as the organic phase was very appealing due to its mild condition, complete oil removal was extremely hard to achieve. Although samples were copiously washed, oil was sti ll present in the particles. Microparticles made using cyclohexane as the organic phase pr esented a narrower size distribution with the majority of particles diameters ranging from 5-15 m. Solvent removal was successfully achieved by evaporation.

PAGE 77

77 The particles-forming properties in the emul sion-gelation technique were significantly altered with changes in the cros slinking ions. Particles crossli nked with copper presented a teardrop shape as oppose to the spherical shape of particles crosslinked w ith zinc and calcium (Figure 4-8). In addition to vari ations in the particles shape, the size of the particles was also affected by the different crosslinki ng ions. Table 4-7 illustrates th e fractions of particles within different size ranges. Also, size distribution histograms of Ca-, Zn-, and Cu-crosslinked particles are shown in Figure 4-9. Alt hough the majority of all the microspheres made were 0 20 m in diameter, the smallest uniform microspheres were obtained with calcium a nd zinc. They had the largest fractions, 69.7 9.77% and 77.2 0.50% re spectively, in the size range of 5 10 m; while copper-crosslinked micropart icles had the largest fraction (46.6 0.96%) in the size range of 15 20 m. The difference in particles morphology and si ze was not only dependent on the crosslinker but also on the solvent used as the organic phase of the emulsion. The sizes of Ca-, Zn-, and Cucrosslinked particles were larger when soybean oil was used. However, Cu-crosslinked alginate particles presented the smallest diameters (dat a from Chapter 3) and a spherical morphology. The contradictory results between the use of soybean and cyclohexane could be explained by differences in the solvents viscosities and in the alginates affinity for the cations. Since alginate has the highest affinity for copper ions, the crosslinking r eaction occurs a lot faster than for the other ions. With high viscous solvents, su ch as soybean oil, a fast crosslinking reaction prevents destabilization of the emulsion and, theref ore, less coalition of sm all particles into large ones. In addition, phase separa tion takes more time in highly viscous solvents, which slows down particles settlement into the aqueous phase resulting in spherical particles. On the

PAGE 78

78 contrary, in low viscous solvents (i.e. cyclohexane) copper-crossli nked particles form rapidly and settle quickly into the aqueous phase generating the tear-drop shape. Ca-, Zn, and Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption technique. The coating always starte d with the cationic polymer and ended with the anionic polymer. Microcapsules with zero, two, six, and ten altern ating layers were fabricated. Figure 4-10 shows photographs of alginate b eads made with the conventional dripping mechanism: by extruding alginate as drops into a crosslinking solution for gelation. These 2 3 mm beads presented a spherical shape and size that were independent of th e type of crosslinker used. However, beads mean diameter decrease d with multiple coatings and their shape changed from spherical to biconcave. Zinc-crosslinke d particles showed a more pronounced biconcave configuration as opposed to calci umand copper-crosslinked partic les. These observations could be explained also by the affinity of algi nate for divalent ions in the order of Pb2+>Cu2+>Cd2+>Ba2+>Ni2+>Ca2+>Zn2+>Co2+>Mn2+>Sr2+ [ 85, 100, 101, 102 ]. The new shape of the coated particles was very similar to the bi concave shape of red blood cells, which represents a great advance in the drug deliver y field. It is well-known that the extraordinary flexibility and respiratory functions of red bl ood cells are attributed to thei r structural characteristics [ 11, 9, 10 ]. The biconcave disk shape of our particles provi des a greater surface area ideal for gas exchange as opposed to the traditional spherical shape of the existing drug delivery particulates. The surface morphology of micropa rticles made with the gelation-emulsion technique was determined by SEM analysis for uncoated and bila yer PEC alginate particles. A difference in morphological structure can be observed among algina te particles formed with calcium, zinc or copper cations. As shown in Fi gure 4-11 (C,D,E,F), the surfa ce morphology of the Znand Cumicroparticles looked smoother before any coati ng was applied. On the contrary, the calcium-

PAGE 79

79 alginate microspheres looked very rough on the su rface even prior to any coating (Figure 4-11 (A,B). This sponge-like structure of the uncoa ted calcium-alginate microspheres could be an artifact created by collapsing of the pore walls due to dehydration. Although the order affinity of alginates for cations has been established as Cu2+>Ca2+>Zn2+, binding sites for these cations are different [ 108 ]. Zinc is able to crosslink less selec tively than calcium and hence produces more extensive crosslinking of alginate [ 109, 108 ]. As a result, calcium cations generate a more permeable alginate matrix with a high water co ntent that is more susceptible to morphological changes after dehydration. Bilayer PEC microparticles pr esented a very rough surface morphology with some aggregations which is attributed to the PEC coatings. Size distribution histograms of Ca-, Zn-, and Cu-crosslinked uncoated and coated microparticles are shown in Figure 4-12. Altho ugh the majority of all the microspheres made were 0 20 m in diameter, the size distributions ch anged significantly with the number of coatings. Uncoated calciumand zinc-alginate microspheres had a very narrow size distribution with more than 70% of the particles in the 5 10 m range; while coated particles presented a more polydispersed size distribut ion. Differences in size di stribution between uncoated and multiple coated microparticles could be a consequence of particle agglomeration which increased with the number of coatings. Stability of the particles was determined by the three different methods previously described. Initially, particles were suspe nded at a concentration of 5% (w/v) in dH2O, RPMI 1640 and DMEM/F12 50:50 cell culture media supplem ented with different concentrations of BSA. Particles were incubated at 25C or 37C under continuous or bital rotation and at predetermined intervals samples were collected, visually inspect ed and subjected to a partial vacuum pressure. All particles incubated in dH2O looked intact independent of the number of

PAGE 80

80 coatings or crosslinkers at the end of the study. Uncoated Ca-, Zn-, and Cu-alginate particles dissolved after one hour incubati on in both 5% BSA cell culture media at 37C (Figure 4-13). By day 1, all calcium-alginate uncoated and coated particles were dissolve d regardless of the cell culture medium (Figure 4-13 A, B). It is apparent from the histograms shown in Figure 4-13 (C,D,E,F) th at the stability of zincand copperalginate particles depended on the number of coatings and the incubation medium. Stability of these part icles was directly proportional to the number of coatings. Also, particles were more stable in DMEM/F12 50:50 cell culture medium than in RPMI 1640. These results were expected due to the electrolyte c ontent difference of each media. DMEM/F12 50:50 electrolyte content is very similar to human plasma (Table 4-2); however, RPMI 1640 has approximately a 6-fold reduction of Ca2+ and a 10-fold excess of phosphate ions. This difference in ions is detrimental for the stability of PEC algi nate particles. The ion exchange properties of alginate gels have been studied by many rese arch groups and the proc ess depends on several factors such as: the amount of crosslinking ions forming the gel, physical properties of the gel, competing ion concentration, pH and ionic strength of the solution [ 97, 98, 99 ]. After removal of crosslinking ions from th e particles by albumin and phosphate ions present in the RPMI 1640 cell cu lture medium, the polyelectrolyte complex system was not stable anymore. The calcium PEC system colla psed regardless of the number of PEC layers because alginate had no affinity for the other elec trolytes present in the medium. The zinc and copper PEC systems were stable slightly longer because of the alginate affinity for the cations and the more extensive crosslinking of these systems. However, the lack of calcium ions that would replace the alginate binding sites left by th e previous crosslinkers accelerated the collapse of the systems. On the contrary, when zinc and copper ions were chelated by albumin in

PAGE 81

81 DMEM/F12 medium, the alginate bi nding sites left were occupied by calcium ions present in the cell culture medium. As a result, the ion exchan ge process along with the PEC layers kept the pre-crosslinked Znand Cu-alg inate particles stable. Figures 4-14 and 4-15 illustrate how other factors alter the ion exchange process in alginate gels. Stability of the particles is directly propor tional to the concentratio n of albumin present in the medium (Figure 4-14), as e xpected. Temperature is another important factor influencing the ion exchange process. The kinetics of ion ex change in our chitosan-a lginate PEC system was slower at 25C than at 37C. Although our re sults were expected, we cannot determine the equilibrium and kinetics of the ion exchange process taking place in the PEC coated zinc-and copper-alginate microparticles at this time. Prev ious studies on the kinetics of metal ion uptake by alginates showed that uptake begins with a rapid phase (minutes to hours) followed by a relative slow phase (up to one day)[ 97 ]. There are many models that describe the equilibrium and kinetics of ion exchange in alginate gels such as the Langmuir, Freundlich, and the surface complex formation models [ 97 ]. Although each of the models has advantages and disadvantages, it may be possible to use one of them to determine the equilibrium of our system. The other two approaches used to determ ine stability of the particles involved centrifugation and shear followed by optical inspection. Znand Cu-a lginate particles with zero, two, six, or ten PEC coatings were incubate d in 5% BSA DMEM/F12 50:50 medium at a 5% (w/v) density overnight. Particles were then exposed to centrifugal speeds ranging from 1000 13,000 rpm (67 11,337g) for 20 seconds or 5 minutes Uncoated Zn-alginate particles failed when centrifuged at 1000 rpm for 20 sec; while uncoated Cu-alginate particles failed at 4000 rpm. All the coated zincand copper-alginate part icles sustained all the ce ntrifugal speeds, even at the high speed of 13,000 rpm for 5 min. Figures 4-16 and 4-17 illustrate optical micrographs

PAGE 82

82 of zincand copper-alginate PEC micropart icles exposed to a shear rate of 200 sec-1 for 1 min. Microparticles had been suspended in 5 % BSA DMEM/F12 50:50 medium at a 12 % (w/v) density prior to testing. Uncoated zinc partic les were completely destroyed (Figure 4-16A); while pieces of very swollen uncoated copper pa rticles could still be seen (Figure 4-17A). Differences between the two, six, and ten multilayer particles were not apparent from Figures 416 (B,C,D) and 4-17 (B,C,D). Although some of the bilayer zinc part icles looked slightly swollen (Figure 4-16B), the difference was not significant. The observations from the centrifugation and shear tests coin cide with our previous results showing again the strength of copper-alginate gels over the zinc-a lginate ones. In addition, it is evident the stabilizing role of PEC coatings in both systems. The micropipette aspiration technique was used as previously described to determine deformability of the calcium-, zinc-, and copperalginate PEC particles. Concurring with our previous results, all coated algi nate particles suspended in dH2O were not deformable when using the micropipette. However, when cross linking ions were removed, zinc and copper particles presented deformable properties, while calcium particles disintegrated regardless of the number of PEC coatings. As Table 4-8 and Fi gure 4-18 illustrate, zinc and copper particles deformed and were completely aspirated into the micropipette tip af ter applying a 20-cm-H2O negative pressure (-19.6x103 dynes/cm2). While uncoated particles were not stable af ter removal of cross linking ions; bi-layer particles were stable mostly in sta tic conditions. Less pressure (-15 cm H2O; -14.7x103 dynes/cm2) was necessary to aspirate PEC bilayer zinc -crosslinked microcapsules. Some of the bi-layer zinc and copper particles ruptur ed under pressures of -15 and -20 cm H2O, respectively. Best results were obtained with microcapsules coated six times. Hexa-layer zinc and copper

PAGE 83

83 microparticles were able to deform under a 20-cm-H2O negative aspiration pressure several times while recovering their shape. Removal of cr osslinking ions was not as efficient for the tenlayer copper microparticles as fo r the lower-layer ones under the c onditions tested. As a result, only portions of the deca-layer Cu-crosslinked microparticles were deformable under a 20-cmH2O negative pressure as oppose to the 10-PEC-laye r zinc microcapsules which were completely aspirated under the same pressure. These obs ervations agree with our previous results, demonstrating once more the difference in strength of copper-, zinc-, and calcium-alginate gels along with the stabilizing role of the PEC coatings. Viscosity of the particles in simulating plasma medium (DMEM/F12 50:50 supplemented with 5% BSA) was measured using the Wells-Br ookfiled Cone/Plate Digita l Viscometer System with a CP-52 conical spindle. Zinc and coppe r microparticles were suspended in simulating plasma medium at a 12% (w/v) density and incubated at room temperature for 2 hours prior to obtaining the rheological measurements. Measurem ents were taken at a fixed spindle speed of 100 rpm and a shear rate of 200 sec-1. Viscosity values for blood, plasma, and serum were obtained from studies done in healthy adults by Rosenson et al [24]. A histogram of the samples viscosity is given in Figure 4-19 and Table 4-9 demonstr ates the viscosity values and standard error of the means obt ained under the conditions tested. It is apparent from the data that viscosity va lues of uncoated zinc and copper microparticle suspensions were significantly higher compared to the viscosity of blood. The uncoated Zn and Cu particle suspensions presented increments in vi scosity of seven and six folds, respectively. It can be seen from Figures 4-16A and 4-17A th at uncoated zinc particles were completely destroyed while there were only remaining pieces of very swollen uncoated copper particles. Rupture of particles followed by a release of alginate molecules into the salt-containing medium

PAGE 84

84 caused an increment of the suspensions viscosi ties. Solutions of bilayer PEC zinc and copper particles had viscosities of 4.92 and 5.90 mPas, respectively. These values were not significantly different from the viscosity values of human bl ood and Figures 4-16B and 4-17B demonstrate the presence of particles in the solutions. While the viscosity average for the solutions containing copper microcapsules with six PEC coatings was 4.92 mPas, six-layer zinc-a lginate microparticle solutions had a high viscosity value of 9.83 mPas for the initial 30 se conds. The viscosity for the latter sample decreased to 4.92 mPas at 60 seconds. The differen ce in viscosity values w ith respect to time in the six-PEC-layer zinc microparticle solutions could be explained by agglomeration of the particles. Figure 4-16C shows some aggregation of hexa-layer Zn pa rticles. In static conditions, particles agglomeration was even higher, causi ng an initial higher resistance to the rotating cone. As the sample fluid was exposed to shear rates of 200 sec-1, particles agglomeration was broken up, decreasing the solution viscosity. The viscosity mean value of the ten-layer copper micropa rticle solution was 8.19 0.87, representing a significant increment compared to the blood viscosity. This result correlates with our previous stability and deformability studi es. The ten multilayer copper-alginate systems formed the strongest microparticles due to the alginate high affinity for copper ions and the increased stability provided by the high number of coatings. Regarding the ten-layer zincalginate microspheres, suspensions containi ng these particles showed an extremely high viscosity value of 26.2 mPas for the initial 30 se conds. Although, viscosity decreased with time to 8.85 mPas at 60 seconds, it was still about 2.7 -fold higher than blood viscosity. Once more, particles agglomeration seemed be the cause of the changes in vi scosity with respect to time. Figure 4-16D shows large aggregates of deca-layer Zn particles.

PAGE 85

85 Adsorption of each polyelectrolyte layer ont o microcapsules was examined by measuring the particle surface charge changes. We first ex amined uncoated alginate particles dispersed in dH2O. There was no significant difference between the zeta potentials of uncoated zincand copper-alginate particles (Table 410). It is apparent from Figure 4-20 that the ze ta potential of microparticles with the chitosan surface layer (o dd numbers) is slightly less negative than microparticles with the alginate surface layer (even numbers). Although the data showed the presence of each polyelectroly te layer, the difference between uncoated and chitosan-surface coated microparticles was not sign ificant until the sec ond chitosan coating fo r zinc particles and the third coating for copper particles. Also, Fi gure 4-20A illustrates that the zeta potential of zinc microspheres with a tota l of eight and ten PEC layers ending in alginate differed significantly compared to uncoated microspheres. It is evident that microbeads retained a negative surface charge throughout the modification of the surface layer. Thickness of the polymeric films on colloidal templates is considered to be in the monomolecular-layer range [ 110 ]. Since the low molecular weight (< 5000Da) chitosan oligosaccharide was used as th e cationic polymer, layer thickness was very thin and the polyelectrolyte molecules failed to overcharge the microbead surface. Therefore, reversing of the microparticle s surface charge did not o ccur. As mentioned earlier, agglomeration of coated zinc-alginate microsphe res increased with the number of layers. Eventually, the formation of aggregates resulted in an irregular coating of the multilayer zinc particles. Since the toxicity of copper ions was a c oncern, we decided to analyze the amount of copper ions released from micropa rticles into the chelating medium To determine the amount of copper ions sequestered by album in a modified bicinchoninic acid (BCA) protein assay was

PAGE 86

86 used. Microparticles were suspended in RPMI 1640 medium, supplemented with 1% penicillin/streptomycin and 5% BSA, at a concentration of 5% (w/v). Particles were incubated at 25C under continuous orbital rotation and supern atants were collected at predetermined intervals. The removed solution was replaced with an equal volume of fresh supplemented medium and samples were returned to the ro tating incubator until the next time point. Concentrations of albumin-reduced copper cations were determined by comparison to a standard curve. Figure 4-21 illustrates the amount of copper rele ased in a period of three days. Only concentrations of copper released from uncoated particles were dete cted. It was determined that a 5% (w/v) suspension of uncoated particles released a total copper concentration of approximately 3 mM in 24 hours (Table 4-11). Although no released copper from uncoated particles was observed on the second day, some amount was detected on the third day. The results obtained for the coated samples were in agreement with visual observations. At day 1, microparticles with two coatings looked very sw ollen as opposed to six-layer particles, while ten-layer beads looked unchanged. At day 2, some of the twoand six-layer particles were disintegrated, whereas micropartic les with ten coatings looked swollen. Since PEC coatings improved the stability of partic les, the ion exchange process took longer as the number of PEC layers increased. Therefore, small amounts of c opper ions were released into the medium during each time point, which were not detected by the m odified BCA assay due to its sensitivity at the microgram level. Since the dose limit of coppe r intake levels in adults is 10 mg/day (157 mol/day) and the prescribed IV dose for patients with copper deficiency anemia is 3 mg/day (47.2 mol/day) [ 111 ], the low amounts of copper released from the 5% (w/v) suspension of coated particles are still below toxic levels.

PAGE 87

87 The cytotoxicity of the uncoated and coat ed calcium-, zinc-, and copper-crosslinked particles was assessed using an in vitro cell proliferation assay. Hu man dermal fibroblast cells were exposed to extracts from the particles and the MTT assay was used to determine cell survival and recovery. Figure 4-22 shows hist ograms of the cell su rvival and recovery percentages as a function of extr act concentration. Absorbance readings from cells incubated in 2% FBS medium were used to calc ulate the survival and recovery percentages. Cells exposed to the extracts from calcium partic les showed up to a 2.5-fold increment in cell growth independent of the coating number or the extract concentrat ion (Table 4-12; Figure 4-22A). Likewise, Figure 4-22B demonstrates that a 24-hour exposure of calcium particle extracts did not alter cell proliferation in the long term. While cells exposed to supernatants from the coated zinc particles showed also up to a 2.5-fold increase in cell pr oliferation, cells exposed to uncoated particles extracts at the 100 and 1000 g/mL concentrations did not surv ive (Figure 4-22C). Similar to calcium, exposure to zinc particle supernatants did not affect cell prolif eration in the long term, except for cells in contact with extracts from uncoated zinc particles at the two highest concentrations (Figure 4-22D). When cells were incubated in copper part icle supernatants, a 2.5-fold increase in proliferation was seen only for cells exposed to the 0.1 g/ml concentration of coated particles extracts (Figure 4-22E). Alt hough Table 4-12 illustrates a higher proliferation for cells in contact with supernatants from co ated particles as opposed to the c ontrol, it is apparent that cell proliferation decreases with the extract concentration. In additi on, cells exposed to the highest concentration of uncoated particle supernatants did not survive. In contrast to cells incubated at high doses of uncoated zinc partic le extracts, cells exposed to toxic levels of uncoated copper supernatants were able to slowly recover w ithin two days (Figure 4-22F). The long-term

PAGE 88

88 histogram shown in Figure 4-22F also demonstrates that proliferation of cells exposed to all the other extracts was not compromised. These results are in agreement with our previous observations for the particle stability and the ion exchange process of our PEC chitosan-algina te particles. It has been shown repetitively that particle stability is enhanced by the numbe r of PEC coatings, except for calcium-crosslinked particles. We have also demonstrated that the copper-alginate matrix is the strongest and, therefore, forms the most stable particles. As a result, PEC layers not only keep the zinc and copper particles stable but also decrease the rele ase rate of crosslinking ions into the medium, which explains differences in cell toxicity be tween the uncoated and co ated particles. In addition, higher toxicity is obser ved in uncoated zinc particles due to less alginate affinity for this cation as opposed to Cu2+. It is apparent from the reco very histograms that the percent values are lower, which coincide with the va lues obtained for cells incubated in 10% FBS medium, our negative control (Tab le 4-12). These observations c ould be explained by the cells reaching confluency, resulting in inhibition of cell pr oliferation due to cell-to-cell contact. Since the MTT assay can only detect mitotic cells, the correct cell number of non-dividing cells cannot be determined through this assay. The use of copper as a crosslinker for algina te gels has always been limited due to its toxicity. However, our results coincide with other research groups which have found that small amount of copper induces cells proliferation, differentiation, and mi gration in vitro [ 112, 113, 114 ]. Hu was able to demonstrate that human um bilical vein endothelial cells incubated in 10% FBS medium supplemented with 500 M copper were able to increase in cell number by 216% compared to a 248% increment induced by 20 ng/ml bFGF [ 112 ]. Also, Rodriguez et al showed that copper at the 50 M concentration decreased the prolif eration rate of human mesenchymal

PAGE 89

89 stem cells while increasing cell differentiation into osteogenic and adipogen ic cell lines. Besides proliferation and differentiation, migration of kera tinocytes was induced by exposure of cells to zinc or copper concen trations of 1.8 or 2 g/mL, respectively [ 114 ]. Conclusion Particles made with air spray crosslinki ng technique presented a mono-dispersed size distribution; however, the smaller particle size obtained was 50 m. Although, particle size decreased with increments in th e outer lumen air pressure, monodi spersity was lost. Particles made with W/O emulsion-gelation technique using cyclohexane as the organic phase presented a narrow size distribution; diameter of uncoated particles ranged from 5-15 m. Microparticles crosslinked with copper presented a tear-drop shape as oppose to th e spherical shape of particles crosslinked with zinc and calci um. Particle size decreased when increasing the number of coatings. After the tenth coating, there was appr oximately a 50% size reduc tion in particles with an initial diameter ranging between 2 3 mm Besides a decreased in size with multiple coatings, particles changed their shape from sphe rical to biconcave. Zi nc-crosslinked particles showed a more pronounced biconcave configur ation as opposed to calciumand coppercrosslinked particles. After removing cross-linkers of coated micr ospheres, zincand copper-alginate capsules were deformable and remained stable using the micropipette techni que under physiological pressures. Best results were obtained with Znand Cu-crosslinke d alginate particles coated six times. In addition, these capsules were stable for a period of up to 10 days in sequestrantcontaining media and sustaine d shear rates of 200 sec-1. Stability of th e zincand copperalginate microparticles was improved by the nu mber of polyelectrolyte coatings. The ten multilayer copper-alginate systems formed the strongest particles due to the alginate high affinity for copper ions and the increased stability prov ided by the high number of coatings. Calcium-

PAGE 90

90 crosslinked particles were not stable after removal of crosslinking ions; even multiple polyelectrolyte coatings di d not improve stability. Viscosity values for solutions containing copper microcapsules with two and six PEC coatings were close to blood viscosity values; while only suspensions of tw o-layer zinc particles showed viscosity values similar to bloods. Co ating microbeads with ch itosan oligosaccharide did not affect the particles surf ace charge as they retained a negative surface charge throughout the modification of the surface layer. From the ion exchange studies, it was determined that a 5% (w/v) suspension of uncoate d copper particles re leased a total coppe r concentration of approximately 3 mM in 24 hours. Copper released from the coated particles could not be determined under the conditions tested. Results from the MTT assay demonstrated that multiple coatings decreased toxicity of heavy-metal cross linked particles. The highest level of toxicity was shown by uncoated copperand zi nc-crosslinked particles at 1000 g/ml. The rest of the samples seemed not to affect ce lls under the conditions studied.

PAGE 91

91 Table 4-1. The PEC chitosan-algi nate particle samples made Batch # Crosslinker # of Coatings 1 Ca 0 2 Ca 2 3 Ca 6 4 Ca 10 5 Zn 0 6 Zn 2 7 Zn 6 8 Zn 10 9 Cu 0 10 Cu 2 11 Cu 6 12 Cu 10 Table 4-2. Concentrations of major electrolyt es (mmol/L) in human pl asma and cell culture media Electrolyte Plasma RPMI 1640 DMEM/F12 50:50 Bicarbonate 24 30 23.81 29.02 Calcium 2.00 2.75 0.42 2.0 Chloride 100 110 108.03 127.96 Magnesium 0.8 1.1 0.4 0.7 Phosphate 0.53 0.90 5.64 0.95 Potassium 4.0 5.6 5.36 4.18 Sodium 130 155 113.95 150.25 Sulfate 0.35 0.75 0.4 0.703 Nitrate --0.84 0.00037

PAGE 92

92 0 100 200 300 400 500 600 700 800 900 30G22G18G16G Needle GaugeParticle Size (um) *** *** ****** p < 0.001 Figure 4-1. Particle size diameter of copper-crosslinked algi nate particles made with the airspray crosslinking technique as a function of the needle gauge. Particles were made with an alginate extrusion rate of 12 mL/h, a 10-cm dr opping distance, and the outer lumen air pressure set at 50 psi. Signifi cant differences between particles made with the 30G needle and all the othe r needle gauges are reported. Table 4-3. Particle size diameter of copper-crosslinked alginate particles as a function of the needle gauge Particle Diameter ( m) Needle Gauge Needle Inner Diameter ( m)MeanSEM 30G 33085.374.39 22G 711132.318.19 18G 1270482.9642.78 16G 1651688.55125.51

PAGE 93

93 0 20 40 60 80 100 120 1.2612 Alginate Flow Rate (mL/h)Particle Size (um) p < 0.05 *** p < 0.001 *** *** Figure 4-2. Particle size diameter of copper-crosslinked algi nate particles made with the airspray crosslinking technique as a function of alginate flow rate. Particles were made using a needle gauge of 30G (330 m), a 10-cm dropping distance, and the outer lumen air pressure set at 50 psi. Signifi cant differences between particles made with 1.2 and 12 mL/h and between 6 and 12 mL/h are reported. Table 4-4. Particle size diamet er of copper-crosslinked algina te particles as a function of alginate flow rate Particle Diameter ( m) Alginate Flow Rate (mL/h) MeanSEM 1.2 85.833.22 6.0 93.223.17 12.0 105.323.61

PAGE 94

94 Figure 4-3. Optical photomi crographs of copper-crosslinked algi nate particles ma de with the airspray crosslinking technique using A) a 1.65-mm (16G) syringe needle and extruding flow rate of 12 ml/hr, and B) a 0.33-mm (30G ) syringe needle and extruding flow rate of 1.2 ml/hr. The dropping distance set at 10 cm and the outer lumen air pressure set at 50 psi. Original magnifi cation = x50, bar denotes 200 m. A B

PAGE 95

95 0 10 20 30 40 50 60 70 80 90 100 40 psi60 psi Air PressureParticle Size (um) 1.2 ml/hr 12 ml/hr p < 0.05 *** p < 0.001* *** *** *** *** Figure 4-4. Particle size diameter of copper-crosslinked algi nate particles made with the airspray crosslinking technique as a function of air pressure. Particles were made using a needle gauge of 30G (330 m) and a 10-cm dropping distance. While testing air pressure or alginate flow rate, all ot her parameters remained constant. Table 4-5. Particle size diamet er of copper-crosslinked algina te particles as a function of alginate flow rate Particle Diameter ( m) Air Pressure (psi) Alginate Flow Rate (mL/h)MeanSEM 40 1.263.584.78 40 12.081.146.16 60 1.247.634.04 60 12.044.503.40

PAGE 96

96 Figure 4-5. Optical photomi crographs of copper-crosslinked algi nate particles ma de with the airspray crosslinking technique: A) air pressure = 40 psi and alginate flow rate = 1.2 mL/h; B) air pressure = 40 psi and alginate flow rate = 12 mL/h; C) air pressure = 60 psi and alginate flow rate = 1.2 mL/h; and D) air pressure = 60 psi and alginate flow rate = 12 mL/h. Original magni fication = x50, bar denotes 200 m. A B C D

PAGE 97

97 0 10 20 30 40 50 60 70 80 CopperZincParticle Size (um) *** p < 0.001*** *** Figure 4-6. Particle size diameter of copper-crosslinked algi nate particles made with the airspray crosslinking technique as a function of the crosslinking ion. Particles were made using a needle gauge of 30G (330 m), a 10-cm dropping distance, the outer lumen air pressure set at 50 psi, and an alginate extrusi on rate of 1.2 mL/h. Table 4-6. Particle size diameter of copper-crosslinked alginate particles as a function of the crosslinking ion Particle Diameter ( m) Crosslinking Ion MeanSEM Copper 47.634.04 Zinc 71.334.41

PAGE 98

98 Figure 4-7. Optical photomicrograp hs of calcium-crosslinked algina te particles made with the air-spray crosslinking technique A) and with the W/O emulsion gelation technique B). A) Original magnification = x50, bar denotes 500 m. B) Original magnification = x100, bar denotes 100 m. AB

PAGE 99

99 Figure 4-8. Optical micrographs of alginate microspheres made with the W/O emulsion gelation technique. Alginate particle s were crosslinked with (A, B) calcium, (C, D) zinc, or (E, F) copper ions. The left row origin al magnification = x100, bars denote 100 m; right row magnification = x200, bars denote 50 m. AB C D E F

PAGE 100

100 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Ca_0_1 Ca_0_2 Ca_0_3 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Zn_0_1 Zn_0_2 Zn_0_3 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Cu_0_1 Cu_0_2 Figure 4-9. Particle si ze distribution. Percentage of to tal number of uncoated microspheres versus particle diameter: A) calcium-, B) zinc-, and C) copper-crosslinked alginate particles. Measurements carried out by a Beckman LS Particle Characterization Coulter. A B C

PAGE 101

101 Table 4-7. Particle size distribution of uncoated Ca-, Zn-, an d Cu-crosslinked alginate particles Fraction (%) Mean SEM Size Range ( m) Ca_0 Zn_0 Cu_0 0.00 4.99 33.4 7.63 5.00 1.12 0.00 0.00 5.00 9.99 69.7 9.77 77.2 0.50 5.14 0.29 10.0 14.9 6.68 2.06 14.4 0.90 33.2 1.72 15.0 19.9 0.28 0.28 3.03 1.04 46.6 0.96 20.0 24.9 0.00 0.00 0.34 0.34 13.4 1.78 25.0 29.9 0.00 0.00 0.00 0.00 1.09 1.09 30.0 35.0 0.00 0.00 0.00 0.00 0.49 0.49

PAGE 102

102 Figure 4-10. Optical photographs of Ca-, Zn-, and Cu-crosslinked alginate particles made with the air-spray crosslinking t echnique. Particles were made using a needle gauge of 16G (1.65 mm), a 10-cm dropping distance, no outer lumen air pressure, and an alginate extrusion rate of 12 mL/h. Particles were coated with 0, 2, 6, and 10 polyelectrolyte layers using the LbL t echnique. The poly cationic solution was composed of 3% chitosan o ligosaccharide and 0.1% CaCl2 while 0.1% alginate solution conformed the polyanionic phase. 0 2 6 10 Ca Cu Zn

PAGE 103

103 Figure 4-11. Scanning electron micrographs of alginate microspheres made with the W/O emulsion gelation technique. Alginate pa rticles were crosslinked with (A, B) calcium, (C, D) zinc, or (E, F) copper ions. The left row displays uncoated crosslinked alginate partic les while the right row disp lays PEC bilayer alginate particles. Magnification = x2500, bars denote 10 m. A B E F C D

PAGE 104

104 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Ca_0 Ca_2 Ca_6 Ca_10 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Zn_0 Zn_2 Zn_6 Zn_10 0 10 20 30 40 50 60 70 80 90 100 0510152025303540 Particle Diameter (um)Frequency (%) Cu_0 Cu_2 Cu_6 Cu_10 Figure 4-12. Particle size distribution of uncoate d and multiple-layer PEC particles. Percentage of total number of microsphere s versus particle diameter: A) calcium-, B) zinc-, and C) copper-crosslinked alginate particles. Measurements carried out by a Beckman LS Particle Characterization Coulter. A B C

PAGE 105

105 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Ca-0 Ca-2 Ca-6 Ca-10 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Ca-0 Ca-2 Ca-6 Ca-10 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Zn-0 Zn-2 Zn-6 Zn-10 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Zn-0 Zn-2 Zn-6 Zn-10 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Cu-0 Cu-2 Cu-6 Cu-10 0 50 100 150 024487296120144168 Incubation Time (hr)Intact Particles (%) Cu-0 Cu-2 Cu-6 Cu-10 Figure 4-13. Stability studies of PEC microparticles. (A, B) calci um-, (C, D) zinc-, and (E, F) copper-crosslinked alginate particles coat ed with 0, 2, 6, or 10 PEC layers were incubated at a 5 % (w/v) density in 5 % BSA supplemented RPMI 1640 (A, C, E) or DMEM/F12 50:50 (B, D, F) cell culture media at 37C under continuous orbital rotation. At predetermined inte rvals; samples were collected, visually inspected and subjected to a partial vacuum pressure w ith a 5-mL air-displacement pipetter. For this study, large particles with a diamet er ranging between one and two millimeters were used to facilitate op tical inspection. The total num ber of intact particles was represented as the percentage of the total number of particles inspected. CD AB E F

PAGE 106

106 Figure 4-14. Optical photomicrographs of bilayer copper-crosslinke d alginate particles incubated 24 hours in A) dH2O, B) 1% (w/v) BSA RPMI 1640, C) 2.5% (w/v) BSA RPMI 1640, and D) 5% (w/v) BSA RPMI 1640. Original magnification = x50, bar denotes 200 m. CD AB

PAGE 107

107 Figure 4-15. Optical photographs of uncoated zinc-c rosslinked alginate particles. Particles were incubated in dH2O A), 5 % BSA supplemented BS A RPMI 1640 B) or DMEM/F12 50:50 C) cell culture media at 25C for 1 hour. ABC

PAGE 108

108 Figure 4-16. Optical micrographs of zinc-crossli nked PEC microparticles e xposed to a shear rate of 200-sec-1 for one minute. A) Zero, B) two, C) six, and D) ten layer microcapsules. Particles were incubated at a 12 % (w/v ) density in 5 % BSA DMEM/F12 50:50 cell culture medium for 4 hrs prior to appl ying shear rate. The Wells-Brookfield Cone/Plate digital viscometer with a CP52 conical spindle was used at a speed of 100 rpm. Magnification = x100 and bars denote 100 m. A B C D

PAGE 109

109 Figure 4-17. Optical micrographs of copper-cros slinked PEC microparticles exposed to a shear rate of 200-sec-1 for one minute. A) Zero, B) two, C) six, and D) ten layer microcapsules. Particles were incuba ted at a 12 % (w/v) density in 5 % BSA DMEM/F12 50:50 cell culture medium for 4 hr s prior to applying shear rate. The Wells-Brookfield Cone/Plate digital viscom eter with a CP-52 conical spindle was used at a speed of 100 rpm. Ma gnification = x100 and bars denote 100 m. A B C D

PAGE 110

110 Table 4-8. Composition and deformability properti es of PEC chitosan-alginate microcapsules Code Max Negative Suction Pressure x 103 (dynes/cm2) Deformation Ca-0 0Disintegrated Ca-2 0Disintegrated Ca-6 0Disintegrated Ca-10 0Disintegrated Zn-0 0Disintegrated Zn-2 -14.71Yes. Some particles ruptured Zn-6 -19.6Yes. Particles completely aspirated Zn-10 -19.6Yes. Particles completely aspirated Cu-0 0Disintegrated Cu-2 -19.6Yes. Some particles ruptured Cu-6 -19.6Yes. Particles completely aspirated Cu-10 -19.6Yes. Portions of the particles aspirated

PAGE 111

111 Figure 4-18. Optical photomicrograph of he xa-layer alginate-chitosan capsule during micropipette aspiration. Or iginal magnification = x400.

PAGE 112

112 0 5 10 15 20 25 30B l ood Plasma S er u m dH2O Media Z n_0 Zn_2 Zn_6 Z n_1 0 Cu_0 Cu_2 Cu _6 C u_1 0Viscosity (mPa.s) 30 sec 60 sec *** *** *** *** * p < 0.05 ** p < 0.01 *** p < 0.001 Figure 4-19. Viscosity values of zincand c opper-crosslinked PEC micr oparticles. Particles (n=3) were incubated at a 12 % (w/v) density in 5 % BSA DMEM/F12 50:50 cell culture media for 4 hrs prior to taking viscosity readings. The Wells-Brookfield Cone/Plate digital viscometer with a CP52 conical spindle was used at a speed of 100 rpm and applying a shear rate of 200 sec-1. Measurements were taken at 30 and 60 seconds. Viscosity values for blood, plasma, and serum were obtained from Rosenson, McCormick and Uretz (p. 1191), Distribution of blood viscosity values and biochemical correlate s in healthy adults Clinical Chemistry 42 (1996), No. 8, 11891195. Significant differences between whole blood and all the samples are reported.

PAGE 113

113 Table 4-9. Viscosity values of PEC chitosan-alginate microcapsu les. Measurements at 30 and 60 seconds were the same for all sample s except for Zn-6 and Zn-10. Viscosity values for blood, plasma, and serum were obtained from Rosenson, McCormick and Uretz (p. 1191), Distribution of blood vi scosity values and biochemical correlates in healthy adults Clinical Chemistry 42 (1996), No. 8, 1189-1195 Viscosity (mPas) Sample MeanSEM Blood 3.260.25 Plasma 1.390.05 Serum 1.270.03 dH2O 0.980.00 5% BSA DMEM/F12 50:50 cell culture media 1.640.16 Zn-0 21.30.66 Zn-2 4.920.00 Zn-6 9.830.00 Zn-10 26.23.28 Cu-0 17.00.33 Cu-2 5.900.57 Cu-6 4.920.00 Cu-10 8.190.87 Zn-6, 60 sec 4.920.00 Zn-10, 60 sec 8.850.00

PAGE 114

114 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 012345678910 Number of CoatingsZeta Potential (mV) p < 0.05 ** p < 0.01 *** p < 0.001 *** *** *** *** ** -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 012345678910 Number of CoatingsZeta Potential (mV) p < 0.05 *** p < 0.001* *** Figure 4-20. Zeta potential values of A) zi ncand B) copper-crosslinked multiple layer PEC particles. Samples were dispersed in dH2O at a concentration of 0.1% (w/v) and measurements were carried out by the Br ookhaven ZetaPlus Analyzer. The outer layer is composed of alginate when the number of coatings is even; while chitosan oligosaccharide is on the su rface of particles when the number of coatings is odd. Significant differences between uncoated pa rticles and all the coated samples are reported. B A

PAGE 115

115 Table 4-10. Zeta potential values of PEC chitosan-alginate microcapsules. Measurements were carried out by the Brookhaven ZetaPlus Anal yzer. The outer layer is composed of alginate when the number of coatings is even; while chitosan oligosaccharide is on the surface of particles when the number of coatings is odd. Zeta Potential (mV)MobilityRelative Residual Sample Mean SEMMean SEMMean SEM Zn-0 -39.64 1.81-3.10 0.140.0393 0.0020 Zn-1 -32.95 0.94-2.57 0.070.0347 0.0024 Zn-2 -33.51 0.90-2.62 0.0070.0401 0.0021 Zn-3 -26.63 0.92-2.08 0.070.0369 0.0024 Zn-4 -34.26 1.51-2.68 0.120.0362 0.0022 Zn-5 -26.42 2.10-2.06 0.160.0384 0.0026 Zn-6 -29.84 2.11-2.33 0.160.0378 0.0023 Zn-7 -24.71 1.75-1.93 0.140.0340 0.0034 Zn-8 -8.45 3.00-0.66 0.230.0441 0.0011 Zn-9 -7.05 2.37-0.55 0.180.0462 0.0008 Zn-10 5.13 2.01-0.40 0.160.0455 0.0010 Cu-0 -34.11 1.46-2.67 0.110.0415 0.0022 Cu-1 -27.74 0.95-2.17 0.070.0408 0.0026 Cu-2 -35.45 1.16-2.77 0.090.0425 0.0020 Cu-3 -29.74 1.51-2.32 0.120.0425 0.0017 Cu-4 -32.10 1.17-2.51 0.090.0413 0.0019 Cu-5 -25.89 2.20-2.02 0.130.0406 0.0024 Cu-6 -28.39 2.20-2.22 0.110.0361 0.0036 Cu-7 -26.54 1.66-2.07 0.130.0425 0.0014 Cu-8 -36.11 0.98-2.82 0.080.0311 0.0022 Cu-9 -22.68 1.18-1.77 0.090.0410 0.0013 Cu-10 -38.69 2.69-3.02 0.210.0413 0.0019

PAGE 116

116 0 500 1000 1500 2000 2500 3000 3500 244872 Incubation Time (h)Copper Concentration (uM) Cu_0 Cu_2 Cu_6 Cu_10 Figure 4-21. Concentration of c opper ions released from chitosan -alginate PEC microparticles. Microparticles were suspended at a 5 % (w/v) density in 5% (w/v) BSA and 1% (v/v) penicillin/streptomycin supplemented RPMI 1640 cell culture medium and incubated at 37C under continuous orbital rotation. At 24, 48, and 72 hours, supernatants were collected. The removed solution was repl aced with an equal volume of fresh supplemented media and samples with micropa rticles were returned to the rotating incubator until the next time point. Analysis of the supernatant was conducted using the bicinchoninic acid reagent from a BCA kit. Concentrations of albumin-reduced copper cations were determined by comparison to a standard curve. Table 4-11. Concentration values of copper ions released from PEC chitosan-alginate microcapsules. Only values for uncoated samples are reported since no values of copper were detected for the coated samples Copper Concentration ( M) Sample Time Point (hr) Mean SEMCu-0 24 2989.35 48.76 Cu-0 48 0.00 0.000 Cu-0 72 239.53 411.7

PAGE 117

117 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Survival (%control) Ca_0 Ca_2 Ca_6 Ca_10 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Recovery (%control) Ca_0 Ca_2 Ca_6 Ca_10 ** p < 0.05 ** p < 0.01 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Survival (%control) Zn_0 Zn_2 Zn_6 Zn_10 ** p < 0.01 *** p < 0.001*** *** *** ** *** *** 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Recovery (%control) Zn_0 Zn_2 Zn_6 Zn_10 *** *** *** *** *** ****** p < 0.001 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Survival (%control) Cu_0 Cu_2 Cu_6 Cu_10 * *** *** p < 0.05 *** p < 0.001 0 50 100 150 200 250 300 0.010.1110100 Eluent of 1 mg/mL Particle Solution (%)Recovery (%control) Cu_0 Cu_2 Cu_6 Cu_10 *** ** ** p < 0.01 *** p < 0.001 Figure 4-22. Survival and recove ry percentages of human dermal fibroblasts. Cells (passage 10) were plated onto 96-well plates at a density of 1,000 cells/we ll and incubated for 24 hours in 10% FBS supplemented media. Fibroblasts were then incubated in 100 l of the different types of media. Cells incubated in 2% and 10% FBS supplemented media served as negative controls; whereas cells incubated in extracts from different concentrations of copper powder were used as positive controls (n= 6). Fibroblasts were exposed to extracts from (A, B) cal cium-, (C, D) zincand (E, F) coppercrosslinked PEC microcapsules. At 24 hour s, cells were either collected for MTT assay (A, C, E) or cultured in 10 % FBS medium for 48 hours (B, D, F). Results are reported as a percentage of the 2% FBS control. Significant differences between uncoated particles and all the coated samples are reported. EF C D AB

PAGE 118

118 Table 4-12. Fibroblast cell surv ival and recovery percentage s as a function of extract concentrations from PEC microparticles. Re sults are reported as a percentage of the 2% FBS control Concentration Survival (%c ontrol)Recovery (%control) Sample ( g/mL) Mean SEMMean SEM 10% FBS --47.58 15.7074.67 6.989 Cu-control-1 0.1 150.7 10.38136.0 15.67 Cu-control-2 10.0 68.66 9.12685.55 5.418 Cu-control-3 1000.0 0.000 0.00059.33 5.948 Ca-0 0.1 217.6 13.31108.7 10.53 Ca-2 0.1 221.8 10.32126.4 15.40 Ca-6 0.1 169.6 22.1781.82 7.960 Ca-10 0.1 249.8 48.79120.4 18.51 Ca-0 1.0 222.1 10.89108.2 9.791 Ca-2 1.0 241.2 14.2396.44 17.14 Ca-6 1.0 235.9 16.5583.83 12.21 Ca-10 1.0 225.4 28.5778.73 5.387 Ca-0 10.0 164.3 14.41141.9 13.11 Ca-2 10.0 118.3 73.0062.01 11.44 Ca-6 10.0 157.9 9.49763.29 12.23 Ca-10 10.0 181.8 8.19479.67 25.25 Ca-0 100.0 185.6 17.3689.44 13.36 Ca-2 100.0 154.2 6.99388.58 12.01 Ca-6 100.0 183.9 17.8583.28 11.66 Ca-10 100.0 172.5 12.6158.82 9.785 Ca-0 1000.0 172.5 23.58137.9 22.48 Ca-2 1000.0 222.2 22.56113.5 8.090 Ca-6 1000.0 171.7 11.7570.98 5.865 Ca-10 1000.0 161.2 20.2298.98 22.60 Zn-0 0.1 217.3 7.881108.6 11.05 Zn-2 0.1 236.5 17.09128.4 8.298 Zn-6 0.1 225.5 10.86117.4 17.15 Zn-10 0.1 218.6 6.869125.6 8.678 Zn-0 1.0 232.2 11.4198.19 5.381 Zn-2 1.0 217.9 9.468117.2 10.56 Zn-6 1.0 248.0 15.65134.4 20.90 Zn-10 1.0 213.5 11.31129.1 9.863 Zn-0 10.0 181.0 23.1768.27 13.82 Zn-2 10.0 241.6 53.3285.02 10.81 Zn-6 10.0 180.4 17.29115.7 14.63 Zn-10 10.0 153.3 17.9882.32 4.997 Zn-0 100.0 0.000 0.0000.000 0.000 Zn-2 100.0 203.0 30.58106.4 9.982 Zn-6 100.0 188.5 11.0696.93 6.714 Zn-10 100.0 167.4 11.23134.9 26.72 Zn-0 1000.0 0.000 0.000 0.000 0.000 Zn-2 1000.0 210.9 30.1399.33 8.423 Zn-6 1000.0 217.2 37.26101.8 14.43 Zn-10 1000.0 163.6 21.51102.2 17.42 Cu-0 0.1 143.4 18.94145.3 3.656

PAGE 119

119 Table 4-12. (continued) Cu-2 0.1 203.7 9.893115.5 14.59 Cu-6 0.1 246.5 36.6699.96 6.448 Cu-10 0.1 243.8 26.83109.2 9.618 Cu-0 1.0 189.8 24.66124.0 6.700 Cu-2 1.0 191.5 17.84121.1 12.35 Cu-6 1.0 219.1 11.45118.3 2.559 Cu-10 1.0 203.6 7.226132.2 13.34 Cu-0 10.0 108.9 40.2769.67 16.10 Cu-2 10.0 144.1 10.6377.70 10.09 Cu-6 10.0 194.7 23.9769.63 8.220 Cu-10 10.0 155.4 14.8692.87 3.657 Cu-0 100.0 104.7 16.8393.97 10.24 Cu-2 100.0 163.1 10.84103.6 14.22 Cu-6 100.0 179.5 9.06685.58 10.79 Cu-10 100.0 179.9 8.05999.56 15.04 Cu-0 1000.0 0.000 0.00023.36 11.62 Cu-2 1000.0 89.73 20.6171.86 6.007 Cu-6 1000.0 161.9 15.2794.25 10.26 Cu-10 1000.0 167.6 20.71108.2 19.69

PAGE 120

120 CHAPTER 5 ENCAPSULATION A ND RELEASE OF BOVINE SERUM ALBUMIN FROM POLYELECTROLYTE CHITOSAN-ALGI NATE MICROPARTICLES CROSSLINKED WITH ZINC OR COPPER IONS Introduction A vast number of proteins, in cluding monoclonal antibodies, growth factors, cytokines, soluble receptors, and hormones, have been appr oved by the FDA to treat a variety of diseases [ 31, 32, 33 ]. However, conventional oral and intravenous (IV) delivery of these drugs is usually not effective because of the inhere nt instability of many proteins [ 33, 31, 32, 34, 35 ]. Proteins have a very short in vivo half-life, are incap able of diffusing through biological membranes and are unstable in the body environment [ 33, 36, 37 ]. Although intravenous protein administration is most effective, daily injections and high pr otein concentrations are required to achieve an effective local concentration for a prolonged time [ 33, 38 ]. Frequent systemic doses increase treatment cost, patient discomfort, and side effects. To improve delivery of proteins, many contro lled-release systems composed of polymeric biomaterials have been developed [ 39, 40, 41 ]. The main goal of deve loping these systems is to control the release of drugs so th at a therapeutic level is achieve d for long periods of time. The most promising delivery approach is the en capsulation of protein within biodegradable polymeric nanoor microspheres, which facilita te drug administration through a syringe needle [ 33, 41, 43 ]. Poly(lactide) or poly(lactide-co-glicolid e)-based microspheres have been the most studied systems due to the excellent biocompatib ility and biodegradability properties of the polymers [ 42, 47 ]. However, the main drawbacks of th ese systems include: denature of some encapsulated proteins due to the manufacturing pr ocess conditions, poor encapsulation efficiency for hydrophilic drugs, and a deficient linear release of the drug.

PAGE 121

121 In this study we examined the application of chitosan-alginate microparticles in the development of a controlled-rel ease system composed of a prot ein-loaded core coated by a polyelectrolyte complex (PEC) shell. The ration ale behind this system was that the shell would provide the pharmacokinetic -release limiting factor [ 42, 48 ]; with a release kinetics governed by diffusion from the core through the degrading shell [ 48 ]. This work was focused on the encapsulation and characterization of bovine seru m albumin, as a model protein, within zinc-or copper-crosslinked alginate microcores coated by multiple PEC layers composed of chitosan oligosaccharide and alginate. Materials and Methods Preparation of FITC-Labeled BSA (F-BSA) BSA was conjugated with fluorescein isothi ocyanate (FITC) following the method from Blau et al [ 115 ]. Briefly, FITC (Sigma, St. Louis, MO) was diluted in dimethyl sulfoxide (DMSO) prior to use. BSA was dissolved at a 10 mg/ml concentration in 500 mM sodium carbonate-bicarbonate buffer (pH 9. 5) and reacted with 10-fold molar excess FITC for one hour in the dark. Dialysis was performed against protein storage buffer ( 10 mM Tris, 1 mM EDTA, 150 mM NaCl, 0.1wt% sodium azide) at pH 8.2 fo r 2 hr. Following this, the protein storage buffer was replaced by fresh buffer and dialysis was continued overnight. F-BSA obtained was stored at -20C. The molar fluorescein/pro tein (F/P) ratio of F-BSA was determined spectrofluorometrically at 280 nm and 490 nm according to the following equations: nm FITC nmA F490 @ 490 nm BSA nm nmA A P280 @ 490 280* 3 0

PAGE 122

122 where FITC@490nm = 68,000/M-cm and BSA@280nm = 43824/M-cm represent the molar extinction coefficients of FITC and BSA at 490 nm and 280 nm, respectively. Micro-Core Preparation The W/O emulsion-gelation techni que was used to synthesize the alginate microcores. A 3% (w/v) sodium-alginate solution was prepared and F-BSA was added to this aqueous solution at a 1.7% (w/w) concentration. This mixture was then emulsified in an organic phase containing cyclohexane (Aldrich, St. Louis, MO) and Plur onic L61 (kindly donated by BASF) at a 0.5% (v/v) concentration. The first emulsion was obtained by ultrasonication of the two phases at 60W for 1 min in an ice bath. A second aqueou s solution containing 0.5M of the crosslinking agent was added to the emulsion by air-spray (4 0 psi, 20ml/hr) at a 4-cm dropping distance while stirring the whole medium slowly w ith a magnetic stirrer. Particles were allowed to cure for ten minutes under continuous stirring. Then medium wa s allowed to rest for 24 hr so that particles would drop to the bottom of the container while the oil phase was left at the top. After separation of the two phases, particles were collected by filtration through a 45m mesh and washed copiously to remove the organic solvent. Particle solution was also warmed at 37C to induce solvent evaporation. After removal of the organic phase, pa rticles were stored at 4C for future coating. Coating of Alginate Particles Znand Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption technique. The coating always started with the cationic polym er and ended with the anionic polymer. Microcapsules with zero, two, six, a nd ten alternating layers were fabricated. The cationic polymer used was a high-deacetylated, low-molecular-weight chitosan oligosaccharide lactate (Mw < 5000, 90% deacetylation) from Aldric h. Alginate was the anionic polymer used. A 3 % (w/v) chitosan solution containing CaCl2 or a 0.1 % (w/v) alginate solution was added to

PAGE 123

123 crosslinked-alginate part icles dispersed in deionized water. After allowing the particles to coat for 30 minutes, they were collected by centrif ugation and washed three times with deionized water to ensure that all free polyelectrolytes were removed. Following the washes, particles were ready for the next coating or stored at 4C until use. Table 5-1 shows the different batches made; each batch was repeated three times. Characterization of F-BSA-Loaded Microcapsules To characterize the F-BSA-loaded microcapsu les, five tests were performed: surface morphology analysis, particle size distribution, coating adsorption, encapsulation efficiency, and protein release kinetics. Surface morphology analysis Surface morphology of the microcapsules was examined by light microscopy and scanning electron microscopy (SEM). Samples for scanning electron micrographs we re obtained after 0, 2, 6, and 10 coatings. Droplets of microsphere solutions were mounted on aluminum stubs, let air dried and sputter-coated with gold and pallad ium particles. The stubs were mounted in a scanning electron microscope at 10.0 kV and imaged at x2500 and x10000. Particle size distribution Size of microcapsules was analyzed by disper sing particles in deionized water at a concentration of 0.1% (w/v). Measurements were carried out by a B eckman LS13320 Particle Characterization Coulter (Beckman Instruments, Fulle rton, CA). Calculation of the particle sizes was carried out using the sta ndard modus of the LS13320 Particle Size Analyzer software (Beckman Instruments, Fullerton, CA). Percentage of particle diameters was used to describe particle size. Each sample was measured in triplicate.

PAGE 124

124 Coating adsorption analysis Adsorption of each polyelectrolyte layer ont o microcapsules was examined by measuring the particle surface charge changes. Samples were dispersed in dH2O at a concentration of 0.1% (w/v) and measurements were carried out by the Brookhaven ZetaPlus Analyzer (Brookhaven Instruments Corp., USA). The zeta potential wa s calculated from the solution conditions and the measured electrophoretic mobility. Each sample was measured in triplicate and the values reported were the mean value for the three replicate samples. Determination of protein loading efficiency The loading efficiency, or the ratio of prot ein encapsulated to the initial protein mass, was determined by decomposition of the microcores fo llowed by PEC shell destabilization. Particles were dispersed in DMEM/F12 50:50 cell culture medium (Cellgro, Herndon, VA) supplemented with 5% (v/v) goat serum albumin, GSA (Sigma, St. Louis, MO), and 1% (w/v) penicillin/streptomycin solution (Cellgro, Herndo n, VA). Multiple-layer particles, at a 2.5% (w/v) concentration, were incubated at 37C fo r 15 min. The suspension was vortexed for 2 min to disrupt particles and extract the F-BSA. Par ticle solution was incubated for another 15 min at 37C and vortexed again before collecting the aqueous solution by centrifugation at 10,000 rpm for five minutes. The remaining pellet was resuspended in the same initial volume of chelating medium and the incubation and vortex process wa s repeated. The supe rnatant collected was stored at -20C for future analysis. Analysis of the thawed supernatant from the microspheres was conducted using the enzyme-linked immunosorbent assay (ELISA) specific for bovine serum albumin (Alpha Diagnostic International Inc., San Antonio, TX). Concentrations of BSA were determined by comparison to a standard curve. All analyses were conducted in triplicate. Encapsulation efficiency, theoretical and actual protein loadings are defined as follows:

PAGE 125

125 100 loading protein l theoretica loading protein actual efficiency ion encapsulat weight e microspher ed encapsulat potein loading protein actual polymer protein total protein total loading protein l theoretica In vitro release kinetics of F-BSA microcapsules Cumulative release kinetic studies were conducte d to determine temporal release of F-BSA from the microparticles. F-BSA-loaded microspheres were suspended in 5% (v/v) GSA cell culture medium at a 5% (w/v) density. The suspension was incubated at 37C under continuous orbital rotation to ensure continuous mixing. At predetermined interval s (12, 24, and 48 hours) samples were removed from the incubator a nd centrifuged at 13,000 rpm for 5 min. The supernatant was removed, reserved in a new labeled tube, and stored at -20 C for future analysis. The removed solution was replaced with an e qual volume of fresh 5% GSA medium. The sample tubes containing microspheres were return ed to the rotating incubator at 37C until the next time point. Analysis of the thawed supernatant from the microspheres was conducted using an ELISA kit specific for BSA (Alpha Diagnostic Internationa l Inc., San Antonio, TX). Concentrations of F-BSA were determined by comparison to a standa rd curve. All analys es were conducted in triplicate. Bovine serum albumin ELISA The Bovine Albumin ELISA kit from Alpha Dia gnostic Intl. Inc. is a competitive type of assay. The test is based on the competitive bi nding of fixed concentrations of conjugated antiBSA-horseradish-peroxidase (ant i-BSA-HRP) to BSA coated on the plate and to BSA in the

PAGE 126

126 sample. Therefore, the higher the BSA concentration in the sa mples, the less amount of antiBSA-HRP bound to the plate; re sulting in a color developmen t inversely proportional to the amount of BSA present in the sample. The a ssay was performed following the kit manual. Twenty microliters of standards or samples was a dded to each well of a 96-well plate coated with bovine serum albumin. Subsequent to sample addition, 80 l of anti-BSA-HRP was added to each well and the plate was incubated at room te mperature for one hour. After washing the wells five times, 100 l of 3,3',5,5'-tetramethylbenzidine (TMB) substrate solution was added to each well and incubated in the dark at room temper ature for 15 minutes. The reaction was stopped by adding 100 l of stop solution to all wells. Optical density of each well was determined at 450 nm on a microplate reader. Data Interpretation Data were expressed as mean values standa rd error of the mean (SEM). To describe statistical differences, one-way analysis of variance (ANOVA) and Tukey-Kramer multiple comparison post test were used. Statistical significance was defined as p 0.05. Results and Discussion Bovine serum albumin was successfully conjugat ed to fluorescein isothiocyanate (FITC) with a final protein concentra tion of 10 mg/ml and a fluorescein -to-protein ratio of 8.1 5.8 (mol/mol). Although BSA was labeled with FITC to facilitate the detection of released protein by spectrofluorometry, it was discovered that coa ting the particles with chitosan oligosaccharide made them autofluorescent (Figure 5-1). The au tofluorescence of the coated particles was very high and beyond the range of typical background; therefore, analyses using the fluorescence properties of FITC-conjugated BSA could not be used. These observations are in agreement with previous studies showing the autofluorescence properties of natural forms of chitin [116, 117, 118].

PAGE 127

127 Figures 5-1, 5-2, and 5-3 illustrate how the fluorescence intensity of coated particles is dependent on the number of coatings and the cross linker used to prepare the alginate microcores. The fluorescence intensity was directly proportional to the number of coatings. In addition, it is evident that particles fluo rescence decreased when F-BSA was encapsulated within coppercrosslinked alginate microcores (Figure 5-3). The intrinsic fluorescence of chitosan seems to be affected by the environment and binding to specific molecules. The binding affinity of albumin for heavy metal ions, especially for copper ions, is known from previous chapters. When F-BSA is encapsulated within th e core, it chelates some of the coppe r ions in the algi nate gel leaving some binding sites available. As a result, th e copper-alginate gel matrix is more porous and binding of the low molecular weight chitosan to th e interior gel network is favored. Since more chitosan is bound to the interior of the core than onto the surface of F-BSA-loaded copperalginate microcores, these microparticles present lower fluorescence intensity. The particles-forming properties in the em ulsion-gelation technique were significantly altered when F-BSA was encapsulated. As can be seen in Figures 5-4A and 5-5A, zincand copper-alginate microcores appear ed spherical in shape and displa yed similar characteristics. Both systems showed a very rough surface morp hology and high particle agglomeration. In addition, it is also evident from Figures 5-4 and 5-5 that particles size decreased with encapsulation of F-BSA. These observations differ from the surface morphology and size distribution presented by blank mi crocores. Particles crosslinke d with copper presented a teardrop shape as oppose to the spherical shape of F-BSA-loaded copper-alginate particles. Besides shape, empty zincand copper-alg inate microparticles did not present any agglomeration problems at the uncoated stage. Moreover, their size dist ribution was between 5 10 and 15 20 m for zinc and copper particles, respec tively; as opposed to the 1 5 and 1 10

PAGE 128

128 m diameter range presented by F-BSA-loaded zi ncand copper-alginate microcapsules. Size distribution of protein-loaded microparticles could not be determined by the Beckman LS13320 Particle Characterization Coulter due to high pa rticle aggregation. The reported values are estimates obtained from scanning electron micrograph images. The observed differences in particle morphology and size distri bution between empty and protei n-loaded particles could be attributed to changes in the interior core networ k. Since encapsulation took place at a neutral pH and BSA (pI 4.9) was negatively charged, the pr otein could compete with the carboxylic acid sites on the alginate for the crosslinker ions. As a consequence, crosslinki ng of the alginate cores took longer, increasing the coal escence of particles and causing high aggregation. This will influence the release kinetics pursued later. Adsorption of each polyelectrolyte layer ont o microcapsules was examined by measuring the particle surface charge changes. We fi rst examined uncoated F-BSA-loaded alginate particles dispersed in dH2O. There was no significant differen ce between the zeta potentials of uncoated zincand copper-alginate particles (Table 52). It is apparent fr om Figure 5-6A that the zeta potential of F-BSA-loaded zinc-micropa rticles with the chitosan surface layer (odd numbers) is slightly less negative than micropa rticles with the alginate surface layer (even numbers). The data showed the presence of each polyelectrolyte layer with a significant difference between uncoated and chitosan-surface coated microparticles. Also, Figure 5-6B illustrates that the zeta potential of F-BSA-load ed copper-microspheres with one and three PEC layers ending in chitosan di ffered very significantly compared to the uncoated microspheres values. It is evident that microbeads retained a negative surface charge throughout the modification of the surface layer. Thickness of the polymeric films on colloidal templates is

PAGE 129

129 considered to be in the monomolecular-layer range [110]. Since the low molecular weight (< 5000Da) chitosan oligosaccharide was used as th e cationic polymer, layer thickness was very thin and the polyelectrolyte molecules failed to overcharge the microbead surface. Therefore, reversing of the microparticles surface charge did not occur. In addition, the formation of aggregates, which increased with the number of coatings, resulted in an irregular coating of the particles. The significant differences seen in F-BSA-loaded copper partic les with one and three PEC coatings could be explained by the high binding affinity of al bumin for copper ions. During encapsulation, albumin competitively binds to coppe r ions, leaving the core alginate gel with binding sites available. Therefore, more chitosan binds to the interior ge l network, lowering the negative surface charged. The encapsulation efficiency of the zinca nd copper-crosslinked alginate microparticles was determined with an ELISA kit specific for bovine serum albumin. Figure 5-7 and Table 5-3 illustrate the difference in encapsulation efficiency as a function of the type of crosslinker used. The total amount of F-BSA encapsulated in copp er-crosslinked alginate particles was 96.71% 7.56% as opposed to 52.93 22.5% for the zi nc-crosslinked alginate particles. The encapsulation of proteins within an anionic po lymer such as alginate depends vastly on the charge of the protein. Positive charged protei ns interact with alginate, forming coacervates; while there is no interaction between alginate and proteins with a negative charge, resulting in poor entrapment efficiency. The low amount of protein encapsu lated within zinc-alginate pa rticles could be attributed to the negative charge of F-BSA during encapsula tion. However, this does not hold true for copper-alginate particles because of the high binding affinity of BSA for copper. BSA competitively binds to copper ions reducing its ne t negative charge; therefore, less repulsive

PAGE 130

130 forces exist between alginate and BSA, resu lting in high encapsulation efficiency. The incorporation of proteins in to alginate microspheres can be improved by changing sample preparation conditions such as pH, salt concentra tions, protein to polymer concentration ratio, polymer molecular mass, and microsphere size. Figure 5-8 shows the release kinetic profiles of F-BSA fr om PEC chitosan-alginate microspheres. Cumulative release amounts of FBSA were determined by an ELISA kit. The ELISA assay permitted us to quantify only the released fraction of BSA specific for the recognition of the conjugated an tibody. The release kinetics of the F-BSA microspheres was contingent on the crosslinker and the number of coatings. Uncoated zinc-alginate particles showed a cumulative linear release profile (Figur e 5-8C) for the 48-hr period studied. It is apparent from Figure 5-8A that most of the F-BSA was released during the first 24 hours. Although Figure 5-8D also shows a linear release rate of th e protein from uncoated copperalginate particles; approximately 50% F-BSA was released in the first 12 hours as opposed to 25% release at 24 and 48 hours. The data shown in Table 5-4 for the pr otein release rate of uncoated zincand copper-alginate pa rticles do not differ significantly. On the contrary, F-BSA released from coated alginate particles diffe red significantly from the release rate observed for uncoa ted microspheres. When zinc-a lginate particles were coated with multiple PEC layers, there was a 2and 10-fold reduction in F-BSA release during the 12and 24-hr incubation periods, resp ectively (Figure 5-8A). The obs erved protein release reduction was not affected by further increment in the number of coatings. F-BSA release decreased by 10-fold at 12 hours when copper-alginate particles were coated with PEC layers (Figure 5-8B). No released protein was detect ed from coated particles at the 24and 48-hour time point.

PAGE 131

131 The release of proteins en capsulated within alginate networks is governed by two mechanisms: (1) diffusion of protein through the pol ymer matrix pores and (2) degradation of the polymer [84]. While diffusion of small molecules is unaffected by the alginate matrix, porosity plays an important role in the diffusion of larger proteins. Many factors va ry porosity of alginate matrices such as polymer concentration, molecu lar mass, monomer ratio, gelation rate, protein loading, protein/polymer ratio, and the presence of salts during the crosslinking reaction. To illustrate, gels made from high -L-guluronic acid alginates presented the most open pore structure and, therefore, the highe st diffusion rates for proteins [84]. Although microspheres morphological factors (such as protein diffusion and polymer erosion) affect the release kinetics; protein instab ility problems (such as aggregation, adsorption, and ionic interactions), protein si ze, pI, solubility and its distri bution within the polymer matrix are also important factors influencing the releas e trends. As mentioned earlier, the encapsulation of proteins within alginate depends vastly on the charge of the protein. Although, positive charged proteins interact with alginate form ing coacervates and resulting in high encapsulation efficiency as oppose to negative charged proteins; these coacervations hinder the protein release profile as well. The high molecular weight of F-BSA, its possible a ggregation and adsorption during the formulation and/or the release period, and the low stab ility of the uncoated zincand copper-alginate particles in the chelating medium could explain th e short release time observed. Besides the morphological factors of the alginate matrix and the physicochemical properties of the encapsulated protein, the releas e kinetics of our core-s hell systems is also affected by PEC layers surrounding the alginate co re. F-BSA-loaded microspheres with multiple PEC coatings showed an incomplete release pa ttern during the 48-hr in cubation time, which can be explained by a strong interact ion between the two polyelectroly tes and a stabili zation of the

PAGE 132

132 anionic and cationic polymer by electrolytes pres ent in the chelating medium. Future release studies with longer incubation times are needed to fully understand the re lease kinetics of our chitosan-alginate PEC micropart icles since no trend could be determined at this time. Conclusion BSA was successfully labeled with FITC to fac ilitate the detection of released protein by spectrofluorometry. However, analyses using th e fluorescence properties of F-BSA could not be performed since coating the particles with chitos an oligosaccharide made them autofluorescent. The fluorescence intensity was directly proportion al to the number of coatings and it was affected by the environment and binding to specif ic molecules. The particles-forming properties in the emulsion-gelation techniqu e were significantly altered when F-BSA was encapsulated. Although zincand copper-alginate microcores app eared spherical in shap e and displayed similar characteristics, both systems showed a ve ry rough surface morphology and high particle agglomeration. Also, particles size decr eased with encapsulation of F-BSA. The total amount of F-BSA encapsulated in copper-crosslinked alginate particles was 96.71% 7.56% as opposed to 52.93 22.5% for the zinc-crosslinked alginate particles. The incorporation of proteins in to alginate microspheres can be improved by changing sample preparation conditions such as pH, salt concentra tions, protein to polymer concentration ratio, polymer molecular mass, and microsphere size. The release kinetics of the F-BSA microspheres was contingent on the crosslinker and the number of coatings. Uncoated zinc-alginate particles showed a cumulative linear release profile for the 48-hr period studied; while F-BSA-loaded microspheres with multiple PEC coatings presented an incomplete release pattern. Future release studies with longer incu bation times are needed to fully understand the release kinetics of our chitosan-alginate PEC microparticles since no trend could be determined at this time.

PAGE 133

133 Table 5-1. The PEC chitosan-al ginate particle samples made Batch # Crosslinker Protein Encapsulated# of Coatings 1 Zn Blank0 2 Zn Blank2 3 Zn Blank6 4 Zn Blank10 5 Zn F-BSA0 6 Zn F-BSA2 7 Zn F-BSA6 8 Zn F-BSA10 9 Cu Blank0 10 Cu Blank2 11 Cu Blank6 12 Cu Blank10 13 Cu F-BSA0 14 Cu F-BSA2 15 Cu F-BSA6 16 Cu F-BSA10

PAGE 134

134 Figure 5-1. Optical micrographs of blank zinc-alginate microparticles made with the W/O emulsion gelation technique. Light and fl uorescence microscopy images of (A,B) zero-, (C,D) two-, (E,F) sixand (G ,H) ten-multilayer chitosan-alginate microparticles. Original magni fication = x100, bars denote 100 l. A B C D E F G H

PAGE 135

135 Figure 5-2. Optical micrographs of blank c opper-alginate microsphere s made with the W/O emulsion gelation technique. Light and fl uorescence microscopy images of (A,B) zero-, (C,D) two-, (E,F) sixand (G ,H) ten-multilayer chitosan-alginate microparticles. Original magni fication = x100, bars denote 100 l. A B C D E F G H

PAGE 136

136 Figure 5-3. Optical micrographs of F-BSA-loaded alginate micr oparticles coated with ten PEC layers. Light and fluorescence microscopy images of (A,B) zincand (C,D) coppercrosslinked alginate microcore. Origin al magnification = x100, bars denote 100 l. A B C D

PAGE 137

137 Figure 5-4. Scanning electron micrographs of zinc-crosslinked alginate microparticles made with the W/O emulsion gelation technique. Al ginate particles were coated with zero A) two, B) six C), or ten D) PEC multilay ers. Original magnification = x10000, bars denote 1 m. A B C D

PAGE 138

138 A B C D Figure 5-5. Scanning electron micrographs of copper-crosslinked alginate microparticles made with the W/O emulsion gelation technique. Al ginate particles were coated with zero A) two, B) six C), or ten D) PEC multilay ers. Original magnification = x10000, bars denote 1 m.

PAGE 139

139 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 012345678910 Number of CoatingsZeta Potential (mV) p < 0.05 *** p < 0.001 * *** *** * -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 012345678910 Number of CoatingsZeta Potential (mV) *** p < 0.001 *** *** *** *** *** *** *** Figure 5-6. Zeta potential values of A) zi ncand B) copper-crosslinked multiple layer PEC particles encapsulating F-BSA. Samples were dispersed in dH2O at a concentration of 0.1% (w/v) and measurements were carried out by the Brookhaven ZetaPlus Analyzer. The outer layer is composed of alginate when the number of coatings is even; while chitosan oligosaccharide is on the surface of particles when the number of coatings is odd. Significant differen ces between uncoated pa rticles and all the coated samples are reported. B A

PAGE 140

140 Table 5-2. Zeta potential va lues of PEC F-BSA-loaded chit osan-alginate microcapsules. Measurements were carried out by the Br ookhaven ZetaPlus Analyzer. The outer layer is composed of alginate when the number of coatings is even; while chitosan oligosaccharide is on the su rface of particles when the number of coatings is odd Zeta Potential (mV)MobilityRelative Residual Sample Mean SEMMean SEMMean SEM Zn-0 -34.712 0.895-2.71 0.070.0387 0.0029 Zn-1 -23.992 0.633-1.87 0.050.0334 0.0019 Zn-2 -37.241 0.791-2.91 0.060.0400 0.0022 Zn-3 -24.785 1.218-1.94 0.100.0403 0.0023 Zn-4 -35.208 1.562-2.75 0.120.0407 0.0021 Zn-5 -28.925 1.385-2.26 0.110.0337 0.0026 Zn-6 -28.739 1.610-2.25 0.130.0039 0.0028 Zn-7 -28.624 0.769-2.24 0.060.0401 0.0009 Zn-8 -33.979 1.018-2.65 0.080.0379 0.0022 Zn-9 -28.942 1.166-2.26 0.090.0358 0.0026 Zn-10 -29.840 1.402-2.33 0.110.0396 0.0026 Cu-0 -35.484 0.722-2.77 0.060.0399 0.0023 Cu-1 -3.104 1.329-0.24 0.100.0384 0.0022 Cu-2 -44.932 1.118-3.51 0.090.0380 0.0024 Cu-3 -4.613 1.404-0.36 0.110.0451 0.0012 Cu-4 -20.938 1.835-1.64 0.140.0466 0.0008 Cu-5 -20.543 0.627-1.61 0.050.0459 0.0005 Cu-6 -20.818 1.267-1.63 0.100.0462 0.0009 Cu-7 -26.188 1.421-2.05 0.110.0414 0.0020 Cu-8 -37.359 1.137-2.92 0.090.0424 0.0016 Cu-9 -30.939 1.869-2.42 0.150.0398 0.0020 Cu-10 -29.960 0.983-2.34 0.080.0406 0.0019

PAGE 141

141 0 10 20 30 40 50 60 70 80 90 100 CopperZincEncapsulation Efficiency (%) Copper Zinc Figure 5-7. Encapsulation efficiency of zincan d coppercrosslinked algi nate microparticles. The total amount of F-BSA encapsulated was determined with an ELISA specific for BSA.

PAGE 142

142 Table 5-3. Encapsulation effici ency of F-BSA loaded micropart icles prepared from zincand copper-crosslinked alginate gels Microparticle Batch Encapsulation Efficiency (%) Mean SEM Zn-alginate 52.9282 22.49 Cu-alginate 96.7085 7.555 0 10 20 30 40 50 60 70 122448 Time (hr)F-BSA Release (ug/mg (w)) Zn_0 Zn_6 Zn_10 ** p < 0.01 ** 0 10 20 30 40 50 60 70 122448 Time (hr)F-BSA Release (ug/mg (w)) Cu_0 Cu_6 Cu_10 0 20 40 60 80 100 120 122448 Time (hr)F-BSA Cumulative Release (ug/mg (w)) Zn_0 Zn_6 Zn_10 p < 0.05 * 0 20 40 60 80 100 120 122448 Time (hr)F-BSA Cumulative Release (ug/mg (w)) Cu_0 Cu_6 Cu_10 Figure 5-8. In vitro release of F-BSA from uncoated and multiple PEC coated (A,C) zincand (B,D) copper-alginate microspheres over a 48hour period. Significant differences between uncoated particles and all the coated samples are reported. A B C D

PAGE 143

143 Table 5-4. Release kinetic profile of F-BSA from uncoated and multiple PEC coated zincand copper-alginate microspheres during a 48-hour study Time F-BSA Release ( g/mg (w))F-BSA Cumulative Release ( g/mg (w)) Sample (hr) Mean SEMMean SEMZn-0 12 22.697 6.00722.697 6.007 Zn-6 12 7.261 1.5197.261 1.519 Zn-10 12 6.936 .20896.936 .2089 Zn-0 24 40.635 23.3063.331 29.31 Zn-6 24 2.397 .64159.658 2.041 Zn-10 24 0.464 .12187.400 .3020 Zn-0 48 5.459 2.19368.791 31.21 Zn-6 48 0.186 .05599.844 2.076 Zn-10 48 0.086 .02887.486 .3087 Cu-0 12 25.679 6.12825.679 6.128 Cu-6 12 1.778 .07791.778 .0779 Cu-10 12 2.887 .45532.887 .4553 Cu-0 24 11.249 1.65036.928 5.102 Cu-6 24 0.080 .03861.859 .0433 Cu-10 24 0.178 .05443.065 .5087 Cu-0 48 13.429 7.73550.357 5.729 Cu-6 48 0.010 .00181.869 .0435 Cu-10 48 0.006 .00273.071 .5103

PAGE 144

144 CHAPTER 6 CONCLUSION AND FUTURE WORK In this work, conjugations of DL-PLA and Pluronics with differe nt compositions and molecular weights were used to investigate th e systems feasibility in the development of deformable particles. Besides different mol ecular weights, Pluronics with diverse HLB ratios ranging from the most hydrophilic to the most hydrophobic copolymer were used. Preliminary results showed that only P105-PLA microspheres presented a well-defined core-shell structure with a highly porous PLA core and a P105 shell. In addition to a different surface morphology, P105-PLA particles displayed higher deformable cap ability compared to th e particles made with PLA and all the other Pluronics. During micropipette experiments, it was clearly seen that the shell was very deformable although the core was not. The differences in morphology and deformability properties are associated with the copolymer composition, indi cating that a 50 % PEO content as well as a high PPO molecular weight are ideal for the synthesis of capsular systems. Although, the P105-PLA microcapsules did not attain the desired deformab ility properties, these novel systems could be used for other applications. Fully characteri zation of these microcapsules is necessary for assessment of their drug delivery capabiliti es as well as their immunogenic effects. In addition to PLA-Pluronic conjugations, we also studied different concentrations of alginate and chitosan for the development a polye lectrolyte complex (PEC) system consisting of multiple-layer microcapsules with a dissolvable gel core. Several variables were studied to determine the set of parameters that would yield the most pr omising red blood cell analog, in terms of deformable capability. We examined the application of two frequently used techniques in the development of alginate particles: th e air-spray crosslinking and the emulsion gelation techniques. Besides discerning on the most appropriate manufact uring process, the effect of

PAGE 145

145 different crosslinkers in the de velopment of PEC microparticulates was also studied. Calcium, zinc, and copper were the divalent cations used as crosslinking agents and their effect on particles morphology, deformability, stab ility, and toxicity was analyzed. Particles made with air spray crosslinki ng technique presented a mono-dispersed size distribution; however, the smaller particle size obtained was 50 m. Although, particle size decreased with increments in th e outer lumen air pressure, monodi spersity was lost. Particles made with W/O emulsion-gelation technique using cyclohexane as the organic phase presented a narrow size distribution; diameter of uncoated particles ranged from 5-15 m. Microparticles crosslinked with copper presented a tear-drop shape as oppose to th e spherical shape of particles crosslinked with zi nc and calcium. Although a size range closer to the erythrocytes diameter wa s obtained with the emulsiongelation technique, monodispersity plays a crucial role in the drug delivery field. Therefore, modification of the applied tech niques or evaluation of other pr ocesses for the development of alginate microparticles is highly suggested. Based on the air-spray crosslinking technique, a different nozzle or microchannel de sign could be used to increase th e linear velocity and thus the productivity of a bead former. Also, a vibrational or cutting device as well as electrostatic forces could be used to decrease the size of the algi nate droplet. Regardi ng the emulsion-gelation technique, a membrane or microchannel could be us ed to control dispersity in the production of microparticles. Ca-, Zn, and Cu-alginate particles were coated applying the layer-by-layer (LbL) absorption technique. The coating always starte d with the cationic polymer and ended with the anionic polymer. Microcapsules with zero, two, six, and ten alternating layers were fabricated. Particle size decreased when increasing the number of coatings. After the tenth coating, there

PAGE 146

146 was approximately a 50% size reducti on in particles with an initial diameter ranging between 2 3 mm. Besides a decreased in size with multiple coatings, particles changed their shape from spherical to biconcave. Zinc-crosslinked particles showed a more pronounced biconcave configuration as opposed to calciuma nd copper-crosslinked particles. The new shape of the coated particles was very similar to the biconcave shape of red blood cells, which represents a great advance in the drug delivery field. The biconcave disk shape of our particles provides a greater surface area ideal for gas exchange as opposed to the traditional spherical shape of the existing drug delivery partic ulates. As opposed to the results observed in larger coated alginate partic les, coating of microspheres ca used an increase in particle aggregation, which represented a greater problem for zinc-alginate microparticles. The agglomeration of coated zinc-alginate microsphe res increased with the number of layers. Eventually, the formation of aggregates resulted in an irregular coating of the multilayer zinc particles. After removing cross-linkers of coated micr ospheres, zincand copper-alginate capsules were deformable and remained stable using the micropipette techni que under physiological pressures. Best results were obtained with Znand Cu-crosslinked alginate particles coated six times. In addition, these capsules were stable for a period of up to 10 days in sequestrantcontaining media and sustaine d shear rates of 200 sec-1. Stability of th e zincand copperalginate microparticles was improved by the nu mber of polyelectrolyte coatings. The ten multilayer copper-alginate systems formed the strongest particles due to the alginate high affinity for copper ions and the increased stability prov ided by the high number of coatings. Calciumcrosslinked particles were not stable after removal of crosslinking ions; even multiple polyelectrolyte coatings di d not improve stability.

PAGE 147

147 Even though an assessment of the particles deformable property was obtained by the micropipette technique, a quantitat ive evaluation of the particles elastic modulus could not be determined at this time. The current model used to calculate the elastic modulus of cells using the micropipette aspiration techniqu e does not apply to our systems. Further rheological studies are necessary to establish elasticity differences be tween our zincand coppe r-alginate particles. Atomic force microscopy or new designs of micr ofluid chambers simula ting capillary beds are alternative approaches to the mi cropipette technique for particles deformability assessment. Viscosity of the particles in simulating plasma medium at a 12% (w/v) density was measured using the Wells-Brookfiled Cone/Plate Digital Viscometer System with a CP-52 conical spindle. Measurements were taken at a fixed spindle speed of 100 rpm and a shear rate of 200 sec-1. Viscosities for solutions containing c opper microcapsules with two and six PEC coatings were close to blood viscosity values; while only suspensions of tw o-layer zinc particles showed viscosity values similar to bloods. As mentioned earlier, additio nal rheological studies are essential to determine how closely these par ticles simulate the behavior of erythrocytes. Viscosity studies of cell suspensions at highe r densities (up to 50% w/v) simulating blood hematocrit as well as determination of the shear rate dependence of viscoelastic properties are required for a full comparison between the flow be havior of blood and these particle-containing fluids. Adsorption of each polyelectrolyte layer ont o microcapsules was examined by measuring the particle surface charge changes. Coating microbeads with chitosan oligosaccharide did not affect the particles surface char ge as they retained a negati ve surface charge throughout the modification of the surface layer. Since the toxicity of copper ions was a concern, we decided to analyze the amount of copper ions released from microparticles in to the chelating medium. To

PAGE 148

148 determine the amount of copper ions sequester ed by albumin a modified bicinchoninic acid (BCA) protein assay was used. It was determ ined that a 5% (w/v) suspension of uncoated copper particles released a tota l copper concentration of appr oximately 3 mM in 24 hours. Copper released from the coated particles could not be determined under the conditions tested. Since the dose limit of copper in take levels in adults is 157 mol/day and the prescribed IV dose for patients with copper deficiency anemia is 47 mol/day, the low amounts of copper released from the 5% (w/v) suspension of coated particle s are still considered below toxic levels. However, understanding the ion ex change process of our system s is very important since it would allow us to determine the equilibrium and kinetics of the metal ion uptake by our PEC alginate matrices. There are many models that describe the equilibriu m and kinetics of ion exchange in alginate gels, such as the Langmui r, Freundlich, and the surface complex formation models. Although each of the models has advant ages and disadvantages, it may be possible to use one of them to determine the equilibrium of our system. The cytotoxicity of the uncoated and coat ed calcium-, zinc-, and copper-crosslinked particles was assessed using an in vitro cell proliferation assay. Human dermal fibroblast cells were exposed to extracts from the particles and the MTT assay was used to determine cell survival and recovery. Results from the assay demonstrated that multiple coatings decreased toxicity of heavy-metal crossl inked particles. The highest level of toxicity was shown by uncoated copperand zinc-crosslinked particles at 1000 g/ml. The rest of the samples seemed not to affect cells under the conditi ons studied. The recovery percen t values as well as the values obtained for cells incubated in 10% FBS medium were lower than expected. These observations are attributed to the cells reaching confluency, re sulting in inhibition of ce ll proliferation due to cell-to-cell contact. Since the MTT assay can det ect mitotic cells more accurately, the correct

PAGE 149

149 cell number of non-dividing cells cannot be determin ed through this assay. To obtain a true cell count, the MTT assay should be used along w ith a DNA quantification assay for future experiments. BSA was successfully labeled with FITC to fac ilitate the detection of released protein by spectrofluorometry. However, analyses using th e fluorescence properties of F-BSA could not be performed since coating the particles with chitos an oligosaccharide made them autofluorescent. The fluorescence intensity was directly proportion al to the number of coatings and it was affected by the environment and binding to specif ic molecules. The particles-forming properties in the emulsion-gelation techniqu e were significantly altered when F-BSA was encapsulated. Although zincand copper-alginate microcores app eared spherical in shap e and displayed similar characteristics, both systems showed a ve ry rough surface morphology and high particle agglomeration. Also, particles size decr eased with encapsulation of F-BSA. The total amount of F-BSA encapsulated in copper-crosslinked alginate particles was 96.71% 7.56% as opposed to 52.93 22.5% for the zinc-crosslinked alginate particles. The incorporation of proteins in to alginate microspheres can be improved by changing sample preparation conditions such as pH, salt concentra tions, protein to polymer concentration ratio, polymer molecular mass, and microsphere size. The release kinetics of the F-BSA microspheres was contingent on the crosslinker and the number of coatings. Uncoated zinc-alginate particles showed a cumulative linear release profile for the 48-hr period studied; while F-BSA-loaded microspheres with multiple PEC coatings presented an incomplete release pattern. Future release studies with longer incubati on times are needed to fully understand the release kinetics of our chitosan-alginate PEC mi croparticles since no trend could be determined at this time. In addition, encap sulation of various proteins differi ng in size and pI is necessary to

PAGE 150

150 determine the controlled release of diverse therapeutic agents from our systems. Besides encapsulation of different proteins, factors affec ting the porosity of alginate matrices (such as polymer concentration, molecular mass, mono mer ratio, gelation rate, protein loading, protein/polymer ratio, and the presence of salts during the crosslinking reaction) need to be studied to establish the set of parameters th at would yield the most promising drug delivery system. In conclusion, we were able to produce a deformable, biodegradable, biocompatible particle for applications in drug delivery and device testing. The nove l microparticle system developed in this work represents an advance in the drug delivery field and the beginning of a new generation of red blood cell analogs. Further rheological studies are necessary to determine how closely these particles simulate the beha vior of erythrocytes. Also, studies on the encapsulation and release of different proteins including hemoglobin along with the enzymes catalase and superoxide dismutase, are needed to establish the desired controlled release of bioactive agents for the proposed delivery system.

PAGE 151

151 LIST OF REFERENCES 1. T. M. S. Chang, Artificial cells for ce ll and organ replacements Artificial Organs 28 (2004), no. 3, 265-270. 2. D. R. Spahn, Current status of artificial oxygen carriers Advanced Drug Delivery Reviews 40 (2000), no. 3, 143-151. 3. R. R. Bartz, "Blood substitutes," vol. 2003, 2002. 4. R. M. Winslow, Blood substitutes Advanced Drug Delivery Reviews 40 (2000), no. 3, 131-142. 5. B. T. Kjellstrom, Blood substitutes: Where do we stand today? Journal of Internal Medicine 253 (2003), no. 5, 495-497. 6. K. C. Lowe and E. Ferguson, Benefit and risk perceptions in transfusion medicine: Blood and blood substitutes Journal of Internal Medicine 253 (2003), 498-507. 7. E. T. K Kobayashi, H Horinouchi, Artificial oxygen carrier its front line vol. 12, 2005. 8. R. M. Winslow, Blood substitutes 1995. 9. Y. C. Fung, Biomechanics: Mechanical pr operties of living tissues Springer, New York, 1993. 10. E. N. Marieb., Human anatomy and physiology 1998. 11. J. F. Dailey., Blood Medical Consulting Group, Arlington, 1998. 12. S. M. Slack, "Propertie s of biological fluids," Biomaterial science 1996. 13. E. W. Merrill, W. G. Margetts, G. R. Cokelet and E. R. Gilliland, The casson equation and rheology of blood near zero shear. Interscience, New York, 1965. 14. S. Chien, S. Usami, M. Taylor, J. L. Lundberg and M. I. Gregersen, Effects of hematocrit and plasma proteins of human blood rheology at low shear rates. Journal of Applied Physiology 21 (1966), 81-87. 15. Y. Isogai, S. Ikemoto, K. Kuchiba, J. Ogawa and T. Yokose, Abnormal-blood viscoelasticity in diabetic microangiopathy Clinical Hemorheology 11 (1991), no. 3-4, 175-182. 16. I. Anadere, H. Chmiel, H. Hess and G. B. Thurston, Clinical blood rheology Biorheology 16 (1979), no. 3, 171-178.

PAGE 152

152 17. H. Chmiel, I. Anadere and E. Walitza, The determination of bl ood viscoelasticity in clinical hemorheology Clinical Hemorheology 10 (1990), no. 4, 363-374. 18. T. M. S. Chang, Future generations of re d blood cell substitutes Journal of Internal Medicine 253 (2003), no. 5, 527-535. 19. L. H. Yeh and A. I. Alayash, Redox side reactions of haemoglobin and cell signalling mechanisms Journal of Internal Medicine 253 (2003), no. 5, 518-526. 20. G. B. Thurston, Viscoelasticity of human blood Biophysical Journal 12 (1972), no. 9, 1205-&. 21. G. B. Thurston, Viscoelastic properties of blood and blood analogs Advances in Hemodynamics and Hemorheology (1996), 1-30. 22. M. H. Ross, L. J. Romrell and G. I. Kaye, Histology a text and atlas Lippincott Williams and Wilkins, Baltimore, 1995. 23. I. Vilastic Scientific, "Plasma viscos ity and blood viscosity," vol. 2007, 2000, p. http://www.vilastic.com/tech10.html 24. Y. C. Fung, Theoretical considerations of elasticity of red cells and small blood vessels Federation Proceedings 25 (1966), no. 6P1, 1761-&. 25. T. Shiga, N. Maeda and K. Kon, Erythrocyte rheology Critical Reviews in Oncology/Hematology 10 (1990), no. 1, 9-48. 26. M. Hamidi and H. Tajerzadeh, Carrier erythrocytes: An overview Drug Delivery 10 (2003), no. 1, 9-20. 27. P. R. Mishra and N. K. Jain, Reverse biomembrane vesicles for effective controlled delivery of doxorubicin hci Drug Delivery 7 (2000), no. 3, 155-159. 28. B. E. Bax, M. D. Bain, P. J. Talbot E. J. Parker-Williams and R. A. Chalmers, Survival of human carrier erythrocytes in vivo Clinical Science 96 (1999), no. 2, 171-178. 29. C. G. Millan, A. Z. Castaneda, M. L. S. Marinero and J. M. Lanao, Factors associated with the performance of ca rrier erythrocytes obtained by hypotonic dialysis Blood Cells Molecules and Diseases 33 (2004), no. 2, 132-140. 30. N. Talwar and N. K. Jain, Erythrocytes as carriers of primaquine-preparation characterization and evaluation Journal of Controlled Release 20 (1992), no. 2, 133-142. 31. T. P. Richardson, M. C. Peters, A. B. Ennett and D. J. Mooney, Polymeric system for dual growth factor delivery Nat. Biotechnol. 19 (2001), no. 11, 1029-1034.

PAGE 153

153 32. S. D. Putney and P. A. Burke, Improving protein therapeutic s with sustained-release formulations Nat. Biotechnol. 16 (1998), no. 2, 153-157. 33. J. A. Schrier, B. F. Fink, J. B. Rodgers, H. C. Vasconez and P. P. DeLuca, Effect of a freeze-dried cmc/plga microsphere ma trix of rhbmp-2 on bone healing AAPS PharmsciTech 2 (2001), no. 3, 18-27. 34. K. Fu, A. M. Klibanov and R. Langer, Protein stability in cont rolled-release systems Nat. Biotechnol. 18 (2000), no. 1, 24-25. 35. M. Shameem, H. Lee and P. P. DeLuca, A short-term (accelerated release) approach to evaluate peptide release fr om plga depot formulations AAPS Pharmsci 1 (1999), no. 3, art. no.-7. 36. Y. J. Park, S. H. Nah, J. Y. Lee, J. M. Jeong, J. K. Chung, M. C. Lee, V. C. Yang and S. J. Lee, Surface-modified poly(lactide-co-gly colide) nanospheres for targeted bone imaging with enhanced labeling and delivery of radioisotope Journal of Biomedical Materials Research Part A 67A (2003), no. 3, 751-760. 37. E. R. Edelman, E. Mathiowitz, R. Langer and M. Klagsbrun, Controlled and modulated release of basic fibroblast growthfactor Biomaterials 12 (1991), no. 7, 619-626. 38. M. Arras, H. Mollnau, R. Strasser, R. We nz, W. D. Ito, J. Schaper and W. Schaper, The delivery of angiogenic factors to the heart by microsphere therapy Nat. Biotechnol. 16 (1998), no. 2, 159-162. 39. M. J. Alonso, "Controlled release of prot eins and antigens fro biodegradable polymer micro and nanospheres," vol. 2002, Dept. Ph armaceutical Technology. University of Santiago de Compostela, Spain. 40. G. Jiang, B. C. Thanoo and P. P. DeLuca, Effect of osmotic pressure in the solvent extraction phase on bsa release pr ofile from plga microspheres Pharm. Dev. Technol. 7 (2002), no. 4, 391-399. 41. G. Crotts, H. Sah and T. G. Park, Adsorption determines in-v itro protein release rate from biodegradable microspheres: Quantitati ve analysis of surface area during degradation J. Control. Release 47 (1997), no. 1, 101-111. 42. B. D. Ratner, A. S. Hoffman, F. J. Schoen and J. E. Lemons, Biomaterials science an introduction to materials in medicine vol. 1, Elsevier Academic Press, California, 2004. 43. G. Z. Zhu, S. R. Mallery and S. P. Schwendeman, Stabilization of prot eins encapsulated in injectable poly (lactide-co-glycolide) Nat. Biotechnol. 18 (2000), no. 1, 52-57. 44. http://www.drugdel.com/polymer.htm "Polymer drug delivery brief review," vol. 2002.

PAGE 154

154 45. S. Schneider, P. J. Feilen, V. Slotty, D. Ka mpfner, S. Preuss, S. Berger, J. Beyer and R. Pommersheim, Multilayer capsules: A promising microencapsulation system for transplantation of pancreatic islets Biomaterials 22 (2001), no. 14, 1961-1970. 46. J. H. Cui, J. S. Goh, S. Y. Park, P. H. Kim and B. J. Lee, Preparation and physical characterization of alginate micropa rticles using air atomization method Drug Development and Industrial Pharmacy 27 (2001), no. 4, 309-319. 47. P. Giunchedi, B. Conti, I. Genta, U. Conte and G. Puglisi, Emulsion spray-drying for the preparation of albumin -loaded plga microspheres Drug Development and Industrial Pharmacy 27 (2001), no. 7, 745-750. 48. A. A. Antipov, G. B. Sukhorukov, E. Donath and H. Mohwald, Sustained release properties of polyelectroly te multilayer capsules Journal of Physical Chemistry B 105 (2001), no. 12, 2281-2284. 49. S. E. Moya, R. Georgieva, H. Baumler, W. Richter and E. Donath, Composite lipid polyelectrolyte capsules templated on re d blood cells: Fabrication and structural characterisation Medical & Biological Engineering & Computing 41 (2003), no. 4, 504508. 50. C. Peniche, W. Arguelles-Mona l, H. Peniche and N. Acosta, Chitosan: An attractive biocompatible polymer for microencapsulation Macromolecular Bioscience 3 (2003), no. 10, 511-520. 51. D. B. Shenoy, A. A. Antipov, G. B. Sukhorukov and H. Mohwald, Layer-by-layer engineering of biocompatible, d ecomposable core-shell structures Biomacromolecules 4 (2003), no. 2, 265-272. 52. P. Couvreur, M. J. BlancoPrieto, F. Puisieux, B. Roques and E. Fattal, Multiple emulsion technology for the design of microspher es containing peptides and oligopeptides Advanced Drug Delivery Reviews 28 (1997), no. 1, 85-96. 53. P. Giunchedi, H. O. Alpar and U. Conte, Pdlla microspheres cont aining steroids: Spraydrying, o/w and w/o/w emulsifica tions as preparation methods Journal of Microencapsulation 15 (1998), no. 2, 185-195. 54. P. Giunchedi and U. Conte, Spray-drying as a preparati on method of microparticulate drug-delivery systems an overview Stp Pharma Sciences 5 (1995), no. 4, 276-290. 55. F. Caruso, C. Schuler and D. G. Kurth, Core-shell particles and hollow shells containing metallo-supramolecular components Chemistry of Materials 11 (1999), no. 11, 33943399.

PAGE 155

155 56. I. L. Radtchenko, G. B. Sukhorukov, S. Le poratti, G. B. Khomutov, E. Donath and H. Mohwald, Assembly of alternated multivalent ion/polyelectrolyte layers on colloidal particles. Stability of the multilayers and encapsulation of macromolecules into polyelectrolyte capsules Journal of Colloid and Interface Science 230 (2000), no. 2, 272280. 57. G. Decher, Fuzzy nanoassemblies: Toward layered polymeric multicomposites Science 277 (1997), no. 5330, 1232-1237. 58. R. Georgieva, S. Moya, E. Donath and H. Baumler, Permeability and co nductivity of red blood cell templated polyelectr olyte capsules coated w ith supplementary layers Langmuir 20 (2004), no. 5, 1895-1900. 59. A. Elaissari, Colloidal polymers surfactant: Synthesis and characterization Science Series 115 (2003). 60. J. Dusseault, F. A. Leblond, R. Robitaille, G. Jourdan, J. Tessier, M. Menard, N. Henley and J. P. Halle, Microencapsulation of living cells in semi-permeable membranes with covalently cross-linked layers Biomaterials 26 (2005), no. 13, 1515-1522. 61. O. Gaserod, A. Sannes and G. Skjak-Braek, Microcapsules of alginate-chitosan. Ii. A study of capsule stab ility and permeability Biomaterials 20 (1999), no. 8, 773-783. 62. R. A. Goldsby, T. J. Kindt and B. A. Osborne, Kuby immunology W. H. Freeman and Company, New York, 2000. 63. S. M. Moghimi, A. C. Hunter and J. C. Murray, Long-circulating and target-specific nanoparticles: Theory to practice Pharmacological Reviews 53 (2001), no. 2, 283-318. 64. S. Stolnik, L. Illum and S. S. Davis, Long circulating micro particulate drug carriers Advanced Drug Delivery Reviews 16 (1995), no. 2-3, 195-214. 65. E. Chambers and S. Mitragotri, Prolonged circulation of large polymeric nanoparticles by non-covalent adsorption on erythrocytes Journal of Controlled Release 100 (2004), no. 1, 111-119. 66. S. M. Moghimi and H. M. Patel, Serum-mediated recognition of liposomes by phagocytic cells of the reticuloendothelial system the concept of tissue specificity Advanced Drug Delivery Reviews 32 (1998), no. 1-2, 45-60. 67. G. Storm, S. O. Belliot, T. Daemen and D. D. Lasic, Surface modification of nanoparticles to oppose uptake by the mononuclear phagocyte system Advanced Drug Delivery Reviews 17 (1995), no. 1, 31-48.

PAGE 156

156 68. D. J. Mooney, R. S. Langer and J. P. V acanti, "Localized deliver y of factors enhancing survival of transplanted cells," Pharmaceutical Patents Children's Medical Center Corp. (Boston, MA);Massachusetts Institute of Technology (Cambr idge, MA), USA, 1994, p. 14. 69. J. L. Cleland, E. Duenas, A. Daugherty, M. Marian, J. Yang, M. Wilson, A. C. Celniker, A. Shahzamani, V. Quarmby, H. Chu, V. Mukku, A. Mac, M. Roussakis, N. Gillette, B. Boyd, D. Yeung, D. Brooks, Y. F. Maa, C. Hsu and A. J. S. Jones, Recombinant human growth hormone poly(lactic-co-glycolic acid ) (plga) microspheres provide a long lasting effect J. Control. Release 49 (1997), no. 2-3, 193-205. 70. F. W. Cordewener, L. C. Dijkgraaf, J. L. O ng, C. M. Agrawal, G. Zardeneta, S. B. Milam and J. P. Schmitz, Particulate retrieval of hydroly tically degraded poly(lactidecoglycolide) polymers J. Biomed. Mater. Res. 50 (2000), no. 1, 59-66. 71. K. F. Pistel, A. Breitenbach R. Zange-Volland and T. Kissel, Brush-like branched biodegradable polyesters, part iii protein release from microspheres of poly(vinyl alcohol)-graftpoly(d,l-lactic-co-glycolic acid) J. Control. Release 73 (2001), no. 1, 720. 72. H. Kranz, N. Ubrich, P. Maincent and R. Bodmeier, Physicomechanical properties of biodegradable poly(d,l-lactide) and poly(d,l-la ctide-co-glycolide) films in the dry and wet states J. Pharm. Sci. 89 (2000), no. 12, 1558-1566. 73. J. E. Matthew, Y. L. Nazario, S. C. Roberts and S. R. Bhatia, Effect of mammalian cell culture medium on the gelation pr operties of pluronic((r)) f127 Biomaterials 23 (2002), no. 23, 4615-4619. 74. A. J. Domb, J. Kost and D. M. Wiseman, Handbook of biodegradable polymers Harwood Academic Publishers, Singapore, 1997. 75. K. T. Oh, T. K. Bronich and A. V. Kabanov, Micellar formulations for drug delivery based on mixtures of hydrophobic and hy drophilic pluronic((r)) block copolymers Journal of Controlled Release 94 (2004), no. 2-3, 411-422. 76. A. V. Kabanov and V. Y. Alakhov, Pluronic (r) block copolymers in drug delivery: From micellar nanocontainers to biol ogical response modifiers Critical Reviews in Therapeutic Drug Carrier Systems 19 (2002), no. 1, 1-72. 77. C. Allen, D. Maysinger and A. Eisenberg, Nano-engineering block copolymer aggregates for drug delivery Colloids and Surfaces B-Biointerfaces 16 (1999), no. 1-4, 3-27. 78. T. G. Park, S. Cohen and R. Langer, Poly(l-lactic acid) pluronic blends characterization of phase-separation behav ior, degradation, and morphology and use as protein-releasing matrices Macromolecules 25 (1992), no. 1, 116-122.

PAGE 157

157 79. P. D. Vos, B. D. Hann, G. Wolters J. Strubbe and R. V. Schilfgaarde, Improved biocompatibility but limited graft survival after pur ification of alginate for microencapsulation of pancreatic islets Diabetologia 40 (1997), 262-270. 80. S. J. De and D. Robinson, Polymer relationships during pr eparation of chitosan-alginate and poly-l-lysine-alginate nanospheres Journal of Controlled Release 89 (2003), no. 1, 101-112. 81. S. B. Zhou, X. M. Deng and X. H. Li, Investigation on a novel core-coated microspheres proteins delivery system Journal of Controlled Release 75 (2001), no. 1-2, 27-36. 82. C. K. Kuo and P. X. Ma, Ionically crosslinked alginate hy drogels as scaffolds for tissue engineering: Part 1. Structure, gelation rate and mechanical properties Biomaterials 22 (2001), no. 6, 511-521. 83. D. Quong and R. J. Neufeld, DNA protection from extracapsular nucleases, within chitosanor poly-l-lysi ne-coated alginate beads Biotechnology and Bioengineering 60 (1998), no. 1, 124-134. 84. W. R. Gombotz and S. F. Wee, Protein release from alginate matrices Advanced Drug Delivery Reviews 31 (1998), no. 3, 267-285. 85. R. H. Walter, Polysaccharide association structures in food 1998. 86. R. v. Schilfgaarde and P. d. Vos, Factors influencing the prop erties and performance of microcapsules for immunoprot ection of pancreatic islets Journal of Molecular Medicine 77 (1999), 199-205. 87. M. Prabaharan and J. F. Mano, Chitosan-based particles as controlled drug delivery systems Drug Delivery 12 (2005), no. 1, 41-57. 88. T. Chandy, G. H. R. Rao, R. F. Wilson and G. S. Das, Delivery of lmw heparin via surface coated chitosan/peg-alginate microspheres prevents thrombosis Drug Delivery 9 (2002), no. 2, 87-96. 89. S. Aldrich, "Chitosan a technologically important bi omaterial," vol. 2005, p. Sigma product information sheet. 90. R. Auzely and M. Rinaudo, Controlled chemical modifications of chitosan. Characterization and investigati on of original properties Macromolecular Bioscience 3 (2003), no. 10, 562-565. 91. O. Gaserod, I. G. Jolliffe, F. C. Ha mpson, P. W. Dettmar and G. Skjak-Braek, The enhancement of the bioadhesive pr operties of calcium alginate gel beads by coating with chitosan International Journal of Pharmaceutics 175 (1998), no. 2, 237-246.

PAGE 158

158 92. N. Acosta, I. Aranaz, C. Peniche and A. Heras, Tramadol release from a delivery system based on alginate-chitosan microcapsules Macromolecular Bioscience 3 (2003), no. 10, 546-551. 93. C. Batich and f. Vaghefi, "Process for microe ncapsulating cells," University of Florida, USA, 1994, p. 18. 94. D. P. Theret, M. J. Levesque, M. Sato, R. M. Nerem and L. T. Wheeler, The application of a homogeneous half-space model in the analysis of endotheli al-cell micropipette measurements Journal of Biomechanical Engine ering-Transactions of the Asme 110 (1988), no. 3, 190-199. 95. S. Thomas, W. Bolch, K. J. Kao, F. Bova and R. Tran-Son-Tay, Effects of x-ray radiation on the rheologic properties of platelets and lymphocytes Transfusion 43 (2003), no. 4, 502-508. 96. J. F. Hoffman and S. Inoue, Directly observed reversibl e shape changes and hemoglobin stratification during centrifugation of human and amphiuma red blood cells Proceedings of the National Academy of Sciences of the United States of America 103 (2006), no. 8, 2971-2976. 97. J. P. Chen, F. Tendeyong and S. Yiacoumi, Equilibrium and kinetic studies of copper ion uptake by calcium alginate Environmental Science & Technology 31 (1997), no. 5, 1433-1439. 98. D. Cozzi, P. G. Desideri and L. Lepri, Mechanism of ion exchange with alginic acid Journal of Chromatography 40 (1969), no. 1, 130-&. 99. L. K. Jang, D. Nguyen and G. G. Geesey, Selectivity of alginate gel for cu over zn when acidic conditions prevail Water Research 33 (1999), no. 12, 2817-2825. 100. Y. Jodra and F. Mijangos, Ion exchange selectivities of ca lcium alginate gels for heavy metals Water Science and Technology 43 (2001), no. 2, 237-244. 101. P. Aslani and R. A. Kennedy, Studies on diffusion in alginate gels .1. Effect of crosslinking with calcium or zinc ions on diffusion of acetaminophen Journal of Controlled Release 42 (1996), no. 1, 75-82. 102. C. Ouwerx, N. Velings, M. M.M. and M. A. V. Axelos, Physico-chemical properties and rheology of alginate beads formed with various divalent cations Polymer Gels and Networks 6 (1998), 393-408. 103. Brookfield, "The wells-brookfie ld cone/plate digital viscomet er operating in structions," Brookfield Engineering Laborator ies, Inc., Stoughton, p. 12.

PAGE 159

159 104. M. F. A. Goosen, G. M. Oshea, H. M. Gharapetian, S. Chou and A. M. Sun, Optimization of microencapsulation parameters semiperm eable microcapsules as a bioartificial pancreas Biotechnology and Bioengineering 27 (1985), no. 2, 146-150. 105. T. Rehg, C. Dorger and P. C. Chau, Application of an atomiz er in producing small alginate gel beads for cell immobilization Biotechnology Letters 8 (1986), no. 2, 111114. 106. J. Klein, J. Stock and K. D. Vorlop, Pore-size and properties of spherical ca-alginate biocatalysts European Journal of Applie d Microbiology and Biotechnology 18 (1983), no. 2, 86-91. 107. P. A. Haas, Formation of liquid drops with uniform and controlled diameters at rates of 103 to 105 drops per minute Aiche Journal 21 (1975), no. 2, 383-385. 108. L. W. Chan, Y. Jin and P. W. S. Heng, Cross-linking mechanisms of calcium and zinc in production of alginate microspheres International Jour nal of Pharmaceutics 242 (2002), no. 1-2, 255-258. 109. P. Aslani and R. A. Kennedy, Effect of gelation conditions and dissolution media on the release of paracetamol from alginate gel beads Journal of Microencapsulation 13 (1996), no. 5, 601-614. 110. V. Dudnik, G. B. Sukhorukov, I. L. Radtchenko and H. Mohwald, Coating of colloidal particles by controlled precipitation of polymers Macromolecules 34 (2001), no. 7, 2329-2334. 111. http://cpip.gsm.com.lp.hscl.ufl.edu/ "Clinical pharmacol ogy world-class drug information," vol. 2007, 2003. 112. G. F. Hu, Copper stimulates proliferation of human endothelial cells under culture Journal of Cellular Biochemistry 69 (1998), no. 3, 326-335. 113. J. P. Rodriguez, S. Rios and M. Gonzalez, Modulation of the proliferation and differentiation of human mesenchymal stem cells by copper Journal of Cellular Biochemistry 85 (2002), no. 1, 92-100. 114. I. Tenaud, S. Leroy, N. Chebassier and B. Dreno, Zinc, copper and manganese enhanced keratinocyte migration through a functiona l modulation of keratinocyte integrins Experimental Dermatology 9 (2000), no. 6, 407-416. 115. S. Blau, N. Levin, B. Schwartz and A. Rubinstein, Adsorption of cationized bovine epithelial crypt fractions of the serum albumin onto rat colon Journal of Pharmaceutical Sciences 90 (2001), no. 10, 1516-1522.

PAGE 160

160 116. A. M. Baty, C. C. Eastburn, Z. Diwu, S. Techkarnjanaruk, A. E. Goodman and G. G. Geesey, Differentiation of chitinase-active and n on-chitinase-active subpopulations of a marine bacterium during chitin degradation Applied and Environmental Microbiology 66 (2000), no. 8, 3566-3573. 117. L. Houhamdi, P. E. Fournier, R. Fang, H. Lepidi and D. Raoult, An experimental model of human body louse infection with rickettsia prowazekii Journal of Infectious Diseases 186 (2002), no. 11, 1639-1646. 118. P. E. Fournier, M. F. Minnick, H. Lepidi, E. Salvo and D. Raoult, Experimental model of human body louse infection using green fl uorescent protein-ex pressing bartonella quintana Infection and Immunity 69 (2001), no. 3, 1876-1879.

PAGE 161

161 BIOGRAPHICAL SKETCH The author was born in Valencia, Venezuela to a general surgeon and an anesthesiologist. After the wonderful experience of living in a fo reign country for eleven years, the author was introduced to many of the things that have shap ed her life. She studied English at Northern Virginia Community College for one year. Afte r finishing her English studies, she moved to Palm Beach, Florida, where she started her college education at Palm Beach Community College. In spring 1998, the author attended the University of Florida where she obtained her bachelors degree in Engineering Science with a minor in Biomechanics in May 2000. The author started her graduate stud ies at the Biomedical Engineeri ng Department of the University of Florida in fall 2000. She obtai ned her masters degree in biom edical engineering in May 2003 and continued with her Ph.D. studies in the fall of 2003.