<%BANNER%>

Superparamagnetic Folate-Immobilized Dye Labeled Microspheres for Oral Cancer Screening


PAGE 1

SUPERPARAMAGNETIC FOLATE -IMMOBILIZED DYE LABELED MICROSPHERES FOR ORAL CANCER SCREENING By BERND LIESENFELD 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 2004

PAGE 2

Copyright 2004 by Bernd Liesenfeld

PAGE 3

ACKNOWLEDGMENTS I would like to thank all the people that contributed to this work, either directly by helping to do the work or indirectly by supporting me. Many thanks go to the members of my committee: my advisor Dr Chris Batich, my external member Dr Ken Wagener, and my departmental members Dr Paul Holloway, Dr Abbas Zaman and Dr Ken Anusavice who have contributed wonderful academic examples as much as splendid technical help, and helped me find my own direction. Additional thanks goes to Dr Karl Soderholm, who agreed to be a substitute member for my defense at a late stage, and Dr Chiayi Shen, who substituted for Dr Soderholm with almost no notice. Dr Shen, like each of my members, made an impressively careful, thoughtful and insightful reading of my dissertation with minimal notice. Perhaps more than that of anyone the always willing and generous help, of Swadeshmukul Santra has enabled many research successes. Also deeply helpful was Jon Dobson for providing expert advice and background information on magnetics. Patrick Leamy was extremely helpful as a labmate, and the quality of the research he left behind helped to enable many subsequent research successes. Special thanks go to the graduate and undergraduate students who participated on my projects in some very meaningful ways. Special thanks among these go to Cindy Rau-Zink for coaxing any number of cell lines through the experiments needed. The twinsJompo Moloye and Taili Thulahelped prevent the cell lab from imploding under its own weight. Bradley Willenberg and Mike Tollon were responsible for much iii

PAGE 4

help on MAIC machinery and the machinery in our own labs. David Chatel was one of a line of French summer students who have all contributed beyond their years of education. Other primary participants include Rekha Nair who sacrificed many hours counting cells and microspheres. JP Bullivant graciously trained any number of students in the magnetite production process that we helped Pat Leamy to develop. Other participants on the project that I would like to thank include Leland Black, Vasana Maneeratana, and members of the Batich research group that have provided excellent technical help: Albina Mikhailova, Nakato Kibuyaga and the balance of the group. Many other people have helped to provide support in ways that deserve recognition. Gill Brubaker has provided tireless training and support for major instrumentation at the Particle Science Engineering Research Center. Kevin Powers and Gary Scheiffele at the same center also provided important assistance. Eric, Wayne, and Erik at MAIC have each provided valuable instrumental assistance and guidance. Jennifer Wrighton has provided critical administrative support that enables the functioning of all the polymer research groups. I would also like to thank my personal supporters, especially Kelly Rooney and my parents who have been very patiently supportive and positive. Additional thanks go to my triathlon and cycling teamsthe Tri-Gators and Team Floridawhich provided sporting mental regeneration. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES...............................................................................................................x LIST OF FIGURES...........................................................................................................xi ABSTRACT.......................................................................................................................xv CHAPTER 1 INTRODUCTION........................................................................................................1 2 BACKGROUND..........................................................................................................3 2.1 Magnetism..........................................................................................................3 2.1.1 Types of Magnetism...............................................................................4 2.1.1.1 Diamagnetism....................................................................................5 2.1.1.2 Paramagnetism..................................................................................6 2.1.1.3 Ferromagnetism.................................................................................6 2.1.1.4 Ferrimagnetism..................................................................................8 2.1.1.5 Antiferromagnetism..........................................................................9 2.1.2 Domain Size Effects.............................................................................10 2.1.2.1 Single domains................................................................................11 2.1.2.2 Superparamagnetism.......................................................................11 2.1.3 Some Applications of Magnetic Particles............................................12 2.1.3.1 Hyperthermic treatments.................................................................13 2.1.3.2 Contrast media.................................................................................14 2.1.3.3 Particle guidance by magnetic forces..............................................14 2.2 Cancer..............................................................................................................17 2.2.1 General Cancer Background................................................................17 2.2.2 Oral Cancer...........................................................................................18 2.3 Folic Acid and Receptor Targeting..................................................................21 2.3.1 Folic Acid.............................................................................................21 2.3.2 Folate Receptors...................................................................................22 2.4 Fluorescence....................................................................................................23 2.4.1 The Fluorescence Process....................................................................24 2.4.2 Fluorescence Techniques.....................................................................25 2.5 Polymerization Methods for Producing Microspheres....................................27 v

PAGE 6

2.5.1 Emulsion Polymerization.....................................................................29 2.5.2 Soapless Emulsion Polymerization......................................................30 2.5.3 Dispersion Polymerization...................................................................30 2.5.4 Precipitation Polymerization................................................................31 2.5.5 Suspension Polymerization..................................................................31 3 PROPOSED STRATEGY AND DESIGN REQUIREMENTS.................................33 3.1 Appropriate Uses of Screening Tests...............................................................33 3.2 Proposed Testing Procedure............................................................................35 3.2.1 Components..........................................................................................35 3.2.2 Procedure..............................................................................................35 3.3 Design Parameters for Microspheres...............................................................37 3.3.1 Microsphere Size Considerations.........................................................37 3.3.2 Ligand Immobilization.........................................................................39 3.3.3 Magnetic Guidance...............................................................................39 3.3.4 Dye labeling.........................................................................................39 4 MATERIALS AND METHODS ...............................................................................41 4.1 Magnetic Material for Microspheres...............................................................41 4.1.1 Materials used in Iron Oxide Preparation.............................................41 4.1.2 Magnetite Production and Treatment...................................................41 4.1.2.1 Method of iron oxide precipitation..................................................42 4.1.2.2 Method of coating iron oxide..........................................................43 4.1.3 Characterization of Iron Oxide.............................................................44 4.2 Microsphere Polymerization and Characterization..........................................44 4.2.1 Incorporation of Fluorescent Dye Into Functional Monomer..............44 4.2.1.1 Materials for fluorescent dye incorporation into functional monomer...............................................................................................45 4.2.1.2 Method of conjugating fluorescent dye to functional monomer.....45 4.2.2 Suspension Polymerization Procedure.................................................48 4.2.2.1 Materials for suspension polymerization........................................48 4.2.2.2 Method of suspension polymerization............................................48 4.2.3 Microsphere Post-Polymerization Processing......................................50 4.2.4 Microsphere Dye Loading by Swelling and Solvent Evaporation.......51 4.2.5 Microsphere Characterization..............................................................52 4.2.5.1 Coulter sizing..................................................................................53 4.2.5.2 Light microscopy.............................................................................53 4.2.5.3 Zeta potential analysis.....................................................................54 4.2.5.4 Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS)..............................................................................54 4.2.5.5 Inductively coupled plasma spectroscopy (ICP).............................54 4.2.5.6 X-Ray powder diffraction (XRD)...................................................55 4.2.6 Microsphere Fluorescence Properties..................................................55 4.2.6.1 Sample preparation..........................................................................56 4.2.6.2 UV-Visible absorbance spectroscopy.............................................57 vi

PAGE 7

4.2.6.3 Fluorescence spectrometry..............................................................57 4.2.6.4 Confocal microscopy.......................................................................57 4.2.6.5 Fluorescence microscopy................................................................58 4.2.7 Preparation of Microspheres for Cell Work.........................................59 4.2.8 Preparation of Microspheres by Dispersion Polymerization................59 4.2.8.1 Method of dispersion polymerization.............................................59 4.2.8.2 Incorporation of magnetic species...................................................60 4.2.9 Preparation of Microspheres by Activated Swelling............................61 4.3 Immobilization of Folic Acid onto Microspheres............................................63 4.3.1 Folic Acid Immobilization Procedure..................................................63 4.3.1.1 Materials for folic acid immobilization...........................................63 4.3.1.2 Method of folic acid immobilization onto microspheres................64 4.3.2 Characterization of Folic Acid Immobilized Microspheres.................65 4.3.2.1 UV-Visible spectroscopy................................................................66 4.3.2.2 Fluorospectrometry.........................................................................66 4.3.2.3 Brookhaven zeta plus......................................................................66 4-4 Cell Testing..........................................................................................................66 4.4.1 Cell Lines.............................................................................................67 4.4.1.1 Malignant cell line CRL-5800 / NCI-H23 human epithelial lung adenocarcinoma....................................................................................67 4.4.1.2 Secondary testing cell line CCL-163 / BALB/3T3 clone A31 mouse fibroblasts.............................................................................................67 4.4.1.3 Control cell line NHDF: normal human adult fibroblasts...............67 4.4.1.4 Oral squamous cell carcinoma cell line SCC-9...............................68 4.4.2 Cell Culture Procedures........................................................................68 4.4.3 Cell Culture Preparation and Testing Procedure..................................68 4.4.3.1 Cells seeded onto multi-well plates.................................................68 4.4.3.2 Cells seeded onto coverslips...........................................................69 4.4.3.3 Cell experiments for microsphere specific binding.........................69 4.4.3.4 Fluorescently labeled microspheres................................................70 4.4.3.5 Image treatment and analysis..........................................................70 4.4.3.6 Microsphere recovery experiments.................................................71 4.5 Tissue Testing..................................................................................................71 4.5.1 Institutional Review Board (IRB) Approval........................................71 4.5.2 Tissue Preparation................................................................................72 4.5.2.1 Fresh tissue samples........................................................................72 4.5.2.2 Snap-frozen tissue samples.............................................................72 4.5.3 Sample Treatment for Testing with Microspheres...............................73 4.5.3.1 Unmounted tissue testing procedure...............................................73 4.5.3.2 Slide mounted tissue testing procedure...........................................73 4.5.4 Microscopy of Prepared Tissue Samples.............................................74 4.5.5 Magnetic Recovery of Microspheres From Tissue Samples................75 5 RESULTS AND DISCUSSION.................................................................................76 5.1 Microsphere Synthesis..........................................................................................76 5.1.1 Dispersion Polymerized Samples...............................................................77 vii

PAGE 8

5.1.1.1 D013 dispersion polymerization with ferrofluid.............................77 5.1.1.2 D030 dispersion polymerization with iron oxide precipitated in situ........................................................................................................78 5.1.2 Activated Swelling...............................................................................84 5.1.3 Suspension Polymerization........................................................................88 5.1.3.1 Magnetic dopant characterization...................................................88 5.1.3.2 Suspension polymerization methods and incorporation of iron oxide.....................................................................................................89 5.1.3.3 Microsphere size control.................................................................91 5.1.3.4 Particle morphology........................................................................94 5.2 Folic Acid Immobilization...............................................................................95 5.2.1 UV-Visible Spectrophotometry............................................................95 5.2.2 Fluorescence Spectrometry..................................................................96 5.2.3 Zeta Potential Measurement.................................................................96 5.3 Microsphere Labeling......................................................................................98 5.3.1 Dye Loading Vs. Covalent Coupling of Dye.......................................98 5.3.2 Dye Loading by Swelling.....................................................................99 5.3.3 Covalent Coupling of Dye to Microspheres.......................................103 5.3.3.1 Microsphere fluorescence...........................................................104 5.3.3.2 Confocal microscopy.....................................................................105 5.3.3.3 Dye content optimization..............................................................105 5.4 Cell Line Testing............................................................................................109 5.4.1 Initial Testing with Cell Line NCI-H23.............................................110 5.4.1.1 Initial testing results......................................................................110 5.4.1.2 Determination of desirable microsphere size from testing............111 5.4.2 Non-labeled Microspheres.................................................................112 5.4.3 Labeled Microspheres........................................................................113 5.4.4 Malignant Cell Lines Testing.............................................................113 5.4.4.1 NCI-H23 human lung adenocarcinoma cell line...........................114 5.4.4.2 BALB/3T3 mouse fibroblast cell line...........................................115 5.4.4.3 SCC-9 oral squamous cell carcinoma cell line..............................115 5.4.5 Control Cell Line Testing...................................................................116 5.5 Tissue Testing................................................................................................117 5.5.1 Tissue Sample 1..................................................................................117 5.5.1.1 Fresh tissue testing with sample 1.................................................117 5.5.1.2 Mounted tissue testing with sample 1...........................................117 5.5.2 Tissue Sample 2..................................................................................123 5.5.3 Results with Tissue Samples..............................................................123 5.6 Microsphere Recovery Experiments..............................................................124 6 DYE DOPED SILICA PARTICLES........................................................................126 6.1 Introduction....................................................................................................126 6.2 Background....................................................................................................126 6.2.1 Stber Process for Producing Nanoparticles......................................126 6.2.2 Nanoparticles applications.................................................................127 6.3 Materials and Methods...................................................................................128 viii

PAGE 9

6.3.1 Materials.............................................................................................128 6.3.2 Methods..............................................................................................128 6.3.2.1 Method of conjugating FITC fluorophores to APTS..................128 6.3.2.2 Method of synthesizing fluorescent silica nanoparticles.............129 6.3.2.3 Method of immobilizing folic acid onto the DDS nanoparticles.......................................................................................130 6.3.2.4 Transmission electron microscopy..............................................131 6.3.2.5 Scanning electron microscopy....................................................131 6.3.2.6 Zeta potential measurement........................................................131 6.3.2.7 Light scattering particle size measurement.................................131 6.3.2.8 UV-visible absorption spectroscopy...........................................132 6.3.2.9 Fluorescence spectrometry..........................................................132 6.3.2.10 Confocal microscopy.................................................................132 6.3.2.11 Cell experiments........................................................................132 6.4 Results and Discussion..................................................................................133 6.4.1 Size of FSNPs.....................................................................................133 6.4.2 Fluorescence of the FSNPs.................................................................135 6.4.3 Folic Acid Immobilized FSNPs.........................................................135 6.4.4 Cell Experiments................................................................................138 7 CONCLUSIONS AND FUTURE WORK...............................................................141 7.1 Magnetic Microsphere Preparation and Characterization..............................141 7.2 Ligand Immobilization...................................................................................142 7.3 Microsphere Labeling....................................................................................142 7.4 Cell Line Testing............................................................................................143 7.5 Tissue Testing................................................................................................145 7.6 Microsphere Retrieval....................................................................................145 APPENDIX A POLYMERIZATION RECORDS FOR SELECTED SAMPLES THAT APPEAR IN THE MANUSCRIPT...........................................................................................147 B CELL LINE DATA ..................................................................................................168 C MAGNETIC SEPARATION DEMONSTRATION................................................178 LIST OF REFERENCES.................................................................................................179 BIOGRAPHICAL SKETCH...........................................................................................185 ix

PAGE 10

LIST OF TABLES Table page 2-1 Units used for magnetic quantities.............................................................................5 5-1 Suspension Polymerization Series Results...............................................................92 5-2 Zeta potential measurement on folate-immobilized and control s19 microspheres. Averages presented are for the three runs depicted, each consisting of 10 cycles...98 5-3 Polymer solubility parameter values for selected solvents, presented in common form of (cal/cm3)1/2, not in SI units........................................................................100 5-4 Group Molar Contribution calculation of Hildebrand polymer solubility parameter for AEMH monomer unit.......................................................................................102 5-5 Swelling of microspheres by selected solvents and solvent-dye compatibility.....102 5-6 FITC concentrations in sample batches prepared to optimize dye content and fluorescence yield...................................................................................................107 5-7 Initial cell experiments statistical evaluation.........................................................111 5-8 Results showing normalized counts of microspheres per unit area for control and immobilized S18 microspheres on NCI-H23 cells.................................................114 6-1 UV-absorption spectroscopy instrument output of absorption maxima for folic acid assay on FITC5 FSNPs whose spectral curves are shown in figure 6-3................138 x

PAGE 11

LIST OF FIGURES Figure page 2-1 A lodestone with nails and magnetite fragments attached.......................................4 2-2 Diamagnetic response..............................................................................................5 2-3 Paramagnetic response curve...................................................................................6 2-4 Hysteresis loop.........................................................................................................8 2-5 Inverse spinel structure of magnetite showing A and B lattice positions................9 2-6 Energy minimization by domain walls..................................................................10 2-7 Domain wall transitions.........................................................................................11 2-8 Schematic for drug delivery system.......................................................................16 2-9 Squamous cell carcinoma of the tongue................................................................19 2-10 Carcinoma of tongue..............................................................................................20 2-11 Mechanism of receptor mediated endocytosis used to target anti-cancer drugs to tumourous cells......................................................................................................22 2-12 Structure of Folic Acid...........................................................................................22 2-13 Jablonski diagram showing energy states for a fluorescence process...................24 2-14 Polyatomic molecule spectra showing excitation and emission intensity equivalence............................................................................................................26 2-15 Normalized fluorescence emission spectra of fluorescein (FL), tetramethylrhodamine (TMR) and Texas Red (TR) dyes......................................28 2-16 fluoro probes hybridized to human metaphase chromosomes...............................29 3-1 Results matrix for disease state vs. test result........................................................34 3-2 Steps of the proposed testing strategy....................................................................36 xi

PAGE 12

3-3 Size considerations for microspheres.....................................................................38 4-1 Methyl methacrylate (MMA) structure..................................................................43 4-2 Fluorescein Isothiocyanate (FITC) structure.........................................................45 4-3 Texas Red-X (TR) structure...................................................................................45 4-4 Aminoethyl methacrylate hydrochloride salt (AEMHS) structure........................46 4-5 Mechanism of FITC conjugation to AEMH monomer..........................................46 4-6 Mechanism of Texas Red-X conjugation to AEMH monomer.............................47 4-7 Suspension polymerization setup using mechanical stirrer and heating mantle....50 4-8 Magnetic separator apparatus................................................................................51 4-9 Spectra for Zeiss filter set 10.................................................................................58 4-10 Folic acid structure.................................................................................................63 4-11 Schematic of folic acid immobilization onto microspheres...................................64 4-12 Schematic detailing the carbodiimide mediated coupling of a carboxyl group to an amine to form an amide linkage....................................................................65 4-13 Tissue slice mounted on slide being rinsed as part of testing procedure...............74 5-1 Sample D013-4; ST-co-DEA particles dispersion polymerized in ferrofluid.......77 5-2 EDS spectra of sample D013-4 showing strong iron peaks...................................78 5-3 Diethyl aminoethyl methacrylate (DEA) structure................................................79 5-4 Samples d030_3.....................................................................................................80 5-5 XRD spectrum from sample d030_4m..................................................................81 5-6 XRD data listing from spectrum shown in figure 5-5 for sample d030_4m.........82 5-7 D009 dispersion polymerized microspheres..........................................................83 5-8 Seed particles D052 for activated swelling made by dispersion polymerization of styrene................................................................................................................85 5-9 Size graph for sample AS05 and its seeds D052_3...............................................86 5-10 SEM micrograph of sample AS05.........................................................................86 xii

PAGE 13

5-11 Sample AS08, polymerized from equal parts ST/HEMA/EGDMA......................87 5-12 SQUID magnetometer magnetic hysteresis curve for uncoated iron oxide...........89 5-13 TiO2 (10 w/v %) doped PMMA particles produced by Shims suspension polymerization process applied for most samples produced in this study.............94 5-14 Sample S04 showing particles formed by magnetite doped suspension polymerization.......................................................................................................95 5-15 UV-Visible absorption spectrum...........................................................................97 5-16 S19 particles dispersed on slide at 10x ...............................................................104 5-17 S19 microspheres, 40x fluorescent image showing fluorescence intensity inhomogeneity within microspheres....................................................................104 5-18 Confocal image of S19 microspheres at 60x.......................................................106 5-19 Light intensity plot for figure 5-18......................................................................106 5-20 Confocal microscope image of sample S25 with Texas Red dye........................109 5-21 NCI-H23 cell line tested with S11 control microspheres (left) and S11 folate immobilized microspheres (right)........................................................................111 5-22 Volume average size graphs for sample S11.......................................................112 5-23 BALB/3T3 mouse fibroblast cell line with folate-immobilized s19 microspheres........................................................................................................116 5-24 Fresh tissue sample 1 with S19 immobilized microspheres................................118 5-25 H&E stained 10 um section of squamous cell carcinoma tumour of tongue/neck region from tissue sample 1.............................................................118 5-26 10 um slice of sample 1 tissue mounted on slide and tested with control microspheres........................................................................................................119 5-27 Panel of 10 um section mounted tissue from sample 1, treated with folate-immobilized microspheres s19, 10x....................................................................121 5-28 Panel of 10 um section mounted tissue from sample 1, treated with control microspheres s19, 10x..........................................................................................122 6-1 Conjugation of fluorescein isothiocyanate to APTS monomer...........................129 6-2 Formation of silica structure by base-catalyzed condensation, incorporating APTS....................................................................................................................130 xiii

PAGE 14

6-3 TEM of dye doped silica nanoparticles, sample FITC2.....................................134 6-4 Coulter LS 230 graph showing size distribution of FITC2 FSNPs.....................134 6-5 UV-Vis absorption spectra of folate immobilized FSNPs (top), 50 uM folic acid in solution (middle) and control FSNPs (bottom)........................................137 6-6 Panel of FSNPs specifically bound to tumorous cells.........................................138 6-7 Optical (left panels) and fluorescence (right panels) confocal images of folate-immobilized FSNPs on BALB/3T3 fibroblasts...................................................139 6-8 NCI-H23 cells treated with FSNPs......................................................................140 7-1 Results matrix for testing on cell lines.................................................................144 xiv

PAGE 15

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 SUPERPARAMAGNETIC FOLATE-IMMOBILIZED DYE LABELED MICROSPHERES FOR ORAL CANCER SCREENING By Bernd Liesenfeld May 2004 Chair: Christopher D. Batich Major Department: Materials Science and Enginering A design concept is presented and developed for a screening test for oral cancer. The application is based on generating specific binding between microspheres and receptors known to be expressed specifically on malignant cells. Quantification of the test is derived from a ratiometric determination of test microspheres immobilized with folate against control microspheres. Microspheres were suspension copolymerized polymethyl methacrylate and aminoethyl methacrylate, and were doped with superparamagnetic iron oxide to permit magnetic separation of microspheres from testing suspension. Magnetic separation was demonstrated. Specific binding was provided by folic acid that was immobilized on the microsphere surface by carbodiimide chemistry. Microsphere labeling was performed by covalent bonding of fluorophores to monomers prior to polymerization, permitting spatial imaging of microspheres by fluorescence microscopy. xv

PAGE 16

Testing of specific binding of folate to tumorous cell lines was performed using cell lines known to overexpress folate receptors. Cell lines used included NCI-H23 human lung adenocarcinoma, with controls provided by normal human dermal fibroblasts. It was found that the folate-immobilized microspheres were preferentially retained by the tumourous cell line, relative to control microspheres (p = 0.0074). There was no significant difference between the retention of folate-immobilized microspheres by the cancerous cell line as compared to the control cell line (p = 0.90) as determined by pooled data. Testing of specific binding to relevant tissue was performed using excised oral cancer tissue that had been frozen and sectioned onto slides. It was found that the folate immobilized microspheres were retained by the cancerous tissue at a higher rate than the control microspheres (p = 0.037). Controls performed with normal tissue shows that the folate-immobilized microspheres were retained by normal tissue at a higher rate than the cancerous tissue. Both cell line data and tissue data show false positive responses, which may be due to non-specific binding of folate-immobilized microspheres to samples. xvi

PAGE 17

CHAPTER 1 INTRODUCTION The incidence of oral cancer is about 30,000 newly diagnosed cases annually for the U.S. meaning that roughly one in 10,000 people are affected. Approximately 90 % of these are squamous cell carcinomas (SCC). (Tabor et al. 2002; Herrero et al. 2003), Epidemiological data shows that smokers are at much higher risk for SCC than the general population, and that excessive alcohol use exacerbates the risk. It is estimated that up to 90 % of oral cancers are associated with these risk factors. Oral SCC often manifest as painless and innocent appearing keratinized ulcerations, so that the initial lesions typically remain undiagnosed until the malignancy metastasizes. When a metastasized malignancy is found, the prognosis is quite poor, and the treatment route has high morbidity. If the initial lesion were diagnosed prior to metastasis, the prognosis would be greatly improved, and the treatment routes available would be less morbid. The combination of readily identifiable risk groups, and improved outcomes for earlier detection make oral cancer an ideal application for a screening test. A novel detection system is developed that could be applied to oral cancer for screening purposes. The primary thrust of the dissertation is the development and characterization of the microsphere system that acts as the reporter particle. Emphasis is also placed on testing performed to verify functionality of the system. Much of the technical development work for this system is not unique to the ligand and target chosen to suit the particular application. Once regard is taken for the peculiarities and specific handling requirements unique to a receptor-ligand binding system, the technology can be 1

PAGE 18

2 applied in a more general manner. It is believed that this type of screening system could be useful for numerous conditions. The structure of the dissertation is to present the background on the application and relevant fields of study pertaining to the research, followed by a detailed description of the envisioned application system that is ultimately the goal of the research group to develop. The technical requirements of the applications are in turn used to develop the design parameters of the particles. The materials and methods section provides intimate detail on how the product was synthesized, characterized, processed and tested. A results and discussion section is provided that details the findings and interprets them as relevant to the research. A largely independent chapter on dye doped silica particles is inserted prior to the conclusions. This chapter discusses synthesis and characterization as well as some experiments conducted with fluorescent silica nanoparticles. This system was applied as a model for the chemistry of the microspheres in situations where physical characteristics of the microspheres proved to be confounding to important characterizations. The silica nanoparticles had analogous surface chemistry to the microspheres. Conclusions are based on the findings from the microsphere-based research, supplemented by information that could only be derived using the model system.

PAGE 19

CHAPTER 2 BACKGROUND 2.1 Magnetism Magnesia (in the Thessaly region of Greece) lent its name to a line of gentle, wise and just centaurs the Magnetes, mythically descended from the Magnesian Mares that birthed the first centaurs. Eventually the word passed into more modern language as magnates, at one point describing landowners or medieval noblemen, but in modern times referring to a great man or one that is particularly important or influential in some fieldparticularly business. Our word Magnetism is also descended from the Thessalian inhabitants the Magnates, due to the local abundance of lodestonea magnetized form of the mineral magnetite (Fe3O4). Lodestone (or its equivalent loadstone) has magnetic polarity, enabling the Chinese mathematician ShenKua (1030 1093 AD) to build the first recorded navigational compass by using the permanently magnetic lodestone to magnetize a soft iron compass needle (since the magnetic response of the soft iron needle would fade over time a lodestone had to be carried to remagnetize the compass needle occasionally). While magnetite is relatively abundant, lodestone is not as common. Lodestone turns out to be an intimate mixture of magnetite and maghemite that is typically found very close to the surfacetypically in volcanic regions (Wasilewski and Kletetschka 1999). As rocks of the proper composition cool, they are magnetized by the earths magnetic field, and freeze in that magnetic orientation. This permits researchers to 3

PAGE 20

4 study deposits of magnetized rock for a historical recording of shifts in the earths magnetic field, or to use these patterns of shifts to date geological structures. Figure 2-1: A lodestone with nails and magnetite fragments attached. Sourced: Moskowitz (1991), Hitchhiker's Guide to Magnetism, http://www.geo.umn.edu/orgs/irm/hg2m/hg2m_index.html, February 2004 2.1.1 Types of Magnetism There are five major groups of magnetic materials, classified by their type of magnetic behaviourtheir response to an applied field. These five are diamagnetism, paramagnetism, ferromagnetism, ferrimagnetism and antiferromagnetism. Of these, the first two do not show collective magnetic interactionsonly the last three have longrange magnetic order. The materials that are generally considered as magnetic are ferromagnets and ferrimagnetsall other groups have relatively weak magnetic properties as expressed by the materials bulk susceptibility: The bulk susceptibility is the slope of the M vs. H curve, and is used as a measure of the strength of a materials magnetic response. Some basic magnetic terms and their units are detailed in table 2-1.

PAGE 21

5 Table 2-1: Units used for magnetic quantities Term Magnetic quantity SI units CGS units B Magnetic Induction T (tesla) G (gauss) H Applied field A/m (ampere / meter) Oe (oersted) M Magnetization A/m (ampere / meter) G (gauss), emu/cm3 In terms of equations (using SI unit formulations) this yields: B = o (H + M) (equation 2-1) M = H (equation 2-2) Where o is the permeability of free space. Detailed below are the types of magnetic responses in terms of the magnetization vs. applied field curves that are generated (the response diagrams are not drawn to scale and indicate only the trend of behaviour, not the magnitude relative to other curves). Some texts show B vs. H while others show M vs. H curves. As far as illustrating the relevant trends, these are largely the same. MH slope = Figure 2-2: Diamagnetic response 2.1.1.1 Diamagnetism All matter is diamagnetically responsive, but the effect is very weak and can be completely masked by stronger magnetic properties (ferroand ferri-magnetism). Diamagnetism arises from the non-cooperative behaviour of orbiting electrons when exposed to a magnetic field, and is not temperature dependent. As seen in figure 2-2, diamagnetic materials have a small negative magnetic susceptibility: 10-5, so that the

PAGE 22

6 magnetic response opposes the applied magnetic field. Superconductors are a special class of diamagnetic materials that have a susceptibility -1. MH slope = Figure 2-3: Paramagnetic response 2.1.1.2 Paramagnetism Some of the atoms or ions in this class of materials have net magnetic moments due to unpaired electrons in partially filled orbitals. For paramagnetic materials the susceptibility is a very small positive value of approximately 10-3 10-5. The magnetization of paramagnets is aligned parallel to an applied magnetic field. The form of the paramagnetic magnetic response curve is shown in figure 2-3. 2.1.1.3 Ferromagnetism Ferromagnets are the most widely recognized group of magnetic materials. Susceptibility values are large and positive: 50 10,000. This is due to strong interactions by electronic exchange forces (a quantum mechanical phenomenon that arises from the relative orientation of the spins of electrons) resulting in parallel or antiparallel alignment of magnetic moments. These materials are capable of being magnetized up to their saturation magnetization (the maximum induced magnetic moment), by relatively weak applied fields. At temperatures above their Curie temperature, the magnetic behaviour changes to paramagnetic (and the remnant

PAGE 23

7 magnetization goes to zero). Ferromagnets exhibit the magnetic property of hysteresis, and are able to retain a certain magnetization when the applied field is removed. Those ferromagnets that retain a large percentage of their saturation magnetization when the applied field is removed are called hard magnets, whereas those that lose most of their magnetization are termed soft magnets. This is illustrated in figure 2-4, which shows a hysteresis curve with relevant parts of the curve labeled. Starting from the origin, the magnetization curve slopes upwards as a filled H is applied at an initial susceptibility of o, eventually reaching saturation magnetization Ms at point A. Reducing and reversing the applied field the magnetization does not follow the same curve again this effect being called hysteresis. The curve proceeds along past point B and at zero applied field retains a magnetization Mr the remnant magnetization. This is the amount of magnetization that the material will retain when not in a magnetic field the permanent magnetization of the material. Following the curve to point D reaches the intersection with zero magnetization at applied field -Hc representing the coercive force Hc, also known as the materials coercivity that is the amount of field (in the opposite direction) required to demagnetize the material. Further increases in the applied field to point E generate a magnetization in the negative direction up to the saturation point Ms, reached at point E. Any further increases in field magnitude over E (negative direction) or A (positive direction) can produce no further increase in magnetization past the saturation points this is the state of the material being fully aligned in its magnetic state. Reversing the curve again produces the same type of curve through points F (the negative remnant magnetization) and Gthe coercivity. Further reversals will traverse the path around the loop indicated by ABCDEFGA This loop is called the hysteresis loop.

PAGE 24

8 Figure 2-4: Hysteresis loop. This shows the magnetic response of ferromagnetic and ferrimagnetic materials below their Curie temperatures 2.1.1.4 Ferrimagnetism Ferrimagnets are ionic compounds that display both ferromagnetic and antiferromagnetic properties, and exhibit a small positive magnetic susceptibility that increases with temperature. Macroscopically, ferrimagnets behave in much the same way as ferromagnets, exhibiting hysteresis and saturation in their magnetic curves. Magnetite is perhaps the best known ferrimagnet, with an inverse spinel crystal structure. For ferrimagnets, there are A and B type ions that occupy different lattice sites (figure 2-5)for magnetite the structural formula resolves from the molecular Fe3O4 to [Fe3+]A [Fe3+, Fe2+]B O4. The trivalent ferric ions occupy tetrahedral sites where they are surrounded by four oxygens, while the divalent ferrous ions occupy octahedral sites and

PAGE 25

9 are surrounded by six oxygens. The A sublattice spins are antiparallel to the B sublattice spins and with the negative AB exchange interaction, the B site ferrous ions contribute the net magnetic moment of magnetite. Figure 2-5: Inverse spinel structure of magnetite showing A and B lattice positions. Sourced: Moskowitz (1991), Hitchhiker's Guide to Magnetism, http://www.geo.umn.edu/orgs/irm/hg2m/hg2m_index.html, February 2004 2.1.1.5 Antiferromagnetism Antiferromagnets have an initial low positive susceptibility (similar to a ferromagnet) up to their Nel temperature, above which susceptibility decreases with increasing temperature (like a paramagnet). This property occurs in certain inorganic compounds where neighbouring atoms interact in such a manner as to produce an antiparallel arrangement of magnetic dipoles that is equal and opposite.

PAGE 26

10 2.1.2 Domain Size Effects Magnetic domains in ferrimagnetic materials are much larger than atomic dimensions, but small in a macroscopic sense, existing at a scale of single to 100s of m. Domains can be the result of particle or grain size, or they can form to minimize energy within a single crystal. Figure 2-6 illustrates the process of domain wall formation as it minimizes the system energy. Figure 2-6: Energy minimization by domain wall. Sourced: Moskowitz (1991), Hitchhiker's Guide to Magnetism, http://www.geo.umn.edu/orgs/irm/hg2m/hg2m_index.html, February 2004 Magnetostatic energy (the energy associated with the surface charge distribution) is minimized by subdividing a single domain into multiple domains. The surface charges at the ends of a domain form a demagnetizing field. When a domain is split into two, the magnetostatic energy is reduced by almost half. The demagnetizing field is reduced by bringing opposite charges closer together. The formation of each new domain wall requires a transition between the alignments of the spins. A gradual transition between spins results in a wide wall (Figure 2-7a) that minimizes the exchange energy. A sharp

PAGE 27

11 transition (Figure 2-7b) between spin states minimizes the anisotropy energy. The balance between these competing forces produces domain wall widths of ~ 100 nm. Figure 2-7: Domain wall transitions. Sourced: Moskowitz (1991), Hitchhiker's Guide to Magnetism, http://www.geo.umn.edu/orgs/irm/hg2m/hg2m_index.html, February 2004 2.1.2.1 Single domains Below a certain grain size, domain walls are no longer energetically feasible, and the grain (or particle) will contain a single domain (SD) that will be uniformly magnetized to saturation magnetization. It is an energetically difficult process to rotate the magnetization of an SD. SD grains are therefore magnetically hard (high coercivity and remnance). For magnetite the transition from multiple domains to SD is around 80 nm (Leamy 2003). 2.1.2.2 Superparamagnetism When an SD particle is small enough, and/or temperature is high enough, then the thermal energy can overcome the anisotropy energy that separates opposite magnetization states, and spontaneous reversal of magnetization occurs. Remanence and coercivity thus become zero, and the grain / particle becomes superparamagnetic. A superparamagnetic particle will have zero net magnetic moment when no field is applied. Under the

PAGE 28

12 application of a field, a net statistical alignment of the magnetic moments will occur. This behaviour is similar to paramagnetism, but on a scale of perhaps 105 atoms rather than for a single atom, generating a much higher susceptibility. The transition size of a particle from SD behaviour to superparamagnetic behaviour has been calculated (see equation 2-3) as between 22 and 23 nm, for spherical pure magnetite at room temperature. 1/t = fo exp ((-Kv)/(kT)) (equation 2-3) where fo = frequency factor (10-9sec-1) Kv = anisotropy constant = particle volume k = Boltzmann constant T = absolute temperature The practical value of particles with superparamagnetic properties is that they will show a strong magnetization in response to an applied field allowing the generation of a significant motile force. When the applied field is removed, no remnant magnetization remains, and the particles have no drive to agglomerate. This type of response is easy to verify on macroscopic samples with equipment no more complex than a magnet and some superparamagnetic particles. 2.1.3 Some Applications of Magnetic Particles There are a large number of applications that exploit the magnetic properties of materials, beginning most obviously with the magnetic compass needle long ago, to such mundane current items as refrigerator magnets. Many high tech applications such as magnetic storage media (cassette tapes through hard drives), and medical imaging technologies exploit the magnetic properties of fine particulates. Detailed below are

PAGE 29

13 samples of biomedical applications that utilize magnetic particles (mostly superparamagnetic). 2.1.3.1 Hyperthermic treatments Hyperthermia is a clinical treatment used to combat tumours that involves the artificial elevation of temperature in the body. Mild hyperthermia (under 42o C) is used to stimulate the immune system of the body, while higher temperatures (about 45o C) are applied to attempt to destroy specific (cancerous) cell populations. The mechanism of hyperthermic treatment is to introduce coated magnetic particles to the site in question and heat them through application of an alternating current (AC) field. This method is called inductive heating, and provides a way to heat only the tissue in immediate contact with the magnetic particles. Targeting the particles to the desired tissue thus represents the key to successful hyperthermia applications. The underlying physical phenomena that produce heating vary by the class of magnetic material utilized, but are usually quoted as a specific absorption rate (SAR) so that comparisons are viable. Ferromagnetic or ferrimagnetic (FM) particles rely on hysteretic losses to generate heat. Quite high magnetic fields can be required to achieve saturation magnetization higher than generally considered clinically viable for humans. Superparamagnetic (SPM) particles generate heat when a field is applied and removed by relaxation either of the particles themselves when in fluid (Brownian rotation,) or by rotation of the atomic magnetic moments within the particles (Nel relaxation). The latter process occurs at higher frequencies, and is the only process available if the particles are not free to rotate physically. Owing to the lower applied field (well within limits considered safe for human clinical use) required to generate heating in SPM

PAGE 30

14 particles, these are the primary avenue of development in the field (Pankhurst et al. 2003). 2.1.3.2 Contrast media Paramagnetic and superparamagnetic particles have been applied as contrast agents for magnetic resonance imaging (MRI) applications, since they alter the relaxation times of the tissue in which they are resident. This occurs through the particles enhancement of the local magnetic field strength due to the influence of their own induced magnetic fields, and by their creation of field inhomogeneities. In order to obtain specific image enhancements for features of interest, the key is to optimize the targeting of the particles to the relevant tissue or features, since the contrast enhancement effects are very short range. 2.1.3.3 Particle guidance by magnetic forces There are many variations on this concept that are in current clinical use. The common thread through these is that there are two components to the applications that are critical to their success. The first is that the particles are responsive to a magnetic field such that they can be retained at some site of interest (a target organ perhaps, or a collection site). The second is a chemical modification of the particle to carry a payload this might be a label or drug or antigen that will either facilitate attachment of the particle of interest to the target or that will provide a therapeutic or analytical service. Applications that can be categorized under this heading include magnetic separations and magnetically targeted drug delivery. To make particle guidance applications feasible, the motive force on particles needs to be sufficient to move the particles over some distance in a viscous fluid. Fluids are often aqueous or blood, and may be flowing at some rate w (velocity of water) such

PAGE 31

15 that a = m w where m represents the magnetic particle velocity (zero if particles are being retained in a fixed position by a fixed magnet). The hydrodynamic drag force Fd then equates to equation 2-4: Fd = 6 Rm (equation 2-4) In the above equation represents the viscosity of the medium, and Rm represents the radius of the magnetic particle. This equation shows the force that must be generated between the applied field and the magnetic particle to retain the particle in a flowing medium. The magnetic force Fm generated by a particle is given by equation 2-5 below: Fm = Vm ( B H) (equation 2-5) Where Vm is the volume of the magnetic particle (if the magnetic core is coated, or a dispersion of magnetic nanoparticles exists within a microparticle then the magnetic particle volume may not be equal to the hydrodynamic volume for the particle). The term ( B H) represents the differential of the magnetostatic field energy density, indicating the direction of magnetic force. Magnetic separations include several applications such as magnetic cell sorting (magnetic particles are coated with a reagent that selectively attaches to target cells, permitting the separation of bound cells) and immunoassays (methods that employ antibodies to measure analyte concentrations). These applications rely on the magnetic component of the particle to enable recovery from a flow system with the efficiency of the process being proportional to the flow rate at which the particles can still be recovered. Targeting can be tailored by careful selection of the species on the particles

PAGE 32

16 that generate specific binding allowing multiple targets to be accessed in a single separation. Figure 2-8: Schematic for drug delivery system. A magnet external to the body retains magnetic carriers at target site Drug delivery is another application to benefit from magnetic carriers to permit more localized delivery. Localized drug delivery is desirable for many drugs such as anti-inflammatory agents or chemotherapeutics that are toxic in the doses required if administered systemically. Localized targeting permits an overall lower drug dose to yield a higher local concentration, reducing sideeffects. Drug deliver applications are typically designed around magnetite or maghemite magnetic particles either with a magnetic core or with magnetic nanoparticles dispersed within a polymer microparticle. Biocompatible polymers often used include polyvinyl alcohol (PVA) and dextran. The drug payload can be either attached to the surface of the particle or stored inside if the particle. The magnetic drug carrier particles are typically introduced into the body as a ferrofluid and then detained at the target region by means of a strong applied field as shown in figure 2-8. Hydrodynamic investigations of blood flow and magnetic retention suggest that successful magnetic drug delivery is more likely for regions of slower blood

PAGE 33

17 flow where a magnet can be placed in close proximity (Voltairas et al. 2002; Pankhurst et al. 2003). 2.2 Cancer 2.2.1 General Cancer Background Cancer as a disease has been known to society for a very long time, with sources as far back as 2500 BC recording surgical treatments in ancient Egypt. It is believed that the ancient Egyptians were able to distinguish malignant from benign tumours (tumour meaning literally a swelling or lump). Only malignant tumours are termed cancerous, their cells invading the basement membrane to spread to other parts of the body in a process called metastasis. Benign tumours do not invade other parts of the body, though they may continue to grow, which can still be a problem in a constrained setting (i.e. the brain). The term cancer is credited to Hippocrates (~430 360 BC), who noted the crablike form of a tumour (karkinoma being Greek for crab). The simplest modern definition of cancer is provided by the American Cancer Society (ACS): cancer is a group of diseases characterized by uncontrolled growth and spread of abnormal cells. Cancerous cells growing out of control are unable to recognize the signals that cause normal cells to stop growing at their natural boundary, to specialize to perform some useful function, or to stop replicating and die. Cancers are classified by two characteristics: their histological type (categorized by tissue type wherein the cancer originated) and location (primary site of origination). There are five major categories of cancers: carcinoma, leukemia, lymphoma, myeloma and sarcoma. Carcinomas are malignancies of epithelial tissue that have a tendency to metastasize to other parts of the body, and account for 80 to 90 percent of all cancer

PAGE 34

18 cases. There are two major types of carcinoma, categorized by site of development. Adenocarcinomas develop within an organ or gland, while squamous cell carcinoma (squamous meaning flat or scale-like) develop within the squamous epithelium. Sarcomas are cancers of supportive or connective tissues, i.e. bone, muscle, tendon, cartilage, muscle and fat. Myelomas originate in plasma cells of bone marrow. Leukemias are also cancers of bone marrow specifically as relates to the production of white blood cells. Lymphomas are cancers of the lymphatic system arising in lymphoid tissue or lymph nodes. 2.2.2 Oral Cancer In the U.S. head and neck squamous cell carcinomas (HNSCC, see figure 2-9) represent approximately 4 % of all newly diagnosed cancers, or about 30,000 new cases yearly (Tabor et al. 2002). Over 90 % of cancers of the oral mucosa and the lips vermilion are squamous cell carcinomasthese are what is classically referred to as oral cancer (Auclair and Rasmussen 2002). HNSCC is a serious health problem with a well defined atrisk population. Heavy alcohol indulgence acts as an independent multiplicative factor for the primary risk factor: tobacco use (Mashberg and Samit 1995). Smokeless tobacco products (such as dip, chew or tobacco pouches) may represent an even greater health challenge than smoked tobacco. In some Asian countries, where chewing of betel leaves supplements or substitutes for use of tobacco products, the oral cancer rate soars to 40 % of new cancers (Pande et al. 2002). Efficient diagnosis of oral cancers presents a number of difficulties: the hidden nature of many sites in the oral environment, the often innocuous appearance of the lesions, a lack of pain or clear lumps to guide the clinician (figure 2-9 and 2-10), and the

PAGE 35

19 abundance of similarappearing harmless ulcerations. Detection frequently occurs only when the tumour has metastasizedoften as a lump in the neck. At this later stage the prognosis is quite poor, worse than for breast, cervical, or prostate cancers (Canto et al. 2002) Figure 2-9: Oral squamous cell carcinoma. The small, granular red lesion shown by the arrow was totally asymptomatic and was only slightly thickened. Sourced: Bristol Biomedical Image Archive (2002), http://www.brisbio.ac.uk/ROADS/subject-listing/carcinomasquamouscell.html April 2004. Clinicians were found, in a survey (Mashberg and Samit 1995), to underestimate the incidence of the disease, and hence not routinely examine as carefully as possible for it. Examination is complicated by the innocuous appearance of the lesions, which is not easily discriminated under the lighting conditions available (head lamps, pen lights). Staining with toluidine blue provides a sound method for detection, but suffers the confound that all ulcerations accept stain, requiring a subsequent confirming treatment some weeks later (Epstein and Scully 1997). The above factors combine to make late diagnosis the more typical situation, as confirmed by a recent perspective (Friedlander 2003). Friedlander concludes that early

PAGE 36

20 detection may be the most important strategy to improve overall survival for patients with head and neck cancer. Currently one third of patients present with HNSCC present with early-stage disease, whereas two thirds present with latestage disease. Contributing factors to late presentation include lack of a definite symptom complex. Figure 2-10: Squamous cell carcinoma of the tongue. Lesions can be painless and indistinct in spite of being in advanced stages. Sourced: Ghorayeb (2004) Pictures of Tongue Cancer, http://www.ghorayeb.com/TongueCancer.html April 2004. At the late stage the simple excision of the offending mass is no longer sufficient treatment and radiation based therapy is indicated in order to prevent further metastases, and/or to combat metastases that have already occurred. Radiation based therapy in the region of the oral cavity is highly undesirable even when successful since a common side effect is the inactivation of the radiationsensitive salivary glands. A deficit in salival production brings a host of complications to the patient, including reduced taste perception, increased tooth decay rates, and drymouthall contributing to reduced quality of life. Calls have been made for research into improved detection methods (Field and Jeffcoat 1995); if a malignancy could be detected before it has a chance to

PAGE 37

21 metastasize, then a purely local surgical intervention should provide a complete and final treatment [the continued existence of risk factors notwithstanding]. 2.3 Folic Acid and Receptor Targeting 2.3.1 Folic Acid Folic acid (also known as vitamin B9 or Bc), is needed by cells in the synthesis of DNA nucleotides and it cannot be synthesized by the cell. Cell populations that are growing rapidly (i.e. tumorous growths) require more folic acid, and upregulate the production of folic acid receptors (Antony 1996; Suh et al. 2001). The resultant overexpression of folate receptors is exploited in cancer cell targeting by attaching folate to genes, contrast agents and drugs. An example is shown in figure 2-11, where an anticancer agent linked to folate makes use of the folate receptor in two ways: first by selecting the proper cell by attaching to the folate receptor, and then gaining entry into the cell via receptor mediated endocytosis in a Trojan Horse strategy. Folic acid, as shown in figure 2-12, is composed of three molecular fragments: (left to right) 6 methylpteridin, P amino benzoic acid (PABA) and glutamic acid. The methylpteridin end of the molecule is the biochemically active moiety that participates in the folic acid cycle (Campbell 1991; Campbell and Smith 1994) and through which the receptor binding occurs. In order to preserve receptor function, folate conjugation has to occur at the -carboxylate group of the glutamic acid fragment (Leamon and Low 1991). Tritium labeled folate attached via the -carboxyl group demonstrated the same rate of endocytosis as free folic acid, while attachment via the amine group of the methylpteridin fragment abrogated the receptor binding (Kranz et al. 1995; Wang and Low 1998).

PAGE 38

22 Figure 2-11: Mechanism of receptor mediated endocytosis used to target anti-cancer drugs to tumourous cells. Sourced: Purdue News (2002), Researchers developing technology to outsmart metastasized cancers, http://www.purdue.edu/UNS/html4ever/020619.Low.Endocyte .html, April 2004 Figure 2-12: Structure of Folic Acid 2.3.2 Folate Receptors Folate receptors exist in a number of isoforms: FR-, FR-, FRand a soluble form (sFBP: soluble folate binding protein). Of these, only the first two are relevant to targeting applications. FRwas recognized first as a tumour marker, with cancers of the

PAGE 39

23 ovaries, kidneys, uterus, testes, brain and colon, as well as adenocarcinomas of the lungs being found to prominently overexpress the isoform of the folate receptor. Significant expression of FRon nonmalignant cells occurs only in the kidneys, lungs, choroid plexus and placenta, where in all but the latter case the receptors were situated on the apical membrane surface. The primary folate transport mechanism for normal cells is thought to occur via the reduced folate carrier a plasma membrane transport protein that reacts preferentially to reduced folate but has no capacity to transport folate-drug conjugates, since it is not anchored to the cell membrane as are the FRand FR(both FRand FRare glycosyl phosphatidylinositol linked membrane proteins). FRis overexpressed in many cancers of the breast, brain, mesenchymal tissue, testes, head and neck (squamous cell) and hematopoietic cells (granulocytic lineage). FRhas a tenfold lower folate binding affinity than FR-, with KD ~ 10-9M and 10-10M respectively (Reddy and Low 1998; Wang and Low 1998). Estimates on folate receptor overexpression rates in malignant tissues are quite variable, and may not distinguish well between heterogeneous cell populations that are grossly part of the same complex. Literature indicates that the expression of folate receptors for squamous cell carcinoma of the oral mucosa exceeds the rate of expression on otherwise identical, non-malignant tissues by >20 fold (Ross et al. 1994) 2.4 Fluorescence Histology is the study of the microscopic structure of plant and animal tissue. It is typically performed by staining tissues to highlight features of interest, or increase contrast between features. The use of dyes permits the more sensitive detection of features. The use of fluorescent dyes permits further increases in sensitivity that can

PAGE 40

24 enable visualization of subcellular features. Fluorescence microscopy makes use of fluorescent dyes, in many cases conjugated to specific binding agents, to generate images that uniquely highlight specific features. Figure 2-13: Jablonski diagram showing energy states for a fluorescence process. Reprinted with permission from Molecular Probes, Haugland, R. Handbook of fluorescent probes and research products, web edition, http://www.probes.com/handbook/figures/0664.html January 2004 2.4.1 The Fluorescence Process Fluorescence is found mainly in heterocyclics or polyaromatic hydrocarbons. It is a three stage process depicted in the Jablonski diagram of figure 2-13 (Haugland 2003). An energy input in the form of a photon of light hEX excites an electron from the S0 ground state to the S1 excited singlet state energy level (1). The excited singlet state exists for a short amount of time (1 10 nanoseconds) before decaying into the relaxed singlet excited state S1 (2). When the electron returns to the ground state a photon of energy hEM is emitted (3). The energy dissipation between S1 and S1 (hEX hEM) is the Stokes shift the difference between the energy of the photon causing excitation hEX and the photon emitted hEM. The emitted photon is at a lower energy and thus at a longer wavelength. This phenomenon is partly responsible for the sensitivity that can be

PAGE 41

25 achieved by fluorescence techniques, since it allows the separate detection of emission electrons against a low background since the excitation signal is at a different wavelength that can be removed (i.e. by optical filters). Another factor that contributes to the high sensitivity achievable is fluorophores are not consumed in fluorescence and can be repeatedly excited to generate many thousands of detectable photons. This is because the excitation emission process is cyclical, and the excited electron returns to the ground state via path (3) in figure 2-13 unless photobleaching occurs (essentially an oxidative degradation process that can occur in the excited singlet state S1). For a polyatomic molecule in solution, the discrete energy level transitions depicted in the Jablonski diagram of figure 2-13 are replaced by broader energy spectra for excitation and emission. These spectra are depicted in figure 2-14, which also shows the equivalency between the emission and excitation intensities: excitation at intensity EX 1 results in emission at that same intensity, while excitation at EX 2 yields emission at EM 2. This same figure also illustrates that the fluorescence emission spectrum is largely independent of the excitation wavelength. 2.4.2 Fluorescence Techniques Fluorescence techniques enable large numbers of very interesting biological characterizations and analytical processes making them big business for the scientific community. There are sizable companies specializing in the production of fluorophores (Haugland 2003), especially high value added products such as fluorophores conjugated either to general reactive groups for easy applications in individual research, or fluorophores labeled proteins (and nucleic acids, lipids, etc.) known to be important to common research and diagnostic applications.

PAGE 42

26 Figure 2-14: Polyatomic molecule spectra showing excitation and emission intensity equivalence Reprinted with permission from Molecular Probes, Haugland, R. Handbook of fluorescent probes and research products, web edition, http://www.probes.com/handbook/figures/0665.html, January 2004 There are four primary types of fluorescence instruments that provide the ability to access different types of information. A fluorescence microscope yields spatial information, where identification of items, regions or reactions of interest is monitored by their fluorescence signals on a microscopic scale. Flow cytometers measure fluorescence of individual particles (cells mostly) in a stream, permitting quantification of labeled groups within a larger population, but can only excite at a single fixed wavelength. Fluorescence scanners identify fluorescence signals from such sources as electrophoresis gels spatially on a macroscopic scale. Spectrofluorometers measure an average signal coming from a bulk sample (such that the data acquired resembles the spectra of figure 2-14 more than the discrete energy bands of figure 2-13. The fluorescence spectrometer is perhaps the most flexible of these instruments since it provides a continuous range of excitation and emission wavelengths that can be monitored. The absorption spectrometer is a common instrument that is used to acquire similar data but is not included in this list of fluorescence instrumentation since it does not measure emission. The excitation

PAGE 43

27 spectrum of a fluorophores species is essentially the same as its absorption spectrum, but the absorption spectrum can be influenced by particulate phenomena such as small particle based scattering. Interesting applications of fluorescence detection in molecular biology include the use of green fluorescent protein (Tsien 1998) as a tag for gene expression, as well as fluorescence in situ hybridization (FISH) for gene mapping, etc. (Haugland 2003). Systems have been designed that enable four to five different fluorescent signals to be resolved with optical filters only more when linked to interferometers and sophisticated software. Multicolor labeling experiments involve the introduction of two or more probes for resolution or monitoring of different objects or reactions. This technique is widely applied in FISH, flow cytometry, DNA sequencing and fluorescence microscopy (Herman 1998). Selecting dyes with narrow bandwidths and maximum spectral separation maximizes resolution. Figure 2-15 shows the emission spectra of three dyes commonly used in fluorescence microscopy that have good spectral separation: fluorescein, tetramethylrhodamine and texas red. Figure 2-16 shows an image created for a human metaphase chromosome hybridization experiment, using red and green fluorophores. 2.5 Polymerization Methods for Producing Microspheres The definition of polymerization types is predominantly phenomenologically based. The methods of interest are heterogeneous processes leading to the formation of small particulates (up to 1 or 2 mm in size) by means of addition polymerization of vinyltype monomers. The initial reagents for all systems include a monomer (or mixture of monomers) to be polymerized, an initiator and a polymerization medium (a liquid phase). In many cases there is a stabilizer added to the mixturewhich may be

PAGE 44

28 called an emulsifier in some systems. Other additions to the mixture include a crosslinker if the final product is desired to be insoluble, as well as dopants or additives to impart desirable properties (such as magnetic response, colour or radio-opacity) or chemistries (i.e. surface chemistries that facilitate binding reactions for labeling or targeting) to the particles. Figure 2-15: Normalized fluorescence emission spectra of fluorescein (FL), tetramethylrhodamine (TMR) and Texas Red (TR) dyes. Reprinted with permission from Molecular Probes, Haugland, R. Handbook of fluorescent probes and research products, web edition, http://www.probes.com/handbook/figures/0833.html, January 2004 There are four fundamentally different techniques commonly employed for the formation of particulate polymers from monomers in solution. These are emulsion, dispersion, suspension and precipitation polymerizations. These can be distinguished by several factors, including the size range of particles each can generate, the physical process applied (including reactor design), and the manner in which polymerization takes place within the system. Solvent evaporation is also a popular process, but this is not a

PAGE 45

29 polymerization process, since polymer is dissolved in solvent and reshaped rather than monomers being polymerized. There are also instances where the above processes are performed with seed particles. Figure 2-16: fluoro probes hybridized to human metaphase chromosomes. Reprinted with permission from Molecular Probes, Haugland, R. Handbook of fluorescent probes and research products, web edition, http://www.probes.com/handbook/figures/0674.html, January 2004 2.5.1 Emulsion Polymerization Emulsion polymerization is known to produce the smallest particle sizes of the methods mentioned typically in the range from approximately 50 nm to just below 1 m. The monomer is only slightly soluble or insoluble in the medium, but a surfactant is added to allow emulsification. An initiator that is soluble in the medium, but not the monomer is used. The monomer exists in the medium as both droplets and as surfactant enabled micelles of nanometer size. Initiated oligomers eventually produce stabilized nuclei, and become the loci of polymerization for the system. The original micelle size does not determine the eventual particle size produced since the initiator is not soluble in the monomer. The product is formed by the stabilized micelles containing radical

PAGE 46

30 bearing oligomers, and these tend to grow to largely the same size, so that nanospheres of low polydispersity can be formed. 2.5.2 Soapless Emulsion Polymerization Soapless or emulsifier free emulsion polymerization is very descriptively named; in the absence of micelle formation, nucleation occurs by precipitation of radical bearing oligomers and macromolecules. These coalesce since they are not stabilized in any way. Literature indicates that growing particles eventually are somewhat stabilized by the charges on their chain end groups forming electrostatic charges. Typically, the product size is slightly larger than for emulsion polymerized particles (up to micron size), and the initial monomer concentration is slightly lower. 2.5.3 Dispersion Polymerization A dispersion polymerization utilizes an initially homogeneous mixture of medium and stabilizer, in which the monomer and initiator are soluble. Phase separation of the growing oligomers / polymer radicals occurs because the medium is designed to be a poor solvent for the polymer forming the nuclei of the primary particles. The locus of polymerization for this type of system lies in the precipitated nuclei, as they swell with monomer from the homogeneous medium / monomer mixture. This mechanism leads to spherical particles that range in size from approximately 100 nm to 10 m. The stabilizer is a critical component in this system without it the particles tend to coagulate during formation. Stabilizers for dispersion polymerizations are polymer compounds with low solubility in the medium and good affinity for the polymer particles formed. Particle size control is affected by identity and concentration of the stabilizer, as well as by control of the polymerization temperature and adjustment of the medium to alter the degree of polymerization at which the particles precipitate out of solution. The medium is often an

PAGE 47

31 aqueous alcohol mixture (with the solubility parameter adjusted by means of water content) tailored to compatibilize the solubility parameter of the medium with the monomer. It is possible to produce particles with very low polydispersity via dispersion polymerization. 2.5.4 Precipitation Polymerization Precipitation polymerization is initially like dispersion polymerization in that the reaction mixture is a homogeneous solution. The important difference is that once the particles precipitate out of solution, they are not swollen polymerization takes place entirely within the continuous phase. The precipitated nuclei coagulate to form irregular particles within a size range that is similar to dispersion polymerization at approximately 100 nm to 100 m. 2.5.5 Suspension Polymerization Suspension polymerizations are generally capable of forming a product that is from a few m to a few mm in size, with significant polydispersity in the eventual size of the formed microspheres. An initiator that is soluble in monomer is used, and the resulting monomer / initiator mixture is insoluble in the polymerization medium. The monomer mixture is injected into medium, where the small droplets that form essentially comprise microreactors. The reaction can be seen as a small-scale bulk polymerization within the droplet, with excellent heat transfer characteristics due to the large amount of surface area available. If the monomer mixture is prepared with a diluent, then the droplet can be seen as a small-scale solution polymerization. The size of particles formed in suspension polymerization is controlled predominantly by stirring and monomer injection. Anything that acts to create smaller

PAGE 48

32 droplets initially or that acts to break up existing droplets leads to the formation of smaller particulates by suspension polymerization. Other factors that influence the size of particles produced include the monomer to suspension medium ratio, the stabilizer concentration and the viscosities of monomers and medium. The control of a suspension polymerization is based largely on empirical observation and iterative solutions due to the complicated interaction of these parameters, making it as much an art as a science. The morphology of particles produced by suspension polymerization relates to the swelling reaction of the polymer to the monomer. Particles composed of polymer that is swellable / soluble in its monomer tend to produce a smooth surface. A rough surface is produced on particles that are not swellable by their monomer. Particle porosity can be tailored by the addition of suitable monomer diluent (causing porosity during particle formation), and control of the amount of crosslinking (to retain porosity on removal of diluent).

PAGE 49

CHAPTER 3 PROPOSED STRATEGY AND DESIGN REQUIREMENTS 3.1 Appropriate Uses of Screening Tests Factors that should be considered in the application of a screening test include socioethical as well as scientific parameters (Grimes and Schulz 2002). Whereas screening in modern society is correctly viewed to be a valuable medical tool, it is important to note that not all screening tests that are technologically feasible would provide any statistical benefit to the population. A good screening test should be sensitive, applicable to a defined at-risk group, and it must hold reward for both the patient and the community. Both the individual and the community must see benefits in terms of improved overall health (statistically), and in terms of reduced expenditures for health care. Screening must be understood to have ethical implications that arise from the fundamental difference between a diagnostic and a screening test. Diagnostic medical tests are performed in response to a patient complaint. Screening is performed on a predominantly healthy population. The negatives associated with screening tests are the expense involved, patient inconvenience / morbidity (i.e. colonoscopy), and the possibility of false positive results (which bring fiscal and emotional consequences). This leaves aside the possibility of false negatives, which at least leave the patient in substantially the same position as before being screened. The positive aspects of screening must outweigh these negatives, by improving general health and reducing total healthcare costs. 33

PAGE 50

34 Screening test sensitivity and specificity are critical parameters, as is the disease prevalence in the population being screened. The positive predictive value of a screening test is equal to the number of true positives divided by the number of total positive test results (see the treatment matrix of figure 3-1, and equations 3-1 through 3-3). The higher the number of disease-negative subjects that are tested, the lower the positive predictive value of a test. Put simply, the positive predictive value of any given test is higher when the incidence of said disease is higher in the tested population. This leads to the conclusion that a screening test is more appropriate for a condition for which an at-risk population can be readily defined. In the case of screening tobacco users this criteria is indeed met. DISEASEPositiveNegativeTruepositiveFalsepositiveFalse negativeTrue negativeabcdTESTNegativePositive Figure 3-1: Results matrix for disease state vs. test result Specificity = d / (b + d) (equation 3-1) Sensitivity = a / (a + c) (equation 3-2) Positive predictive value = a / (a + b) (equation 3-3) The situation of false positive results is highly dependent on the disease. In some cases a false positive diagnosis can have grave impact on a patient such as a screening test for STDs when a false positive would not only traumatize the patient but have the

PAGE 51

35 potential to wreck a marriage. A false positive robs the patient of their perceived health, and diverts precious resources for the unnecessary treatment of healthy individuals highlighting the need for sensible application of any screening test within a suitable environment. A further requirement for the sensible application of a screening test is that an earlier diagnosis improves the patients prognosis again a condition that is met with the target group and condition, where early detection can lead to a simpler and more successful treatment. 3.2 Proposed Testing Procedure 3.2.1 Components The brick wall structure in figure 3-2 represents epithelial cells, with normal cells depicted as unfilled black rectangles and tumorous cells shown as containing cyan (pink) ovoids. Open circles = microspheres with dye A, no surface folate = controls Closed circles = microspheres with dye B, folic acid coupled = sample 3.2.2 Procedure Performed after a dental cleaning to limit confounds. Step 1: Equal portions of microspheres A and B are administered as a suspension in mouthwash and swished for 30 s to ensure proper distribution through the oral cavity. Step 2: The subject expels the mouthwash of step 1, and rinses with water several times to remove entrapped microspheres. Folate bound microspheres (B) should adhere to tumorous cells by physically binding to folate receptors expressed by the cells.

PAGE 52

36 = microspheres with dye A, no surface folate= microspheres with dye B, folate immobilizedStep 1) Add microspheres solutionnormal cellstumorous cellsT r u e p o s i t i v e T r u e n e g a t i v e Step 2) Remove unbound microspheres by wash Step 3) Remove bound microspheres with acidwash and capture magnetically magnet magnet acid washStep 4) Analyze ratio of A to B microspheres = 5= 1 Ratio:vs.positivenegative Figure 3-2: Steps of the proposed testing strategy

PAGE 53

37 Step 3: An acid wash releases the receptor ligand binding, and a rinse recaptures the microspheres. These can now be collected magnetically, and assayed spectroscopically to determine the ratio of binding microspheres (B) to controls (A). Since some portion of microspheres will always be retained in the oral cavity by entrapment, the control particles are essential to prevent false positives. Entrapment should retain equal amounts of A and B, and any excess of B will be due to receptor-mediated binding. 3.3 Design Parameters for Microspheres Microspheres synthesized for the application described above require certain properties to accomplish the task. These properties, and the manner in which they can be designed into the microspheres were optimized for the particular application described. The design approach and the chemistry involved are very broadly applicable however, with only minor modifications to the immobilized ligand and microsphere optimization required to adapt the system to other applications. 3.3.1 Microsphere Size Considerations The ideal size of microspheres for this application was initially unknown. Known parameters that helped determine the boundary conditions are shown in the diagram of figure 3-3. The lower boundary is based on endocytosis. Folate receptors exist to conduct folic acid into the cell, and have been shown to endocytose drug molecules (i.e. proteins) and small (100 nm) particles (Wang and Low 1998). The upper size limit of particles that are endocytosed represents the smallest particle that could be recovered thus the smallest particle size that could be of interest to the application. The lower bound on

PAGE 54

38 microsphere size was postulated as being on in the single m rangeroughly from 2 5 m. Specific surface are a log size in um0.1 1.0 10 100 10001 10 100 1000 Surface area Minimum sizeto prevent endocytosis ~ 2-5 um Figure 3-3: Size considerations for microspheres The upper bound on size has several components. First, as the microspheres increase in size the specific surface area decreasesyielding less area to immobilize the binding agent on. Second, larger particles will be increasingly susceptible to draginduced shear stresses due to fluid movement near the surface. Both these phenomena dictate that larger particles become increasingly difficult to retain via physical binding to a surface. The upper boundary of microsphere sizes that could be retained by receptor mediated binding is postulated as 80 100 m. The size considerations represented a theoretical starting point for devising a suitable polymerization process for microsphere production. It was postulated that if a broad size distribution of microspheres in the above size range were produced, then the microspheres in the correct size range would be the same population that adhered to the cells of interest. This empirical solution made a suspension polymerization process the

PAGE 55

39 obvious choice, due to the broad distribution of particle sizes within the range of single to ~100 m that can be achieved. 3.3.2 Ligand Immobilization The microspheres must be capable of having folic acid immobilized onto their surface, with retention of receptor recognition. For folic acid this requires coupling via a carboxylic acid residue on the glutamic acid end. To enable the required coupling chemistry, a monomer with a terminal amine functional group that can couple to the carboxylic acid group is polymerized into the microspheres. 3.3.3 Magnetic Guidance The application requires sufficient magnetic loading to permit magnetic recapture of the microspheres for analysis. The nature of the magnetically responsive species must be superparamagnetic to avoid agglomeration of the microspheres. Viable approaches to forming magnetically responsive microspheres include doping monomers with magnetite prior to polymerization, polymerization in the presence of a ferrofluid, or in situ precipitation of magnetite onto microspheres. For this application magnetite was doped into the monomers prior to suspension polymerization, and the microspheres were collected magnetically ensuring the exclusive retention of magnetically responsive microspheres. 3.3.4 Dye labeling This application requires the preparation of microspheres labeled with two readily distinguishable dyes. The usefulness of the control particles would be diminished if the dyes were not incorporated in the same manner, or if the dye identity influenced the surface chemistry of the particles. Two approaches to dye labeling were investigated: microsphere swelling and dye loading, and conjugation of dye to monomer prior to

PAGE 56

40 microsphere polymerization. The dyes were chosen based on spectral separation and fluorescence intensity, so that they could be both detectable and distinguishable.

PAGE 57

CHAPTER 4 MATERIALS AND METHODS 4.1 Magnetic Material for Microspheres The microspheres require a magnetic component to enable magnetic recapture and separation. The magnetic component had to be superparamagnetic magnetite in order to maximize the magnetic response and to avoid agglomeration of particles. Superparamagnetic iron oxide was produced and coated with oleate and polydimethyl siloxane (PDMS) to facilitate dispersion in the main monomer: methyl methacrylate (MMA). This doped monomer was used in the suspension polymerization process in exactly the same manner as pure monomer would have been. 4.1.1 Materials used in Iron Oxide Preparation Oleic acid (Aldrich 36,425-5, [112-80-1]), polydimethyl siloxane (PDMS, 10 cs, Hls America PS039, [63148-62-9]), ammonium hydroxide (25 % aqu solution, Acros 255210025, [1336-21-6]), cyclohexane (Aldrich 15,474-1, [110-82-7]), hydrochloric acid (Acros 12463-0010, [7647-01-0]), chloroform (Fisher C297-4, [67-66-3]), ferric (iron II) chloride tetrahydrate (Aldrich 22,029-9, [13478-10-9]) and ferrous (iron III) chloride hexahydrate (Aldrich 20,792-6, [10025-77-1]) were all used as received. Methyl methacrylate monomer (MMA, Aldrich M55909, [80-62-6]) was distilled at reduced pressure to remove inhibitor. 4.1.2 Magnetite Production and Treatment All magnetite used was produced in our own laboratory, in a process that has been described extensively by Leamy (2003) and was originally adapted from methods of 41

PAGE 58

42 Robineau and Zins (1995). Briefly, stable suspensions of iron oxide were formed by the base catalyzed precipitation of iron chlorides. Ferrous and ferric chlorides were dissolved in water and hydrochloric acid (HCl), and added to stirring ammonia solution to induce precipitation of iron oxides. Iron oxide product was retained magnetically while most of the aqueous supernatant was decanted. Oleic acid (10 wt % relative to the oxide) in cyclohexane was added to the magnetic slurry to coat the iron oxide and render it hydrophobic. Addition of methyl alcohol permits a solvent exchange to cyclohexane as iron oxide falls to the bottom and the aqueous/alcohol phase can be decanted. The iron oxide was air dried overnight and suspended in chloroform. PDMS (10 wt % relative to the oxide) was added at this stage. The solvent was again evaporated off by air drying to recover the solid product, which was dispersed in monomer. 4.1.2.1 Method of iron oxide precipitation An aqueous solution of iron oxides was produced by combining 2.03 g ferrous iron chloride tetrahydrate (FeCl24H2O) with 4.88 g ferric iron chloride hexahydrate (FeCl36H2O) and 0.887 ml 37 % aqueous HCl in 20 ml distilled water. A 250 ml beaker was filled with 8.3 ml aqueous 28 % NH4OH in 155 ml of DI water. This solution was mechanically mixed for 5 min by a 4-blade stainless steel stirrer (PT1035 stirrer, Kinematica GmBH, Germany), set to approximately 400 rpm. The iron oxide solution was poured into the 250 ml beaker and left stirring for 10 min., after which stirring was stopped and a 1 inch square rare earth magnet (NeFeB) under the bottom of the beaker was used to collect the magnetite produced. The supernatant was decanted, leaving approx 60 ml magnetite solution, which was stirred at low speed (approx. 200 rpm).

PAGE 59

43 4.1.2.2 Method of coating iron oxide The minimum effective oleic acid concentration had been determined previously in cooperation with Leamy (2003), as 10 w/w %. 0.221 g oleate dissolved in 11.14 ml cyclohexane was added to the slurry of iron oxide under stirring. The role of oleate is to coat the iron oxide particles resident in the aqueous phase and render them hydrophobic, thus forcing segregation into the cyclohexane phase. 15 ml of methanol was added after stirring for 15 min, reducing the density of the aqueous phase to permit the ferrofluid in cyclohexane to segregate to the bottom of the beaker more easily. The cyclohexane ferrofluid was retained in the beaker through placement of the same rare earth magnet as above, under the beaker, while the aqueous supernatant was decanted. 10 w/w % (relative to the iron oxide content) polydimethylsiloxane (PDMS) was added directly into the cyclohexane ferrofluid slurry under agitation. Air-drying overnight in a petri dish evaporated off the cyclohexane. The coated magnetite was dispersed in methyl methacrylate (MMA) monomer (see structure in figure 4-1) by sonicating (Branson 2510) and vortexing (Fisher Vortex Genie 2) repeatedly, at a concentration of 6 w/v % (of iron oxide, not including weights of coatings). This product was stored in the refrigerator until further use. Figure 4-1: Methyl methacrylate (MMA) structure

PAGE 60

44 4.1.3 Characterization of Iron Oxide The iron oxide produced was based on processes well characterized both in literature and within our own labs (Leamy 2003). The magnetic properties were established by Leamy using superparamagnetic quantum interference device (SQUID) magnetometry. The physical size of the particles of iron oxide produced was characterized by transmission electron microscopy (TEM) and by scanning electron microscopy (SEM). The magnetic response was also anecdotally verifiable, as any permanent magnet could strongly attract the microspheres. 4.2 Microsphere Polymerization and Characterization The microspheres were prepared by a suspension polymerization process, using doped monomers. The monomers were distilled as appropriate and processed to include dopant species. The main backbone monomer, methyl methacrylate (MMA), had magnetite dispersed in it as described above at a concentration of 6 w/v %. The functional comonomer, aminoethyl methacrylate hydrochloride salt (AEMHS) was conjugated to fluorescent species as described below in certain polymerizations. In some instances, dye loading was performed by means of solvent swelling, in which case dye loading occurred after polymerization and washing steps. 4.2.1 Incorporation of Fluorescent Dye Into Functional Monomer Two separate strategies were applied to incorporate fluorescent dyes into the microspheres. The strategy described here involved the covalent coupling of dyes fluorescein isothiocyanate (FITC, structure shown in figure 4-2) or Texas Red-X (structure shown in figure 4-3) to the functional copolymer utilized in the polymerization reactionaminoethyl methacrylate hydrochloride salt (AEMHS, structure shown in figure 4-4).

PAGE 61

45 Figure 4-2: Fluorescein Isothiocyanate (FITC) structure Figure 4-3: Texas Red-X (TR) structure 4.2.1.1 Materials for fluorescent dye incorporation into functional monomer Fluorescein isothiocyanate (Molecular Probes F-143, Eugene OR), Texas Red-x succinimidyl ester (Molecular Probes T6134, Eugene OR), absolute ethanol and 2 aminoethyl methacrylate hydrochloride salt (AEMH, Acros 357810250, [2420-94-2]) were all used as received. 4.2.1.2 Method of conjugating fluorescent dye to functional monomer The dyemonomer conjugate was prepared by magnetically stirring (overnight) a mixture of dye (FITC or Texas Red-X succinimidyl ester) and AEMH in solvent (absolute ethanol for FITC and THF for Texas Red-X). The dye was conjugated to at

PAGE 62

46 least 3.5x molar excess of monomer. The fluorophores were delivered with readily reactive leaving groups so that conjugation to monomer was simple. Figure 4-4: Aminoethyl methacrylate hydrochloride salt (AEMHS) structure The reaction of FITC with AEMH is illustrated schematically in figure 4-5. The isothiocyanate group reacts with the primary amine on the monomer. The reaction occurs by attack of the nucleophile on the electrophilic central carbon of the isothiocyanate group. This forms a thiourea linkage between the FITC and AEMH with no leaving group (Hermanson 1996). AEMHNH2 SCNFITC monomer with primary amineisothiocyanate fluorophoreNCNFITCSAEMHHH Isothiourea bond Figure 4-5: Mechanism of FITC conjugation to AEMH monomer The reaction of Texas Red-X with AEMH is shown schematically in figure 4-6. The succinimidyl ester of the Texas Red reacts with the primary amine on the monomer to form a stable amide bond, and the N-hydroxysuccinimide forms the leaving group (Hermanson 1996).

PAGE 63

47 monomer with primary amineNCTexas RedOAEMHH Amide bondNOOOCTexas RedO AEMHNH2 succinimidyl esterfluorophore NOOOH NHS leaving group Figure 4-6: Mechanism of Texas Red-X conjugation to AEMH monomer Experiments were performed to optimize the fluorescent dye concentration (using FITC) by conjugating different amounts of dye to 100 mg AEMH. For regular production of microspheres the dye was conjugated to either 20 or 100 mg AEMH. The stirring mixture was protected from light to avoid photobleaching of the fluorophores. The ethanol was removed either by rotovapor, or by flash freezing the monomer-dye conjugate in liquid nitrogen and then drawing the solvent off at reduced pressure and temperature. The monomerdye conjugate was dispersed in 980 or 900 mg AEMH, to bring the total amount of monomer to 1000 mg, and mixed at ~35o C to ensure that mixing occurred intimately in the liquid phase. The AEMH monomer with dispersed dye conjugate was used together with the magnetite doped MMA monomer in the suspension polymerization recipe.

PAGE 64

48 4.2.2 Suspension Polymerization Procedure The suspension polymerization procedure was adapted from Shim and Kim who devised a simple but effective suspension polymerization protocol to produce methyl methacrylate particles incorporating inorganic oxides for sunscreen applications(Kim et al. 2002; Shim et al. 2002). 4.2.2.1 Materials for suspension polymerization Polyvinyl alcohol (PVA, 87-89 % hydrolyzed, 85 146,000 MW, Hoechst-Celanese) was used as delivered. The originally applied Airvol 523 polymer was supplanted (by the producer, who indicates that the new Celvol product is precisely the same as the old Airvol equivalent) by CelVol 523 and CelVol 823the latter being a new easier-dissolving grade that was adapted for use in all polymerizations subsequent to lot S19. Monomers were used as prepared above. Crosslinker divinyl benzene (DVB, Sigma D-0916 lot 50K3652, [1321-74-0]) was NaOH washed three times. Initiator Azo-bis isobutyronitrile (AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1]) was recrystallized in methanol. Additional monomers aminoethyl methacrylate (2-aminethyl methacrylate hydrochloride, AEMH or AEMHS, Acros AC357810250, [2420-94-2]), aminostyrene (AmST, Acros AC30848, [1520-21-4]) and Maleic anhydride (MA, Dajac 7579, [108-31-6]) were used without further purification. 4.2.2.2 Method of suspension polymerization The suspension medium was 100 ml of 2 w/v % aqueous polyvinyl alcohol (PVA) solution, prepared by dissolving celvol 823 PVA (89 % hydrolyzed, 85-146,000 MW) in distilled water and stirring mechanically (Caframo stirrer RZR1, Ontario Canada) at moderate speed overnight at 70o C. This mixture was prepared in a 3 necked 300 ml round bottom flask, with a Teflon stirring rod admitted through the central neck, and a

PAGE 65

49 thermometer fitted for temperature monitoring, and the third neck capped with a rubber stopper when not in use for reactant addition. Heating was provided by a heating mantle controlled by a Variac power adjuster. The monomer mixture was prepared by combining the monomers to be polymerized in a 20 ml glass vial. In most cases 80 90 v/v % main monomeralways methyl methacrylate, was used with 10 20 w/v % functional monomer (maleic anhydride, aminostyrene or aminoethyl methacrylate). An additional 2.5 v/v% divinyl benzene (DVB) was added as crosslinking agent, as well as 2.5 w/v % azobisisobutyronitrile (AIBN) as initiator. These ingredients were combined at mildly elevated temperature (~40o C) to prevent coagulation of AEMH, and mixed until homogeneously dispersed. The suspension medium temperature was elevated to 75o C, and the stirrer was switched for a homogenizer (Kinematica PT1035, Germany) in later suspension polymerizations, with speed adjusted so that mild or no frothing of the mixture occurreda setting of 4 on the controller. The erratic electric supply of the building required that an uninterruptible power supply with automatic voltage regulation (APC XL series 1500 with voltage boost and trim functions to limit variation < 5 % in voltage output) was installed to prevent the homogenizer from either stopping due to low voltage or frothing the mixture excessively due to overvoltage. The monomer mixture was injected through a large gauge stainless steel needle affixed to a syringe. After injection the suspension mixture was left stirring for 4-8 h at 75o C, after which time the heating mantle was switched off. The mixture was left stirring until it had cooled to near room temperature, then decanted into 50 ml centrifuge vials for washing and storage. The polymerization apparatus is shown in figure 4-7 below.

PAGE 66

50 4.2.3 Microsphere Post-Polymerization Processing The microspheres were recovered by magnetic separation and three washes with distilled water. The washed microspheres were stored in solution, away from light (for those microspheres that had either dye content or had folic acid immobilized onto them). Magnetic separation was performed using custom produced magnetic separators (see figure 4-8), which were largely the work of Leamy (2003). These permitted efficient separation of the microspheres from medium at high speed (seconds to minutes). Microspheres that were not covalently dye labeled were ready for swelling with dye at this stage. Figure 4-7: Suspension polymerization setup using mechanical stirrer and heating mantle. Panel a shows the Caframo stirrer, with teflon paddles. Panel b shows the Kinematica homogenizer. Some microsphere batches were prepared with maleic anhydride, which polymerizes across its inner bond (with free radical initiation). This ring is readily

PAGE 67

51 reactive towards primary amines, forming an amide bond and a carboxylate group on the other opened site. This carboxylate group can in turn react with another primary amine in a condensation reaction to form a second amide bond. The method of performing the maleic anhydride modification was to add excess ethylene diamine (approx 4 ml for <1 g microspheres) to microspheres from which water has been removed by magnetic separation (excess water was removed, but no further drying was done), stir for 15 min, and wash repeatedly with DI water. Figure 4-8: Magnetic separator apparatus. The steel cylinder contains four magnets held in place by a machined insert. Magnets are magnetized through their width: opposite faces have like charges and adjacent faces have opposite charges. 4.2.4 Microsphere Dye Loading by Swelling and Solvent Evaporation Solvent was chosen by assessing the microsphere swelling (in solvent), and dye solubility (in solvent). Microsphere swelling tests were performed using a Zeiss Axioplan 2 imaging microscope. Microscope slides were affixed with double sided tape, and dried microspheres were scattered at low concentration onto the tape. Analysis was performed using a 10x objective, with 1300 x 1030 pixels used. The camera was operated in time lapse mode at one frame taken every 12 seconds for 10 20 minutes. The time was

PAGE 68

52 determined by a test sequence for ascertaining the time to ultimate degree of swelling. The images were then sized using Zeiss Axiovision 3.1 software with scaling module. The solubility of dye in solvent was assessed by adding 1 w/v % dye into solvent and vortexing (30 s) and sonicating (5 min), repeating the process twice. Evaluation was based on visual observations of precipitates / undissolved fractions of dye remaining in solvent after mixing. Triple washed microspheres were resuspended in dye/solvent solution by vortexing (30 60 s) and sonicating (10 min) several times. The microspheres were left in the dye solution overnight. Microspheres were collected magnetically, and supernatant decanted, after which they were dried in a vacuum oven overnight at low temperature (~40o C). Dried particles were water washed repeatedly until no further dye was visible in the supernatant after washing. The microspheres were then washed 3 more times. 4.2.5 Microsphere Characterization The microspheres produced were universally characterized for size, since the suspension polymerization process has a significant potential for size variation. Several standard techniques were applied. All batches produced were sized using a Coulter light scattering particle size analyzer, which provides a size histogram of the particles tested. Most batches of particles were also sized by light microscopy, and in some cases electron microscopy was used, providing additional shape and surface data, including surface morphology. The microspheres were also characterized for fluorescence response when doped with fluorophores. A set of experiments to provide some optimization data for fluorophores concentration was also conducted. Evaluation was provided by fluorescence spectroscopy.

PAGE 69

53 4.2.5.1 Coulter sizing Size histograms were generated using Coulter instruments: initially a Coulter LS 230, and then its successor machine, a Coulter LS 13 320. Both were equipped with small volume fluid modules, where the particles were sized suspended in water. The procedure for testing involved selecting (or building) the appropriate optical modela collection of physical constants including the real and imaginary components of the index of refraction for the material composing the particles, and the index of refraction of the medium (water in this case). After acquiring a background measurement, the particles to be sized were loaded, and measured in three consecutive discrete runs. The newer instrument provided an automatically calculated average for all runs performed. All tests were done with stirring at the minimum rate (30 % for the LS 230, 20 % for the LS 13 320), which prevented the particles from settling during the 90 s runs. In the event of inconsistent results from consecutive runs the measurements were repeated. The optical model built for the methyl methacrylate based magnetite doped particles was constructed to use a real index of refraction of 1.5, and an imaginary index of refraction of 5. These values were determined from the Coulter manual description of indices for materials with similar properties when no data for the exact material was available. The variation in results with the indices selected was projected as minimal, since the impact of the indices on the sizing calculations decreases as particle size increases over 1 m. 4.2.5.2 Light microscopy For light microscopy particles were typically suspended in ethanol solution, to enable quicker drying times onto a cover slide. A very dilute suspension of particles was

PAGE 70

54 placed on a slide, which was then viewed under the microscope. All microspheres were viewed in transmitted light mode. The microspheres with fluorophores could also be usefully imaged by fluorescence microscopy. In all cases images generated for characterization were acquired using a Zeiss Axioplan 2 imaging microscope and the attendant Zeiss Axiovision 3.1.2 software. 4.2.5.3 Zeta potential analysis A Brookhaven Zeta Plus zeta potential analyzer and particle sizer was used to acquire information on the surface charge of the microspheres, both before and after the immobilization procedure. Microspheres were dilutely suspended in pH 7.4 PBS. 4.2.5.4 Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) Samples were prepared for electron microscopy by drying (either in vacuum oven or by freeze dryer), and sputter coating for 1 min with carbon. Images were then acquired with either a JEOL 6400 scanning electron microscope or a JEOL 6335CF field emission scanning electron microscope. Both microscopes also featured EDS attachments with LINK ISIS software capable of generating spectra on constituents present and semi-quantitative analyses. 4.2.5.5 Inductively coupled plasma spectroscopy (ICP) A Perkin Elmer Plasma 3200 IRL inductively coupled plasma spectroscopy (ICP) system was used to determine the concentration of iron in the microspheres. This was interesting because the iron oxide concentration is a primary factor responsible for the strength of the magnetic response. Samples were prepared by dissolving the microspheres to atomic constituents. Dissolution occurred either in a solution of hydrofluoric acid (HF) and sulfuric acid

PAGE 71

55 (H2SO4), or in an acid mixture sometimes called aqua regiaconsisting of three parts hydrochloric acid and one part nitric acid. The samples were compared to Fe standards to establish an iron concentration in the sample. 4.2.5.6 X-Ray powder diffraction (XRD) XRD was performed using a Philips APD 3720 in MAIC. Samples were prepared by washing microspheres and drying in a vacuum oven. The resulting diffraction patterns were analyzed by comparison to known patterns for possible compounds present, as referenced from the internal database. 4.2.6 Microsphere Fluorescence Properties Optimization was carried out on a series of suspension polymerized batches of microspheres composed of the standard 80/20 mix of methyl methacrylate and aminoethyl methacrylate monomers. 2.5 wt % initiator and 2.5 vol % crosslinker were added, also the standard amounts. A table in the results section (table 5-6) shows the amount of fluorescein isothiocyanate added to each sample, as well as the molar ratio of fluorophores to monomers. The fluorophores to monomer ratio was calculated by applying the assumption that all fluorophores and monomer in the feedstock was successfully integrated into the microspheres. This assumption is not valid, but it does provide a consistent error. The absolute numbers are probably off by some tens of percent, but the ratios between the samples should have the same error to them so that the comparison between samples remains valid. In addition to assessing the fluorescence generated by each dye concentration, the polymerized samples were used to assess the effectiveness of the dye incorporation process by leaching out dye through a DMSO solvent wash. The DMSO supernatant was

PAGE 72

56 examined for dye content and for the nature of the dye species that had been leached out of the microspheres. Fluorescence intensity was examined for both microspheres in aqueous solution, and for DMSO washed microspheres. UV-absorbance spectroscopy was used to determine the concentration of dye in DMSO. Fluorescence spectroscopy was used to determine the identity of the dye molecules: whether they were native fluorescein isothiocyanate or if they were conjugated to monomer or oligomers. 4.2.6.1 Sample preparation Four samples of FITC-AEMHS conjugate were prepared. Each amount x of FITC was conjugated to 100 mg of AEMHS in 2 ml of absolute ethanol, and left stirring for 48 h, while protected from light exposure. These solutions were made with 2, 5, 10 and 20 mg of FITC. The solutions were flash frozen in liquid nitrogen, and placed under vacuum within an ice bath for 8 h to remove the solvent. The monomer-dye conjugate that remained had 900 mg AEMHS monomer added to it under gentle heatingapproximately 40o C. The dye-conjugated monomer was dispersed within the liquid pure monomer by both mechanical stirring and sonication. Suspension polymerization was carried out as for other samples (this series of samples was labeled S21 through S24, in order of increasing FITC concentration). The polymerized samples S21 through S24 were collected into two 50 ml centrifuge vials each. Each vial was washed three times with 10 ml distilled water, with the particles being magnetically separated in magnetic separators previously constructed in our labs (Leamy 2003). The particles were magnetically retained and the supernatant was decanted, after which the particles were redispersed in new media by vortexing (1

PAGE 73

57 min) and sonication (10 min). 5 ml of each triple washed samples was withdrawn by pipette the water decanted. The samples were dispersed in 5 ml DMSO by vortexing and sonicating as above, then magnetically separated and the supernatant was collected. This process was repeated for one wash after the DMSO supernatant was clear to the naked eye of any dye. All supernatant was saved for analysis. 4.2.6.2 UV-Visible absorbance spectroscopy UV-Visible absorbance spectrometry was performed on a Shimadzu 2401PC UV-Visible recording spectrophotometer, with 6-cell attachment. Experiments for characterization were conducted using spectrum mode, while concentration experiments were done in photometric mode. PMMA cuvettes were used for aqueous media, while Polystyrene cuvettes were used for DMSO based organic solutions. 4.2.6.3 Fluorescence spectrometry Fluorescence spectrometry experiments were performed on a Fluorolog Tau-3 (Jobin Yvon Spex Instruments, S.A. Inc) with a 450 W xenon lamp as the excitation source. For characterization of FITC fluorescence, emission was monitored at the 515 nm emission peak. Excitation was provided at the 488 nm excitation peak. When particle based scattering or raman scattering caused increased in noise beyond the discernible signal, it was possible to utilize lower excitation bands of fluorescein at 350 nm to generate a signal that was lower in intensity, but improved with regard to signal/noise ratio. 4.2.6.4 Confocal microscopy Confocal microscopy experiments utilized an Olympus IX81 microscope with Fluoroview software to assess dye distribution within the microspheres. Images were acquired using a default 20x objective, as well as a 60x oil-immersion objective (PLAPO

PAGE 74

58 60XO). In each case optical and fluorescence images were acquired concurrently, with excitation provided by a 488nm argon laser for fluorescein labeled samples, and a red laser for Texas Red labeled samples. 4.2.6.5 Fluorescence microscopy Fluorescence microscopy was performed using a Zeiss Axioplan 2 imaging microscope with Axiovision 3.1 software. For fluorescence images, or combined (dual) mode images appropriate filter cubes were utilized. When fluorescein was the fluorophore, filter set 10 was used, with the spectral characteristics shown in figure 4-9. This filter set is expressly sold for use with fluorescein. Figure 4-9: Spectra for Zeiss filter set 10, used for fluorescence microscopy of fluorescein containing particles. Sourced from Carl Zeiss, Inc. (2002), Upright microscopes technical data, http://www.zeiss.de/4125681F004CA025/Contents-Frame/286BA4D22B14DEE985256B4A007C3686, January 2004

PAGE 75

59 4.2.7 Preparation of Microspheres for Cell Work All microspheres that were used for cell work were repeatedly washed and suspended in distilled water to remove any possible impurities. No further preparation was made in cases when the cells would be disposed of directly after testing. If the cells were to be used further the microspheres were sterilized by exposure to UV light for several hours under a biological hood. A sample of UV-exposed microspheres was compared to non UV-exposed controls by fluorometric methods to determine if the sterilization exposure generated photobleaching meaningful enough to provide significant reduction in fluorescence intensity. 4.2.8 Preparation of Microspheres by Dispersion Polymerization Microspheres were prepared by dispersion polymerization using only a single general method. This method was partly developed by Leckey (1997). 4.2.8.1 Method of dispersion polymerization The solvent phase was prepared from a mixture of absolute ethanol and water with stabilizer mixed in. The stabilizer used was polyvinylpyrrolidone, MW 40,000 (PVP40), utilized at a concentration equal to 12 w/v % relative to the monomers mixture amount added to the solvent phase. This mixture was stirred for at least 5 min to permit good mixing. The amount of water in the dispersion media was a primary lever that was utilized to control the size of the particles formeda typical composition was 80 v/v % ethanol with 20 v/v % water. The monomer phase was prepared by adding together the desired combination of monomersin most cases approximately (measured by volume) 2 parts styrene and 1 part diethyl aminoethyl methacrylate (DEA). Divinyl benzene (DVB) was added as crosslinker, usually around 1 v/v %. The initiator used was azo-bis isobutyronitrile

PAGE 76

60 (AIBN), at a concentration equal to 3 w/v %, and stirred for at least 2 min to fully dissolve. The monomer mixture was combined with the dispersion phase at 5 v/v % monomers and 95 v/v % dispersion phase, to make 5 or 10 ml samples that were sealed into glass vials of at least twice the capacity of their filling. The combination was vortexed (30 s) and purged with nitrogen gas before sealing. The sealed glass vials were placed in a pre-warmed shaking water batch (Haake SWB 25), with n = 100 cycles per minute, and temperature set between 55 and 80o C, with the most common temperature used being 70o C. The vials were zip tied diagonal to the direction of reciprocation, fully submerged in water. In almost all cases the polymerization time was 24 h ( ~ 4h). Samples were removed from the polymerization apparatus and cooled. A crust of polymer sometimes formed under the cap, and any agglomerations or crust formations were disposed of. The microspheres formed were washed in DI water by centrifuging and redispersing in water at least 2 x. All microspheres were sized by Coulter light scattering. 4.2.8.2 Incorporation of magnetic species Several different techniques of imparting magnetic response to the microspheres were applied. These can be divided into processes that involved addition during polymerization, and post-polymerization modifications. When magnetic additions were made prior to polymerization, a portion of the dispersion medium was replaced with a suspension of magnetite in water; in some cases the magnetite was silica coated.

PAGE 77

61 Post polymerization addition of magnetite occurred by base catalyzed precipitation of iron oxide in situ. The base catalyzed precipitation of iron chlorides to form iron oxide particles was discussed in detail by Leamy (2003). 4.2.9 Preparation of Microspheres by Activated Swelling Activated swelling procedures were adapted from literature sources that detailed procedures that were separately developed by groups headed by Tuncel and Frchet, based on original development of the process by Ugelstad (Ugelstad 1978; Ugelstad et al. 1979; Ugelstad et al. 1988; Wang et al. 1994; Tuncel et al. 2002a; Tuncel et al. 2002b). All procedures used seed particles that could be prepared by emulsion polymerization, or as in the present research by dispersion polymerization. In almost all cases the seeds were low polydispersity uncrosslinked styrene microspheres, and the swelling agent used was dibutyl phthalate (DBP). The procedure is essentially a suspension polymerization of preformed swollen seed particles, with droplet size controlled by swelling rather than by stirring rate. This is performed in a medium designed to preclude particle coalescence. The protocols from the Frchet group (Wang et al. 1992; Wang et al. 1994) used a 4 to 10-fold excess (v/w) of swelling agent over seed particles. The swelling suspension was prepared by dispersing 60 mg seeds with at least 200 l DBP in 30 ml 0.25 w/v % aqueous sodium dodecyl sulfate (SDS) solution by sonicating (using a cell disruptor probe on 50 % intensity for 30 s). The suspension of swollen seeds was stirred for at least one whole day. The above solution had added to it 0.03 w/v % sodium nitrate as solution phase inhibitor and approximately 5 ml water with enough PVP40 stabilizer to adjust the total solution to 1 w/v % stabilizer concentration, and stirred for 10 min. A

PAGE 78

62 monomer mixture of 4.4 ml styrene with 2.5 w/v % benzoyl peroxide (BPO) initiator was added, and emulsified in the solution via probe sonication for 20 s. The whole mixture was then polymerized in a sealed glass vessel at 70o C for 24 h. The product was used in the next step without purification. Styrene and DVB were mixed 1:1 to make 10 ml monomers, and had 1 w/v % BPO added as initiator. The monomer mixture was emulsified in 40 ml of 0.25 w/v % aqueous SDS solution as previously, by probe sonication for 20 s. The emulsified monomer mixture was stirred with the polymerized swollen seeds for 5 h at room temperature. Polymerization took place in sealed glass vessels for 24 h at 70o C. The product was washed and purified like any other microsphere preparation. The Tuncel protocols that were applied to prepare batches of microspheres were drawn from several publication by Tuncel et al. (Tuncel 1999; Tuncel et al. 1999; Tuncel and Cicek 2000; Tuncel et al. 2000; Tuncel et al. 2002a; Tuncel et al. 2002b). The essential protocol was similar to the above Frchet protocol, with a primary difference being that the swelling agent (Dibutylphthalate, DBP) was used at a 1:1 ratio (v/w) with the seed particles. The typical Tuncel-based protocol used 160 mg of uncrosslinked styrene seeds, swollen with 160 l DBP that was emulsified in 20 ml of aqueous 0.25 w/v % SDS solution, and stirred together for 24 h. A monomer emulsion was prepared that contained equal parts of styrene, hydroxyethyl methacrylate (HEMA) and DVB to make 1.2 ml monomers, mixed with 60 mg benzoyl peroxide initiator until dissolved, and emulsified in 20 ml aqueous 0.25 w/v % SDS solution. The monomer emulsion was stirred together with the swollen seeds emulsion for 24 h at ~300 rpm and room temperature. Sufficient 10 w/v % stabilizer (either PVA or PVP40) was added to bring

PAGE 79

63 the concentration of stabilizer in mixture to 1 w/v %. The preparation was nitrogen purged before polymerization, then sealed in glass vials and polymerized for 24 h at 70o C at 120 cpm in a shaking water bath, with the vials arranged diagonally to the direction of motion. The product was then washed and purified as for other polymerizations. 4.3 Immobilization of Folic Acid onto Microspheres 4.3.1 Folic Acid Immobilization Procedure Folic acid (figure 4-10) must be bound via the carboxylic acid groups that exist on the glutamic acid end of the molecule rather than via the amine group on the opposite end (at the methylpteridin). This has been proven necessary in order to preserve receptor binding functionality, as binding to the receptor occurs across the amine group. The procedure has been adapted from protocols by Zhang et al. (2002), who have shown success binding folic acid onto particles with surface chemistries similar to the microspheres being used. Figure 4-10: Folic acid structure 4.3.1.1 Materials for folic acid immobilization Dimethyl sulfoxide (DMSO, Acros 12779 [67-68-5]), N-hydroxy succinimide (NHS, Aldrich 13,067-2 [6066-82-6]), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC, Aldrich 16,142-2, [25952-53-8]), folic acid (Fisher Bioreagents BP251925, [59-30-8]), Ethylene triamine (triethylamine / Et3N, Aldrich 13,206-3, [121-44-2]), were all used as delivered from the respective suppliers.

PAGE 80

64 4.3.1.2 Method of folic acid immobilization onto microspheres Folic acid (COOH)NH2 Et3NEDC, NHS NH(CO)Folic Acid Figure 4-11: Schematic of folic acid immobilization onto microspheres Folic acid immobilization proceeded along the plan shown in figure 4-11. The chemistry utilized is depicted schematically in figure 4-12. The carbodiimide compound EDC is used as a zero-length crosslinking agent, mediating the formation of an amide between the carboxyl group on folic acid (at the glutamic acid end) and the terminal amine on the microsphere (from the aminoethyl methacrylates comonomer). Ethylene triamine was used to activate the amine grouprendering it more reactive. The carbodiimide modifies the carboxyl group to form a short-lived O-isoacylurea intermediate, which is highly reactive towards nucleophilic amine groups, to form the strong amide linkage and an isourea by-product. NHS is added to the reactants to increase the stability of the active intermediate (Hermanson 1996). The microspheres were washed repeatedly prior to immobilization, and 20 mg of microspheres were suspended in 2 ml DMSO. 20 l Et3N was added to activate the microsphere surface amine groups, and the suspension was mixed (sonicated and vortexed for 10 min). A second solution was made consisting of 1 ml 10 M folic acid in DMSO with 1.5 ml each of aqueous 15 mM NHS and 75 mM EDC. This second solution was mixed (sonicated and vortexed) for 10 min. Both solutions were combined, and mixed for 20 min. The microspheres were subsequently separated from the solution magnetically, and washed in distilled water 8x, and finally stored in 5 ml distilled water, away from light.

PAGE 81

65 +NCNNCH3CH3HCH3 R'OHO CarboxylatecompoundEDCforms O-isoacylureaintermediateCH3NCNOCR'ONHCH3CH3H RNH2 amidecompound CH3NCNONH+CH3CH3HH R'ONHR amide bonded productisourea by product Figure 4-12: Schematic detailing the carbodiimide mediated coupling of a carboxyl group to an amine to form an amide linkage. 4.3.2 Characterization of Folic Acid Immobilized Microspheres The folate-immobilized microspheres were characterized to demonstrate the attachment of folate onto the surface. The model system detailed in chapter 6 was utilized for this process also, particularly in cases where larger and more polydisperse microspheres generated significant light scattering to create extra noise in light spectroscopy systems. For spectroscopic methods, there are two independent fluorescent species contained in folic acid: p-aminobenzoic acid (PABA, the leftmost species in the folic acid

PAGE 82

66 structure shown in figure 4-8), and methylpteridin (MTE, the central portion of the folic acid structure). PABA has an absorption band at 265 nm and emits at 336 nm (Tanojo et al. 1997). MTE has absorption bands at 275 nm and 352 nm, and emits at 447 nm (Espinosa-Mansilla et al. 1998). The absorption spectrum of folic acid shows maxima at 281 nm and at 355 nm. 4.3.2.1 UV-Visible spectroscopy Absorption spectra for control and immobilized microspheres were compared to the absorption spectrum of aqueous folic acid. Experiments were performed in spectrum mode on dilute suspensions of microspheres in DI water. 4.3.2.2 Fluorospectrometry Experiments were performed on a dilute suspension of microspheres in DI water, using a Fluorolog Tau-3 (Jobin Yvon Spex Instruments, S.A. Inc) with a 450 W xenon lamp as the excitation source. 4.3.2.3 Brookhaven zeta plus The Brookhaven Zeta Plus zeta potential and particle size analyzer provided data on the charges of the microspheres before and after immobilization. Phosphate buffered saline (PBS), 7.4 pH was used as medium, and particle sizes from Coulter data were specified as parameters. Software used was BIC Zeta Potential Plus, set to 1 run of 10 cycles. Measurements were repeated at least 3 timesmore if results were inconclusive or questionable. 4-4 Cell Testing In order to evaluate the microspheres produced for specific binding to folate receptors, testing on cell lines was an important model system. This required cell lines that were known to express folate receptors (FR) as well as cell lines known to not

PAGE 83

67 express folate receptors in order to serve as controls. Cells were cultured according to standard techniques while observing any instructions specific to the individual cell lines, and grown on circular cover slips to provide a limited and controlled testing environment. 4.4.1 Cell Lines The cell lines used were primarily sourced from the American Type Culture Collection (ATCC), and much of the information on the cell characteristics was directly drawn from their database (www.atcc.org). Cells were also acquired from Clonetics (Clonetics was acquired by Bio Whittaker, and operates under Cambrex Life Sciences). Appendix B provides full information sheets on each cell line, as provided by the source company. 4.4.1.1 Malignant cell line CRL-5800 / NCI-H23 human epithelial lung adenocarcinoma The NCI-H23 cell line is a human epithelial cell, non-small cell lung adenocarcinoma originally derived from a 51 year old black male. Lung adenocarcinomas are known to overexpress FR (Weitman et al. 1992; Franklin et al. 1994), making it a good first model to utilize for testing. 4.4.1.2 Secondary testing cell line CCL-163 / BALB/3T3 clone A31 mouse fibroblasts The BALB/3T3 cell line is a mouse fibroblast that is known to be highly susceptible to transformation in tissue culture. It was utilized as a secondary testing cell line. 4.4.1.3 Control cell line NHDF: normal human adult fibroblasts Normal human adult fibroblasts from Clonetics were used as controls. There is no indication that these cells have any overexpression of folate receptors.

PAGE 84

68 4.4.1.4 Oral squamous cell carcinoma cell line SCC-9 Oral squamous cell carcinoma (OSCC) cells were utilized to provide a tissue-specific cell type for experiments. The cells were from the ATCC SCC-9 cell line, and were revived after extended cryo-storage (>10 years). 4.4.2 Cell Culture Procedures The cell lines were grown in RPMI 1640 1x medium without L-glutamine (Mediatech Inc., Herndon, VA). The medium was supplemented with 10 % fetal bovine serum purchased as manufactured for investigational use (Mediatech, Inc., Herndon, Virginia) and 1% Penicillin-Steptomycin 10,000 IU/ml and 10,000 g/ml purchased as sterile filtered for in vitro diagnostic use (Mediatech, Inc., Herndon, Virginia). Medium was changed every 2-3 days and cells were passaged weekly using 0.05% trypsin-EDTA (Mediatech, Inc., Herndon, Virginia). All cells were maintained in 6 multi-well plates, T-25 flasks and/or T-75 flasks with respectively, 3ml, 5ml and 13ml of growth medium while being kept at 37C in a humidified 95% air: 5% CO2 atmosphere. 4.4.3 Cell Culture Preparation and Testing Procedure 4.4.3.1 Cells seeded onto multi-well plates This procedure was utilized for NCI-H23 cells. Cells were grown in 6-well culture plates according to the procedures above and in appendix B. For testing, 200 l of each suspension (particles in water, ~ 7 w/v %) was pipetted into wells (added directly to medium), followed by 6 min gentle agitation of the plate, after which the wells were aspirated. Each well was rinsed by addition of Hanks balanced salt solution, and 60 s agitation to dislodge and non-bound particles. The rinse procedure was repeated 3 times. A fourth and final rinse with cell growth medium was performed. The medium was

PAGE 85

69 aspirated from the cell wells, and the plate was inverted for observation on a Zeiss Axioplan 2 imaging microscope with Axiovision 3.1.2 image acquisition software. 4.4.3.2 Cells seeded onto coverslips Cells were seeded onto ~ 6 mm diameter coverslip circles that were hole-punched out of Fisher 22 mm plastic coverslips (Fisher catalog # 12-547). The coverslips were placed at the bottom of cell wells in multi-well plates, and cell seeding took place as for wells. Fisher specified the coverslips as being cell-adherent, but specific materials data was unavailable. For testing, the coverslips were picked up with tweezers and immersed into wells of suspended microspheres and swirled for 60 s. The tweezers were cleaned with 70 % ethanol solution before each manipulation to ensure no transfer of microspheres occurred by tweezers, and for better hygiene. Approximately the same concentration suspension of microspheres was used, but this factor was not strictly controlled for. The rinse procedure involved swirling the coverslips in the same rinse solutions as above for 60 s. Rinsing was repeated 3 x with Hanks balanced salt solution and once with medium. In later experiments, the rinse utilized was changed to medium. For analysis the coverslips were placed onto microscope slides for support and examined with the Zeiss Axioplan 2 imaging microscope. 4.4.3.3 Cell experiments for microsphere specific binding A series of experiments with cells were conducted to determine the capacity of the folate immobilized microspheres to specifically bind to the tumorous cell lines. Coverslips were treated as described above with either folate-immobilized microsphere or with controls (no folate immobilization). The coverslips were then examined under the microscope for microspheres that were specifically bound to cells. Images were acquired

PAGE 86

70 on the Zeiss Axioplan 2 microscope using a 10 x Neoplan-fluar objective and a collection area of 1300 x 1030 pixels. The image captured was 1.0683 m per pixel (horizontal and vertical same, calibration from microscope supplier)thus approximately 1400 x 1100 m giving an area of 1.53 mm2. 4.4.3.4 Fluorescently labeled microspheres Microspheres that were fluorescently labeled possessed an additional mechanism by which they could be distinguished from cells. The capture of this information required additional imaging steps. Experiments involving specifically bound microspheres on cells were imaged by transmitted light microscopy, and then by fluorescence microscopy for the same frame and magnification, with appropriate adjustments to exposure settings. In some cases it was possible to create a combined, or dual image involving low intensity transmitted light to provide feature definition, as well as fluorescence imaging to provide spatial information on the fluorophores. 4.4.3.5 Image treatment and analysis Images were acquired with the goal of generating data to turn into a test statistic to evaluate the performance of microsphere specific binding. This process required a number of steps. First, cell culture samples were treated with microsphereseither folate-immobilized or controls. Then images were taken of representative areas from each sample. Areas were chosen that could produce the best quality image (thus the most reliable data). Images of these areas were acquired, and then analyzed manually. The images were divided into smaller areas using Axiovision 3.1 software, and microsphere contact points to cells were counted. Additional data collected included average

PAGE 87

71 microsphere and cell sizes and ranges. This same general procedure was followed for all the cell lines, as well as for the tissue samples. 4.4.3.6 Microsphere recovery experiments Experiments involving the recovery of the test microspheres were performed in two ways. In all cases, the samples were treated as described above, with either folate-immobilized or control microspheres. In most instances, samples were then viewed under the microscope to determine the extent of specific binding that had occurred. In some cases, the samples were not imaged previous to microsphere recovery. Recovery of the microspheres was performed using several procedures to dislodge the microspheres from their specific binding. It was known that acidic conditions caused endosomal release within cells (Low et al. 2001), so acidic solutions were sampled to dislodge the microspheres. In all cases, the wash procedure was performed on the sample using the same procedure as for rinsing (described in section 4.4.3.2). Agents used as the rinse included 1 through 5 v/v % aqueous acetic acid (2.6 to 2.2 pH). Other agents experimented with included trypsin and mucolytic agents, as well as acidic drink products (Coke). Microspheres were collected by placing a magnet under the rinse container for 30 s and the decanting the rinse solution. The microspheres retained could then be counted and/or imaged. 4.5 Tissue Testing 4.5.1 Institutional Review Board (IRB) Approval IRB approval is required for all research projects involving human subjects. The definition of human subjects includes excised tissue. IRB approval was granted on the basis of Expedited IRB #488-2003. This approval included a waiver of the requirement of patient consent forms, since the study posed no possible hazard to the patient. All

PAGE 88

72 tissues used were excised solely on the basis of clinical diagnoses for pathological reasons. No records exist linking the patients to the tissues studied, and the study personnel were blinded to the records. Only information concerning the acquisition date, the pathological diagnosis (by study pathologists and Shands pathology assistants), and the nature of the tissue (type and source location, pathological status) were recorded. This information was recorded on a standardized form to improve records keeping and provide compliance with IRB office policy. 4.5.2 Tissue Preparation The tissues used for testing were acquired exactly as outlined in the application for Expedited IRB #488-2003. Tissue was utilized fresh, or snap frozen after surgical excision. Several types of samples were prepared according to recommendations by pathology personnel. 4.5.2.1 Fresh tissue samples Fresh samples were tested in several ways. Initial testing took place using sections cut from a larger resectioned tumour. These samples were approximately 3 mm x 5 mm x 2-4 mm thick. Later samples were cut to the same dimension at lesser thickness of ~ 1 mm to provide a flatter surface that would allow better focus with the optical microscope. 4.5.2.2 Snap-frozen tissue samples Tissue that could not be used as fresh was snap-frozen in liquid nitrogen to preserve for later testing. The frozen samples could then be sectioned to size for analysis, or mounted on slides. Standard histological preparation methods were followed in mounting the slide samples. The frozen sample was attached to a backing slide and embedded in a polymeric grout. This assembly was placed into a cryomicrotome and

PAGE 89

73 sectioned to desired thickness. Samples used were sectioned to 10 m, 20 m and 30 m thickness. These samples were then mounted onto microscope slides. This procedure was performed by the histological assistants and technicians in the gross pathology laboratory at Shands Health Care Center. Standard H & E (Hematoxylin and Eosin) stained slides of the sample were also prepared. 4.5.3 Sample Treatment for Testing with Microspheres The samples prepared were treated with the procedures established using the cell lines. Variations existed to accommodate the different physical shapes of the unmounted tissue samples and the prepared slides. 4.5.3.1 Unmounted tissue testing procedure The unmounted samples were able to be treated in the same manner as the cover slips that the cells had been grown on, since their size permitted the physical transfer of the samples between cell wells in multi-well plates. The samples were dipped and swirled in a suspension of the testing microspheres in DI water, for 60 s, then dipped and swirled in 3 rinses of PBS 7.4 pH buffer for 60 s each. 4.5.3.2 Slide mounted tissue testing procedure The samples that were mounted on slides could not be physically transferred between wells as above. For these samples, the same concentration of microspheres suspension was prepared, and the slide was placed at an angle in a petri dish, as shown in figure 4-13. A transfer pipette was used to pour suspension over the slide. The suspension collected in the dish and was poured over the slide again for 60 s. This same procedure was repeated with 3 successive rinses of PBS 7.4 pH.

PAGE 90

74 4.5.4 Microscopy of Prepared Tissue Samples The prepared tissue samples were examined in the Zeiss Axioplan 2 imaging microscope. Pictures were taken at 10 x magnification. The images that were recorded included pure optical images (transmitted light, no filter), transmitted light images using the fluorescein filter cube (bandpass filter allowing only green component of light to pass), fluorescence mode (fluorescence excitation) and mixed/dual mode (low level transmitted light together with fluorescence excitation). Whenever possible, each of these four types of images would be acquired for the same area at the same focus to most clearly delineate the structures being viewed. Figure 4-13: Tissue slice mounted on slide being rinsed as part of testing procedure

PAGE 91

75 4.5.5 Magnetic Recovery of Microspheres From Tissue Samples Following the sample preparation procedure outlined in 4.5.3, microspheres were dislodged from receptor mediated binding through the application of an acid wash. 1 v/v % aqueous acetic acid was utilized in the manner appropriate to the sampleapplied in the same manner as the PBS rinse. A magnet was applied to the bottom of the collection vessel (either a dish or a cell well) into which the acetic acid rinse terminated. The acetic acid was then decanted, and the microspheres retained magnetically. The microspheres could then be examined in situ or transferred to permit microscopy or spectrometry.

PAGE 92

CHAPTER 5 RESULTS AND DISCUSSION The design requirements presented in chapter 3 called for a size range of microspheres between approximately 2 and 100 m, with a primary amine group available for ligand coupling and superparamagnetic inclusions to permit separation. The Microspheres were prepared in this size range by a suspension copolymerization of methyl methacrylate (MMA) and aminoethyl methacrylate (AEMH) monomers, the former doped with coated magnetite. The design requirements also called for a method of dye labeling, for which two approaches were demonstrated: a direct covalent fluorophore labeling, and a swelling-based dye incorporation. Characterization for size, surface chemistry, fluorescence, magnetic recovery and ligand adhesion was performed, demonstrating that all the design requirements were fulfilled. The microspheres were tested using various cell lines to demonstrate receptor specific binding. Recovery of the bound particles was demonstrated using an acid wash to sever the receptor binding followed by magnetic recovery of the microspheres. Fluorescence microscopy was utilized both to analyze the microspheres themselves and to acquire spatial information on the particle-cell interactions. 5.1 Microsphere Synthesis Suspension polymerization was ultimately chosen as the best synthesis route to form the desired particles. This conclusion was based on work done producing microspheres by suspension polymerization as well as by dispersion polymerization, including activated swelling procedures that used dispersion polymerized seed particles. 76

PAGE 93

77 5.1.1 Dispersion Polymerized Samples Dispersion polymerized microspheres had been produced in our group on many occasions (Leckey 1997), in some cases with many of the required characteristics for the present application. Styrene based microspheres polymerized with diethyl aminoethyl methacrylate were prepared in a dispersion medium containing iron oxide to yield the product shown in the SEM micrograph of figure 5-1. Appendix A provides detailed polymerization records on each of the polymerized samples that are discussed in the text. Dispersion polymerized samples offered the opportunity to investigate a number of incorporation mechanisms for iron oxides. Several of the strategies that produced viable magnetically responsive microspheres are detailed below, along with the microspheres generated. Figure 5-1: Sample D013-4; ST-co-DEA particles dispersion polymerized in ferrofluid 5.1.1.1 D013 dispersion polymerization with ferrofluid The dispersion polymerized batch of microspheres D013 was produced from a feedstock of 69 vol % styrene (ST), 30 vol % diethyl aminoethyl methacrylates (DEA)

PAGE 94

78 and 1 vol % crosslinkerdivinyl benzene (DVB). This monomer mixture was polymerized in 70 % methanol and 30 % water, with 3 wt % polyvinyl pyrrolidone (PVP-40) as stabilizer. For the particular sample shown, half of the water in the dispersion solution had been replaced with a 3.5 wt % ferrofluid. Figure 5-1 clearly shows that the product from batch D013 was very polydisperse with irregular shape. While the shape and size of the product was unsatisfactory, the EDS spectrum in figure 5-2 below demonstrates a strong response for iron in the labeled peaks. Phase separation apparently resulted, yielding a large portion of segregated, pure white polymer and a smaller, fairly agglomerated mass of black polymer-iron oxide mixtureboth of which can be seen in the micrograph of figure 5-1. This system did not yield an overall satisfactory microsphere for any application, but it did demonstrate that iron oxide could be polymerized into particles in a one-pot polymerization procedure. Figure 5-2: EDS spectra of sample D013-4 showing strong iron peaks 5.1.1.2 D030 dispersion polymerization with iron oxide precipitated in situ A second strategy for the incorporation of iron oxides that was also investigated was direct in situ precipitation of iron oxide (after microsphere formation). The iron

PAGE 95

79 oxide was formed by the same mechanism as described in section 4.1.2a base catalyzed precipitation of iron oxide in magnetite from iron chlorides. The formulation for the microspheres utilized the same monomer feedstock as the previous batch presented: 69/30/1 ST/DEA/DVB. For the particular sample shownD030-3Mthe dispersion medium was 85 vol % denatured ethanol and 15 vol % distilled water, with 12 w/v % stabilizer (PVP40). The DEA monomer (structure shown in figure 5.3) provided sites for the iron chlorides to coordinate to at its amine junctions, as per Ugelstad patent literature (Ugelstad et al. 1988). Figure 5-3: Diethyl aminoethyl methacrylate (DEA) structure The results with this process were microspheres of 1.88 m volume average diameter, with a standard deviation of 0.29 m, as determined by Coulter LS 230 light scattering particle size analysis. Figure 5-4 shows the original microspheres formed in panels a and b. A suspension of the microspheres had iron oxide precipitated onto (and/or into) themthese samples are shown in panels c and d and clearly show a finer surface roughness probably due to precipitated iron oxide. The magnetized D030_4m samples were sized at a volume average of 4.42 m diameter with a standard deviation of 5.90 mindicating agglomeration as shown in the particle adhesions in panels c and d of figure 5-4.

PAGE 96

80 Figure 5-4: Samples D030_3. Panels a and b show the dispersion polymerized microspheres as produced. Panels c and d show the microspheres after iron oxide precipitation. The magnetic character of the particles was determined in several ways. First, XRD analysis shows, as illustrated in figure 5-5 (spectrum) and figure 5-6 (spectrum data listing), that the identity of the precipitated species was not only magnetite, but included maghemitean oxidation product of magnetite. This behaviour was in no way contradictory to observations that the particles reacted appropriately for superparamagnetic species since both magnetite and maghemite are known superparamagnetics (Barthelmy 2004). Second, as the microspheres were magnetically separated from suspension, and they did not agglomerate when outside a magnetic field,

PAGE 97

81 the magnetic character of the material could be taken as superparamagnetic on this empirical evidence. Third, the iron content of the sample was analyzed by ICP, and found to amount to 4.2 wt % iron, which in turn equates to 5.8 wt % iron oxide (this calculation is based on an assumption of 100 % magnetite. This is not strictly accurate, but calculation with even 100 % maghemite would not alter the results beyond the magnitude of experimental error, which is about 5 % for the instrumental measurement, and at least another 5 % for handing and processing). Figure 5-5: XRD spectrum from sample D030_4m. The conclusions from the preparation and characterization of dispersion polymerized samples D013 and D030 was that small particles with superparamagnetic character could be produced by dispersion polymerization methods. These particles were however neither as monodisperse as non-magnetic dispersion polymerized particles shown in figure 5-7, nor as regularly shaped. Additionally, the dispersion polymerized particles could not be formed at an average size of greater than a few m.

PAGE 98

82 Figure 5-6: XRD data listing from spectrum shown in figure 5-5 for sample D030_4m.

PAGE 99

83 Figure 5-7: D009 dispersion polymerized microspheres. 69/30/1 ST/DEA/DVB feedstock composition. Dispersion polymerization as a process for making the particles required for the application detailed in chapter 3 was thus set aside. This was due partly to the above results, and partly due to the nature of the dispersion polymerization process. The required monomers with amine functionalities are typically more water soluble than the main monomer used: methyl methacrylate. This mismatch in water solubility would lead to a significant loss of the more soluble reactant to solution. Compensation can be made in the monomer feedstock ratio, but the development time entailed in devising a dispersion system (such systems are largely empirical in nature) was known to be significant. It was thus judged unlikely that the time would be well spent, in view of the above difficulties that would still apply to dispersion polymerized microspheres.

PAGE 100

84 5.1.2 Activated Swelling The activated swelling process was developed by Ugelstad, who found that the degree of swelling that a polymer particle was able to undergo was radically increased through activation by a low molecular weight solvent (Ugelstad 1978). While non-activated particles are able to swell by only approximately 8-fold (as measured by either mass or volume) when swollen with monomer, activated particles were able to swell by more than 100-fold (Ugelstad et al. 1979). The swelling process relies on a seed particle that is amenable to swellingmeaning that it has no crosslinker in the composition. A common particle type used for swelling was dispersion or emulsion polymerized styrene, since monodisperse seeds permit the formation of a monodisperse swollen particle population. Seed particles made by dispersion polymerization of styrene are shown in 5-8. Final particle composition can be controlled through the swelling mixture, permitting a range of chemistries that would not otherwise be amenable to dispersion or emulsion polymerization techniques. The particular batch of particles pictured in figure 5-8 served as seeds for activated swelling reactions AS02 through AS06, and nicely monodisperse as shown by the size histogram in figure 5-9, where the size of AS05 is shown alongside. The volume average particle size was measured by Coulter LS 230 as 1.25 m, with a standard deviation of 0.16 m. These seeds were activated by forming an emulsion of dibutyl phthalate (DBP) in aqueous 0.25 % sodium dodecyl sulfate (SDS), and adding the seeds as per (Tuncel 1999; Tuncel et al. 1999; Tuncel and Cicek 2000; Tuncel et al. 2002a; Tuncel et al. 2002b). Once the seeds were activated, the emulsion of monomers and

PAGE 101

85 stabilizer was added to the activated seed emulsion. It was found that the system was robust to changes in stabilizer identityforming the same result with either polyvinyl alcohol (PVA) or polyvinyl pyrrolidone (PVP) stabilizer. The size graph in figure 5-9 shows the seeds (D052_3) and final swollen and polymerized particle (AS05) sizes as measured by volume average. Sample AS05 showed a bimodal distribution of sizes, with the larger size initially believed to be a doublet of the more numerous smaller size. SEM pictures as shown in figure 5-10 showed that there was indeed a population of larger sized particles amongst more numerous smaller particles. The particles in sample AS05 had been swollen with a 50/50 volumetric mixture of styrene and divinyl benzene. This showed that we were able to reproduce the swelling results of (Wang et al. 1992; Galia et al. 1994; Wang et al. 1994; Tuncel 1999). Figure 5-8: Seed particles D052 for activated swelling made by dispersion polymerization of styrene

PAGE 102

86 1101000510152025303540 AS05 sample D052_3 seedsVolume %Volume average size in um Figure 5-9: Size graph for sample AS05 and its seeds D052_3 Figure 5-10: SEM micrograph of sample AS05

PAGE 103

87 The activated swelling method was also used to prepare particles containing more water soluble monomers that are important to the compositions deemed desirable for the ultimate application for this project. Sample AS08 used 100 % ST seeds of 1.43 m volume average diameter, with a standard deviation of 0.19 m. These particles were again activated with DBP and swollen with equal parts styrene, hydroxyethyl methacrylate (HEMA) and ethylene glycol dimethacrylate (EGDMA). Figure 5-11 shows an SEM micrograph detailing the product from this experiment. There are again two apparent populations of particles, but the particles are less regularly shaped than for AS05. No elemental analysis was performed on this sample to determine the true composition, but a significant amount of HEMA may partition into the water phase prior to polymerization, due to high water solubility. Figure 5-11: Sample AS08, polymerized from equal parts ST/HEMA/EGDMA Experiments involving diethyl aminoethyl methacrylate as one of the swelling monomers were unsuccessful as far as forming regularly sized particles. This was also true for admittedly very limited attempts at swelling methacrylate based seed particles, although literature indicates that some groups have been successful with systems of this type (Ugelstad et al. 1988; Tuncel et al. 2002; Tuncel et al. 2002).

PAGE 104

88 The activated swelling experiments demonstrated an avenue for arriving at desirable compositions from a seed population with desirable size characteristics. It appears that if limitations on the degree of swelling are acceptable, then this method may prove a viable avenue for post-polymerization modification of the chemistry of particles. Whether this is limited to the surface chemistry would be determined by the porosity and the degree of crosslinking of the seeds, since more porous and uncrosslinked particles should permit more monomer into the interior of their structures. 5.1.3 Suspension Polymerization 5.1.3.1 Magnetic dopant characterization The oleate-coated magnetite had been fully characterized in cooperation with, and as reported by (Leamy 2003) in our labs for magnetic response and size. Leamy confirmed by XRD and SQUID magnetometry that the particles formed were indeed superparamagnetic. The XRD pattern found by Leamy agreed with that presented in figure 5-5 that indicated magnetite, maghemite or a mixture or the two. This observation was in concurrence with results of (Robineau and Zins 1995), whose determination based on chemical analysis. The average size was estimated at approximately 11 nm, based on TEM and XRD evidence. The SQUID magnetometry curve determined is shown below in figure 5-12. The average value of the saturation magnetization for the three samples tested was 68.2 emu/g.

PAGE 105

89 Figure 5-12: SQUID magnetometer magnetic hysteresis curve for uncoated iron oxide 5.1.3.2 Suspension polymerization methods and incorporation of iron oxide The particles that were ultimately used for cell and tissue testing were all produced via the suspension polymerization method detailed in chapter 4. Suspension polymerization systems derived from numerous works (Leckey 1997; Shim et al. 2002; Sivakumar and Rao 2002) were investigated. The (Shim et al. 2002) system had several advantages, including a chemically uncomplicated suspension medium. Protocols were tried repeatedly, and results for systems without seed particles were generally acceptable for all systems (data on results supplied in table 5-1). The preference for the Shim protocol emerged due to the chemical similarity of the dopant required by this project and the seeds applied by them. Shim and Kim suspension polymerized PMMA particles with inorganic oxide seeds for sunscreen applications. They found that their inorganic oxides (TiO2 and ZnO, hereafter referred to as seeds) were homogeneously dispersed throughout the microspheres formed. They also found that the integration of the seed particles was

PAGE 106

90 improved significantly by coating with compatibilizing agents. This is because conventional inorganic oxides have high surface polarityresulting in very low stability in organic media. Coating with an amphiphilic molecule such as stearic acid in the case of Kim (Kim et al. 2002), or oleic acid, as applied in our labs, provides a better surface interface between the particles and organic media (such as MMA monomer). Kim (Kim et al. 2002) found that when they coated their inorganic oxides with stearic acid and then PDMS they were able to increase the loading efficiency of inorganic oxide into their microspheres (in situ suspension polymerized) from 21 % for the bare, uncoated inorganic oxide to 95 % for the coated product. This same trend was observed anecdotally within the experimental series tabulated below. No measure of the loading efficiency was performed, but the findings of Shim and Kim, together with the observation that the iron oxide suspended better in MMA monomer led to the adoption of this technique. It was found that this change led to polymerization with less phase separation into pure polymer (white phase) and mixed product (dark phase). All suspension polymerizations after S04 were thus performed using iron oxide that was coated not only with oleic acid, but also PDMS. Appendix C contains a movie demonstrating the effective separation of magnetic particles (sample S16) from a suspension in 12 mm thick centrifuge tubes. The suspension polymerized microspheres required an aminated surface to enable to immobilization of the species that would provide specific binding, as detailed in chapter 3. Experiments with aminostyrene as the functional monomer formed some microspheres that appeared adequate, but the amine species resident on the benzene ring is not as reactive as a primary amine due to electron delocalization on the ring. Maleic

PAGE 107

91 anhydride was also utilized in several compositions (S11 through S15), but was ultimately superceded by AEMH as the functional monomer of choice. The drawback of using Maleic anhydride was primarily that it required a secondary treatment with ethylene diamine to modify the reactive group to primary amines. The advantage of AEMH was the already resident primary amine that could be utilized in advance of polymerization for conjugation of dye molecules, as well as the methacrylate chemistry that held the promise of incorporating readily into a methacrylate chain. 5.1.3.3 Microsphere size control The evolution of mechanical treatments for the suspension polymerizations revolved around the stirring mechanism, which is known to be a significant lever for altering the product of a suspension polymerization. Initial trials with suspension polymerization omitted magnetic dopants and permitted the use of magnetic stirring. This was not viable for magnetically doped polymerizations, and a mechanical stirrer was applied. When it became desirable to produce a smaller microsphere size, a homogenizer with greater shear power was utilized with success, as illustrated by the size results shown on table 5.1-1. The volume average particle size was reduced from 70 to 30 to under 10 m by variation in the stirring parameters.

PAGE 108

92 Table 5-1: Suspension Polymerization Series Results Sample / Protocol Monomers % / mixture Polymerization time, stirring Vol avg size (m) S01 Sivakumar 12.5 vol %; MMA + 10 % each BPO, EGDMA ~ 3 h Magn stirring 300+ S02 Sivakumar 3 vol %, 80/20 Styrene / aminostyrene + 10 % each AIBN, DVB ~ 15 h Magn stirring 100 S03 Sivakumar 6 vol %, MMA m1 + 10 % each AIBN, DVB ~ 4 h 500 rpm mech stirring 100 200 S04 Sivakumar 5.5 %; MMA m1+ 10 % each AIBN, DVB ~ 3 h 500 rpm mech stirring 100 S05 Shim 3 %; 2 ml MMA m2 + 1 % AIBN, 10 % EGDMA ~ 10 h 500 rpm mech stirring 30 avg, wide distr. From
PAGE 109

93 Table 5-1: Continued Sample / Protocol Monomers % / mixture Polymerization time, stirring Vol avg size (m) S16 9.25 ml; 8 ml MMA m2, 1000 mg AEMH, 125 mg AIBN, 125 l DVB ~ 2.5 d 700 rpm mech stirring ~ 30 S17 Samples destroyed in preparation S18 5.25 ml; 4 ml MMA m2, 1000 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h Homogenizer ~ 12 S19 5.25 ml; 4 ml MMA m2, 20 mg FITC conjugated to 20 mg AEMH, 980 mg AEMH, 125 mg AIBN, 125 l DVB ~ 4 h Homogenizer ~ 9 S20 5.25 ml; 4 ml MMA m2, 1000 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h homogenizer ~ 14 S21 5.25 ml; 4 ml MMA m2, 2 mg FITC conjugated to 100 mg AEMH, 900 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h Homogenizer ~ 14 S22 5.25 ml; 4 ml MMA m2, 5 mg FITC conjugated to 100 mg AEMH, 900 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h Homogenizer ~ 12 S23 5.25 ml; 4 ml MMA m2, 10 mg FITC conjugated to 100 mg AEMH, 900 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h Homogenizer ~ 4 S24 5.25 ml; 4 ml MMA m2, 20 mg FITC conjugated to 100 mg AEMH, 900 mg AEMH, 125 mg AIBN, 125 l DVB ~ 6 h homogenizer ~ 15 some clumping S25 2.63 ml; 2 ml MMA m2, 5 mg Texas Red-X conjugated to 200 mg AEMH, 300 mg AEMH, 63 mg AIBN, 63 l DVB ~ 6h homogenizer ~ 5 Notes: Sivakumar and Shim protocols are detailed in chapter 4; Polymerization temperature was in all cases 75o C; MMA m1 = MMA doped with 5 w% oleate coated magnetite; MMA m2 = MMA doped with 5 wt % oleate and pdms coated magnetite;

PAGE 110

94 5.1.3.4 Particle morphology Microspheres produced by suspension polymerization are known to have significant size polydispersity (Arshady 1992). Figures from literature also show that particle morphology for doped suspension polymerized particles can include significant surface roughness, as shown by figure 5-13 below. This and other literature sources agree with present findings on the regularity (both size and morphology) of suspension polymerized doped particles. Figure 5-14 shows sample S04, made with oleate coated magnetite dispersed in methyl methacrylate, also shaped rather irregularly, and probably featuring some degree of agglomeration. Figure 5-13: TiO2 (10 w/v %) doped PMMA particles produced by Shims suspension polymerization process applied for most samples produced in this study. Note surface roughness and size polydispersity. The matter of irregular morphology does not diminish the functional value of the microspheres, insofar as the surface chemistry is as expected and the magnetic response is adequate. A caveat for non-spherical particles is, however, that light-scattering based particle sizing methods will generate less accurate data, since the assumption built into the sizing models is for spherical particles.

PAGE 111

95 Figure 5-14: Sample S04 showing particles formed by magnetite doped suspension polymerization 5.2 Folic Acid Immobilization Microspheres were synthesized to possess an aminated surface, such that water soluble carbodiimide based bioconjugation could be easily performed to link folic acid onto the microsphere surface. Microspheres were extensively washed after the immobilization procedure was performed, and subjected to characterization to confirm the presence of folic acid. 5.2.1 UV-Visible Spectrophotometry Folic acid contains two independent chemical moieties that are known to be fluorescent. P-amino benzoic acid (PABA) absorbs at 265 nm and emits at 336 nm. Methylpteridin (MTE) has absorption bands at 275 nm and at 352 nm, and emits at 447 nm (Tanojo et al. 1997; Espinosa-Mansilla et al. 1998). The folic acid absorption spectrum represents the combined absorption of these species. The spectrum that was measured for 50 M folic acid in water using a Shimadzu 2401PC UV-Visible recording spectrophotometer is shown in figure 5-15 on the green curve. This folic acid spectrum

PAGE 112

96 showed absorption maxima at 281 nm and at 355 nm, providing good agreement with the literature values. The maroon curve shows S19 control particles (no surface immobilization), and the olive curve shows S19 immobilized particles. This particular image shows that there is no response from the UV-Vis absorbance spectrometer that would provide evidence of immobilized folate on the microsphere surface. Since the folic acid immobilization utilized what was believed to be reliable and proven chemistry, it was theorized that the high background from particle based scattering, combined with the low surface area of the microspheres may not have allowed a signal strong enough for detection to be generated. This led to the use of the model system described in detail in chapter 6utilizing silica nanospheres with much higher surface area and more uniform size to demonstrate that the binding chemistry was functional. The result presented in figure 5-15 is at best inconclusive with regard to confirming folate immobilization on the surface of the microspheres, which led to a search for other methods of characterizing the surface. 5.2.2 Fluorescence Spectrometry Fluorescence spectrometry suffered the same light scattering based confounds as UV-visible absorption spectroscopy, and yielded no data of conclusive value (data not shown). 5.2.3 Zeta Potential Measurement Zeta potential was measured in PBS 7.4 pH, comparing folate immobilized microspheres to control microspheres. Despite repeated measurements, the quality of the data was such that no significant difference could be determined between the two groups. A collected data set is shown in table 5-2, with the individual runs that comprise the averages depicted in order to show the data variability. The result of a 2 tailed t-test is

PAGE 113

97 appended at the bottom of the table: the p-value of 0.95 could not be more clear in showing that the difference between the two sample values is not of significance. The data was collected from 3 separate samples that were run for 10 cycles each to determine a value. Repetitions of the same experiment yielded roughly the same results. The primary factors contributing to the lack of significance in the differences between the groups are that the variation between measurements on the same samples was very high, and that the averages of the values did not turn out to be very different. Possible explanations for this include settling of the large, heavy particles during measurement as well as charge heterogeneity for the particles within each sample. Figure 5-15: UV-Visible absorption spectrum shown by 50 uM folic acid (green), S19 control particles (maroon) and S19 immobilized particles (olive).

PAGE 114

98 Table 5-2: Zeta potential measurement on folate-immobilized and control s19 microspheres. Averages presented are for the three runs depicted, each consisting of 10 cycles. Zeta potential (mV) Half width (mV) Sample value std dev value std dev S19F1 6.32 8.4 4.52 0.47 S19F2 -13.6 1.71 3.53 0.16 S19F3 0.35 1.85 4.44 0.09 averages -2.31 3.99 4.16 0.24 S19C1 2.48 2.21 3.82 0.45 S19C2 -10.96 11.09 4.11 0.36 S19C3 0.08 3.76 3.84 0.21 averages -2.80 5.69 3.92 0.34 P value of 2 tailed t-test: 0.95 = No significant difference! In light of the poor quality data that was obtained directly characterizing the folate immobilized microspheres, the model system presented in chapter 6 was shown to be all the more important. The smaller size and greater specific surface area of the DDS nanospheres, as well as the low polydispersity all contribute to producing cogent characterization data. 5.3 Microsphere Labeling 5.3.1 Dye Loading Vs. Covalent Coupling of Dye Microspheres were labeled by two different methods: by loading with dye and by covalent binding of dye into the microsphere. Each system has theoretical virtues as well as practical considerations weighing for and against them. Both of the methods were tested, and the covalent binding of dye into microspheres was chosen for development work because it offers improved spatial resolution during microscopy work. The advantages of dye loading are that the dye can be separated from the microspheres. This is important because the polydisperse microspheres cause significant

PAGE 115

99 light scattering that results in large amounts of noise for spectroscopy systemsenough to limit sensitivity at the least and to confound measurement completely at worst. The advantages of covalent dye binding are twofold. First, the dyed particle is coloured for easy recognitionproviding spatial information that is not available with a dark (because of the iron oxide content) particle that is loaded with dye. When fluorophores are used as dye then this enables fluorescence microscopy and fluorescent detection systems. Since covalently bound dyes can be integrated during the particle synthesis (conjugated to monomer), no further chemistry has to be performed for labeling. If a surface species is to be immobilized for specific binding, then it is particularly advantageous that the species not be subjected to solvents that may compromise its activity or biochemical recognition capability. Dye loading by swelling requires a choice of which chemistry to perform firstswelling in a solvent to load the dye and then surface immobilization, or vice versa. The correct choice of order for performing this chemistry may still entail a tradeoff between surface immobilization efficiency and dye loading effectiveness. During development of the microsphere binding process it was critical to monitor the binding effects accurately. The correct system for the application ultimately envisioned may not be the best system for the requirements of development work. Research concentrated mainly on the development system, since receptor binding and its correct characterization was judged as the major challenge of the project. 5.3.2 Dye Loading by Swelling The process of dye loading by swelling requires determination of several parameters for the system. Once a microsphere composition is chosen, it must be

PAGE 116

100 determined what solvents can provide swelling for this composition. Polymer solubility tables and calculations represent the quantification of this data (in this work, the common term polymer solubility parameter is used to refer to the Hildebrand version that takes only individual chemical groups into consideration, not more detailed parameters such as by Hansen that include dispersive, dipole and hydrogen bonding forces). The suspension polymerized microspheres were in each case 80 vol % Methyl Methacrylate (MMA), and the known polymer solubility value for MMA is 9.5 (cal/cm3)1/2. Good solvents suggested for pure MMA include dibutyl phthalate, tetrahydrofuran (THF), and possibly ethanol). Table 5-3 below shows polymer solubility parameters of selected solvents (Sperling 1992; Sigma 2002). Table 5-3: Polymer solubility parameter values for selected solvents, presented in common form of (cal/cm3)1/2, not in SI units. Solvent Chemical formula Solubility parameter (delta) Polarity parameter (P') Chloroform CHCl3 9.3 4.1 THF C4H8O 9.1 4.0 Acetone CH3COCH3 9.9 5.1 Dimethylsulfoxide (CH3)2SO 12.0 6.3 Ethanol C2H5OH 12.7 4.3 Methanol CH3OH 14.5 5.1 Acetic acid CH3COOH 10.1 6.0 Water H2O 23.4 10.2 Polymer solubility parameter values for many common monomers, and their attendant homopolymers are readily available. Less commonly applied monomers such as AEMH do not have readily accessible polymer solubility parameter data. In such cases the data can be constructed using group molar interaction theory (Sperling 1992). The solubility parameter is equal to the square root of cohesive energy density, c. The

PAGE 117

101 cohesive energy density is quantifiable through either of the representations in equations 5-1 and 5-2: c = (H RT) / Vm (equation 5-1) = c1/2 = (E / V)1/2 (equation 5-2) where: H = heat of vaporization R = gas constant T = absolute temperature Vm = molar volume E = energy of vaporization to a gas at zero pressure Tables exist that quantify the contributions of various chemical groups, but the tables are incomplete, and the data are not interchangeable between the tables because the measurements were taken utilizing different physical properties to arrive at the data (the individual contributions were calculated by Hoy from vapor pressure data and by Small from heat of vaporization data.). A value was calculated for AEMH based on the data of Hoy, since only that data table provided values for the contributions of all chemical groups present in the monomer. Data was calculated for a unit of monomer (without the coordinated HCl), with the density approximated as 1.0 since no published value was found. This calculation and data is shown in table 5-4. It shows that the polymer solubility value of AEMH is numerically slightly higher than that of MMA (when calculating for MMA in the same way and using the same data table) (Sperling 1992). Swelling experiments were performed using the Zeiss Axioplan 2 imaging microscope, in time-lapse mode so that swelling could be tracked visually in time-compressed fashion. The findings are tabulated in table 5-5 detailing solvent based swelling, as well as results of dissolving fluorescein dye in the solvent. In the case of methanol, the fluorescein solvation trial was performed prior to the swelling experiment, obviating the need to test swelling. The findings of these experiments were that DMSO

PAGE 118

102 and THF both were better swelling agents than the alcohols. The confidence in the data collected was relatively low, as the experiments appeared to have sizable error associated with them. The error sources include movement of the particles in solution and problems with maintenance of precise focus on the particles in solution. The conclusion as regards swelling is that THF and DMSO both appear to swell the polymer particles over 2 % (measured linearly). Ethanol does the same, but was eliminated form contention for not dissolving the dye in question. Table 5-4: Group Molar Contribution calculation of Hildebrand polymer solubility parameter for AEMH monomer unit = group contributions density / MW = 19.31 MPa1/2 9.44 (cal/cm3)1/2 Group contributions Hoy values Type # of Single Sums =CH2 1 259 259 =C< 1 173 173 -CH3 1 303 303 -CO1 538 538 -O(ether) 1 235 235 -CH22 269 538 -NH2 1 464 464 sum of contributions 2510 AEMH MW = 100 AEMH density = 1 Table 5-5: Swelling of microspheres by selected solvents and solvent-dye compatibility Solvent Hildebrand solubility parameter Swelling (linear %) Fluorescein compatible? (1 w/v % solution) THF 9.1 >2.5 Yes DMSO 12.0 2.3 Yes Methoxyethanol >ethanol <1.5 Yes Ethanol 12.7 2.1 No Methanol 14.5 Not tested No In order to test loading and release, microspheres from sample S20 were loaded with dye using both DMSO and THF. In each case 16 mg of microspheres were swollen with 5 ml of 1 w/v % fluorescein-solvent solution. It was expected that 5-8 washes

PAGE 119

103 would leave the microsphere surface clean of dye. The DMSO swollen sample leached dye into water even after 20 washes. The THF washed sample washed clear after 8 washes. A discernible release of dye was found for both samples with swelling in solvent. No attempt was made to rigorously quantify release. The THF swollen and released sample released more dye based on visual inspection. It was theorized that the DMSO swollen sample may have never completely had the solvent evaporated off, since DMSO has a lower vapour pressure than alcohols and THF and is relatively difficult to remove. This recommended THF as the best solvent candidate for swelling and release. Samples were prepared by swelling in a 3 w/v % fluorescein in THF solution, separated magnetically and dried in vacuum. Theses samples were washed 8x with water and dried again. The samples were then swollen in THF to extract the dye. The supernatant was separated from the microspheres magnetically and analyzed by UV-visible absorption spectroscopy. The results indicated that an improbably large amount of dye had been released, measuring up to 20 w % relative to the microspheres. The improbably high value together with reading variability casts doubt on these results. There is no doubt about the observation that an amount of dye was extracted from the microspheres that is clearly clinically and applicationally relevant and significant. 5.3.3 Covalent Coupling of Dye to Microspheres Dye was covalently coupled to microspheres by means of amine functionalities on the dyes. Both dyes: Fluorescein Isothiocyanate and Texas Red-X were purchased with chemistries that permit easy conjugation. In both cases the dyes were conjugated to the aminated functional monomer AEMH. The monomer-dye conjugate was added to the monomer feedstock and polymerized in the normal fashion. This produced satisfactory results, as are shown below.

PAGE 120

104 5.3.3.1 Microsphere fluorescence Microscopy of the dyed microspheres shows, as depicted in the panel of figure 5-16 below, that virtually all microspheres contained some dye. Not all microspheres showed equal fluorescence intensity, nor was the fluorescence intensity of an individual microsphere homogenous throughout that microsphere, as shown in figure 5-17. Figure 5-16: S19 particles dispersed on slide at 10x magnification; a) transmitted light image showing microspheres; b) fluorescence image showing emission from virtually all particles. Figure 5-17: S19 microspheres, 40x fluorescent image showing fluorescence intensity inhomogeneity within microspheres. The unequal fluorescence intensity, as well as the heterogeneous fluorescence emission is readily explained through the presence of the magnetic species and the

PAGE 121

105 manner of incorporation of the fluorophores. The magnetic species are fundamentally small particles of perhaps 20 nm average size, but coatings of oleic acid as well as relatively high (~100,000) molecular weight polydimethyl siloxane (PDMS) can lead to agglomeration of many of these small particles. The dispersal of the PDMS coated magnetic particles within the methyl methacrylate monomer may also be imperfect, leading to yet larger agglomerations. Further, the polymerization of the monomers is expected to proceed largely at random, but the different polymer solubility indices of the main monomer (MMA) and the functional monomer (AEMH) may create regions of higher and lower concentration of each. This last point would impact both the distribution of the opaque magnetic species, as well as the distribution of the fluorophores. 5.3.3.2 Confocal microscopy Confocal microscopy supports the above findings on fluorescence intensity distribution throughout and between particles. The data represented by figures 5-18 and 5-19 shows fluorescence heterogeneity between and within the particles. 5.3.3.3 Dye content optimization Dye content was optimized by preparing various fluorophore concentrations in the microspheres, and assessing fluorescence intensity. This was performed only for FITC, since the fluorescence intensity of FITC is known to be very similar to that of Texas Red-X, and the cost of Texas Red-X was prohibitive for this series of experiments (Haugland 2003). The amounts of FITC shown in table 5-6 were conjugated to AEMH monomer, and microspheres were polymerized in the usual fashion using the dye-conjugated monomer as part of monomer feedstock.

PAGE 122

106 Figure 5-18: Confocal image of S19 microspheres at 60x magnification. Left panel shows fluorescence image, right panel shows transmitted light image. Purple line visible on images defines measurement path for light intensity, graphed in figure 5.3-4. 0510152025303540050010001500200025003000350040004500 Transmitted light intensity Fluorescent light intensityIntensity (arbitrary units)Position, um Figure 5-19: Light intensity plot for figure 5-18, as measured along purple line. Red line represents fluorescence intensity from left panel. Black line represents transmitted light intensity from right panel.

PAGE 123

107 Table 5-6: FITC concentrations in sample batches prepared to optimize dye content and fluorescence yield Sample FITC (mg) Moles FITC [FITC] / [monomers] Avg fluoro intensity Std dev p-value in t-test S21 2 5.14E-06 1.12E-04 3.60E+05 3.52E+03 1.54E-02 S22 5 1.28E-05 2.79E-04 3.55E+05 3.35E+04 6.34E-05 S23 10 2.57E-05 5.58E-04 4.83E+05 7.11E+03 8.19E-01 S24 20 5.14E-05 1.12E-03 4.87E+05 2.32E+04 Note: Fluorophore concentration calculated based on assumption that the full amount of feedstock polymerizes. Fluorescence intensity measurements are in arbitrary units, and are measure above variable background noise, preventing conclusions based on absolute measures of fluorescence intensity. P-values calculated are for 2-tailed t-test comparing groups with group below. The starting points for the fluorophore optimization were chosen in view of data known from the DDS model system detailed in chapter 6. For that system, data not presented confirm the findings of (Vanblaaderen and Vrij 1992) that the ideal FITC concentration was approximately 1 x 10-4 M. The fluorophore utilization in the DDS system is quite efficient because the silica particles at that size are translucent to lightpermitting excitation and emission from all fluorophores entrapped within the particle. This is not true for the microspheres, where fluorophores not near the surface are likely to be obscured by the opaque iron oxide species. The optimal FITC concentration from the model system was utilized as the lowest tested concentration for the microspheres. Fluorescence intensity of the microspheres in this series of experiments was compared by use of the fluorospectrometer. It was found that the size heterogeneity of the particles contributed to significant noise in the system, obscuring the emission peak at 515 nm when the FITC was excited at the primary excitation wavelength of 488 nm. Adjusting the excitation wavelength to 450 nm still resulted in a strong shoulder across the 515 nm emission peak. It was suggested that this might be due to water based Raman

PAGE 124

108 scattering. The Raman line from excitation at 488 nm is produced at 525, and is strong enough to mask low concentrations of fluorescein (Biosciences 2004), so excitation at 450 nm may still induce an obscuring shoulder, since the excitation of fluorescein will be weaker at the lower frequency alsoyielding a lower fluorescence signal that is more easily overpowered by noise. Excitation at 350 nm finally produced lower fluorescence yields, but permitted collection of data that was not completely obscured by noise. The data presented in table 5-6 shows fluorescence intensity measurements averaged from 3 runs, with standard deviations. The data indicate that there is a small but statistically significant (at 95 % confidence) difference between samples s21 and s22, with p = 0.015. Between s23 and s24 there is no statistically significant difference with p = 0.82. There is a statistically significant difference between S22 and S23, with p < 0.001. The danger in professing statistical significance is losing sight of physically meaningful differences, and this applies to the measured difference between S21 and S22, which achieves statistical significance without having physical significance. Only the difference between S22 and S23 represents both statistically significant and physically meaningful data. Based on the FITC dye content optimization data, the dye content of the S23 samples was selected as providing adequate fluorescence intensity. The Texas Red-x labeled conjugate was thus prepared by the same procedure as S23. Fluorescence spectrometry of the S25 samples yielded inconclusive data due to particulate scattering. Confocal microscopy images are shown in the panels of figure 5-20, which clearly demonstrate strong fluorescence from the Texas Red dye. It is apparent from the image that the fluorescence intensity may be more heterogeneous for Texas Red labeled microspheres than for FITC labeled microspheres.

PAGE 125

109 Figure 5-20: Confocal microscope image of sample S25 with Texas Red dye. Left panel shows fluorescence image, right panel shows transmitted light image. 5.4 Cell Line Testing Cell line testing was performed initially with the cells grown in 6-well cell plates. This was found to be inconvenient with regard to performing microscopy on the cells, since the focal distance of the objectives was not great enough to image the cells at the bottom of the wells without entering the wellslimiting the possible movement of the objective. To resolve this, the entire plate was inverted, and imaged through the bottom. The primary drawback of this technique was that the fluid had to be drained from the wells, causing cells in adjacent wells to dry out quickly. The protocol was modified to utilize cover slips, on which the cells were grown. These cover slips were circular, approximately 6 mm diameter, and were treated much the same as the cell wells. An advantage was that the coverslips could be manipulated more independently, and placed on a microscope slide for examination. A disadvantage was that the coverslips, having been prepared by hole punch, had small uneven lips that sometimes prevented them from lying completely flat. Since the optical microscope has

PAGE 126

110 a very narrow plane of focus, some of the coverslips did not permit focus on a very wide area at any one time. Another disadvantage was that while the coverslips were easy to manipulate and immerse in testing and rinsing solutions, it was impossible to prevent retention of some few particles on the bottom of the coverslips, requiring judicious focus and interpretation of the visual results. 5.4.1 Initial Testing with Cell Line NCI-H23 The NCI-H23 cell line was grown in cell wells and tested using suspension polymerized microspheres from batches S11 and S12both immobilized with folic acid on the surface and controls (without surface treatment). Testing was performed on cells that were subconfluent in culture. Briefly, 200 l of microspheres in DI water suspension was added to each well of a 6-well plate, and the plate was gently agitated for 5 min, after which the wells were aspirated and rinse through 3 x with Hanks balanced salt solution, and 1 further time with media. The wells were drained again and the plate was inverted for imaging. 5.4.1.1 Initial testing results Initial testing with samples S11 and S12 in the cell wells determined that the folate-immobilized microspheres did indeed bind preferentially to the cells of the cancerous cell line. Figure 5-21 shows a panel comparing the control (no folate) with the folate-immobilized microspheres from sample S11. There is clearly a significant difference in the number of retained microspheres between the control and the test sample. This is visible in figure 5-21 and is quantified in the table generated from this result: table 5-7. Quantification resulted from counting of the contact points between

PAGE 127

111 microspheres and cells that were visible, and demonstrated a difference that was judged to be clearly significant, both statistically and practically. Figure 5-21: NCI-H23 cell line tested with S11 control microspheres (left) and S11 folate immobilized microspheres (right). Scale bar 50 m. Table 5-7: Initial cell experiments statistical evaluation Sample Number of cells Number of adherent microspheres Average size of adherent microsphere (m), std dev Control: S11C3: 67 1 11.5 2.5 Sample: S11F1: 39 29 8.0 2.0 Once specific binding of folate-immobilized microspheres onto NCI-H23 cells was established (relative to the control group of microspheres), the experiment was verified using another batch of microspheres, and the procedure was refined to permit collection of data for statistical comparisons. 5.4.1.2 Determination of desirable microsphere size from testing Figure 5-22 illustrates two representations of the size distribution of the microspheres from sample S11F1. The lower graph shows the volume size average on a log scale to illustrate that there is a fraction of smaller particles. The volume average size determined for this sample was 210 m, with a standard deviation of 77 m. The seemingly small fraction of smaller particles is miniscule in terms of volumetric average,

PAGE 128

112 but still significant in terms of numerical average. It turns out that the number of smaller particles in the sample was sufficient to provide a meaningful data on the size of adherent particles from the cell testing experiment. Figure 5-22: Volume average size graphs for sample S11 from Coulter LS 230 light scattering particle size analyzer. Upper graph has volume % scale normal, lower graph has volume % on log scale. The most immediately useful data that was derived from the experiment quantified in table 5-7 was the size of microspheres that adhered to the cells. The number average size of particles that were characterized as being specifically bound to cells in the

PAGE 129

113 experiment was 8.0 m, with a standard deviation of 2.5 m. The size determination of microspheres that specifically bound to cells was very meaningful in terms of the design parameters for further microsphere preparation. The information on the average size of binding microspheres was utilized as an iterative solution for the determination of optimal microsphere size to produce, or aim at producing. This solution was enabled by the wide range of microsphere sizes that is produced in a suspension polymerization. 5.4.2 Non-labeled Microspheres Major sources of data from unlabeled microspheres were samples S11, S12 and S18. S18 was utilized with several cell lines, and was of the same composition as all microsphere batches that succeeded it. 5.4.3 Labeled Microspheres Of the labeled microsphere compositions, S19 was the preferred sample batch that was used for data acquisition. 5.4.4 Malignant Cell Lines Testing Cell line data was generated from series of pictures taken across many testing days. Counting for the non-labeled microspheres involved subjective identification of microspheres, which were not always simple to distinguish from cell junctions that also generated dark spots. To minimize the influence of subjective judgments on the counting data, the same person was used to analyze data for an immobilized sample and the controls for that sample.

PAGE 130

114 Table 5-8: Results showing normalized counts of microspheres per unit area for control and immobilized S18 microspheres on NCI-H23 cells Normalized microsphere counts per unit area: in microspheres/m2 S18 controls S18 folate-immobilized 1.09E-04 1.47E-03 7.50E-04 2.79E-04 3.33E-04 3.77E-04 9.09E-04 6.85E-04 1.45E-03 7.26E-04 1.08E-03 4.47E-04 1.15E-03 1.77E-04 8.68E-04 4.08E-04 1.46E-03 4.76E-04 4.98E-04 4.86E-04 Statistics average value: 5.64E-04 9.21E-04 std dev 3.61E-04 4.53E-04 p value 8.48E-02 5.4.4.1 NCI-H23 human lung adenocarcinoma cell line The data for S18 controls and immobilized microspheres respectively is shown in table 5-8. These microspheres were used on the NCI-H23 cell line expressing elevated folate receptors. The statistical data was generated by counting the microspheres bound to cells, and normalizing that count for the area from which the microspheres were counted to arrive at a number for microspheres found per unit area. This method was suitable for the data since all samples were taken at comparable growth times for the cells, so that the number of cells scaled well with area. The deposition rate for the immobilized microspheres was higher at 9.21E-4 vs. 5.64E-4 (microspheres per m2) for the controls. The p-value that was found for this data set was 0.085, indicating that the difference was not significant at the 95 % confidence level.

PAGE 131

115 When a pool of all reliable data was made for the NCI-H23 cell line including data gathered with microsphere lines S11, S18 and S19, the data was normalized by microsphere count. This was done because the cell lines were at different degrees of growth for the various tests. For the pooled NCI-H23 data, the average numbers of microspheres per cell were 0.89 for the folate immobilized microspheres and 0.55 for the controls. The p-value for a two-tailed students t-test with unequal variances was 0.0074. This permits the claim that the difference is statistically significant. Comparing the actual differences in the numerical results shows that while the larger size (n) of the pooled data set enables statistical significance, the difference is not large enough to judge as particularly meaningful for an application. 5.4.4.2 BALB/3T3 mouse fibroblast cell line Microspheres S19 were used to test the BALB/3T3 mouse fibroblast cells. These cells were known to turn malignant in culture. Figure 5-23 shows images of folate-immobilized microspheres bound to the cells in culture. Unfortunately, the cell line met an untimely demise due to contamination before the adequate control tests were made, so that there is no meaningful statistical data from this cell line, though binding is evident. 5.4.4.3 SCC-9 oral squamous cell carcinoma cell line Testing with the SCC-9 cell line was performed to more closely model the cells relevant to the ultimate application. Due to long dormancy, the cells were slow to grow, precluding inclusion for statistical evaluation. On a purely observational level, the microspheres appeared to interact with the SCC-9 cell line in much the same manner as with the NCI-H23 cell line.

PAGE 132

116 Figure 5-23: BALB/3T3 mouse fibroblast cell line with folate-immobilized s19 microspheres. Left image shows transmitted light, right image shows fluorescent light. 5.4.5 Control Cell Line Testing The control cell line that was used was normal human dermal fibroblasts (NHDF). These cells were tested with S19 folate-immobilized microspheres. A test statistic was generated by normalizing the microspheres found per cell, and comparing this information with the figure generated for the NCI-H23 cell line. For n=10 samples counted, the results were an average of 0.89 microspheres per cell for the NCI-H23 cell line, and 0.95 for the NHDF cell line, with standard deviations of 1.20 and 1.39 respectively. The p-value of 0.90 shows that these are not significantly different statistically. Further, the higher value for the control cell line shows that a false positive is being produced. The adherence of the folate-immobilized microspheres to the control cell line has several possible explanations. First, the cell line may be expressing folate receptors at a rate higher than expected, causing specific binding where none was expected. Second, the binding action of the folate-immobilized microspheres may be due to non-specific binding.

PAGE 133

117 5.5 Tissue Testing Tissue samples were freshly excised squamous cell carcinoma of the head and neck region that were excised on pathological grounds by the Department of Surgical Pathology. 5.5.1 Tissue Sample 1 Tissue sample 1 was a human squamous cell carcinoma excised February 6 2004 from the tongue/neck region. The cells were pathologically judged to be poorly differentiated, indicating that they had belonged to a low-grade tumour. 5.5.1.1 Fresh tissue testing with sample 1 The tissue was tested as received and treated with microspheres by the unmounted tissue protocol. Fluorescence microscopy was performed and showed that there was evidence of microsphere binding to the receptor sites. Fluorescence microscopy was impeded by tissue autofluorescence. Figure 5.24a shows tissue autofluorescence in a control sample. Despite the extra noise introduced by the native tissue fluorescence, it was possible to distinguish the immobilized microspheres as shown in figure 5-24b. 5.5.1.2 Mounted tissue testing with sample 1 The micrographs in figures 5-24 a and b show the difficulties associated with microscopy of fresh tissue samples as excised. The thickness of the sample was judged to be excessive at > 3 mm. This definitely limited the transmitted light microscopy that could be done, and may have contributed to the degree of native fluorescence that the tissue displayed. The thickness of the sample also prevented it from lying flat, so that a uniform plane of focus could not be achieved with the 10 x objective on the fluorescence microscope. It was theorized that a thinner sample could minimize the impact of both these factors, but the tissue toughness prevented manual preparation of thinner samples.

PAGE 134

118 Figure 5-24: Fresh tissue sample 1 with S19 immobilized microspheres. a) control microspheres on tumour; b) immobilized microspheres on tumour. Note microsphere fluorescence over background tissue fluorescence Figure 5-25: H&E stained 10 um section of squamous cell carcinoma tumour of tongue/neck region from tissue sample 1

PAGE 135

119 Thinner samples were tested as prepared by pathology technicians on slides, by slicing micrometerthick sections of snap-frozen tissue from sample 1 by microtome. Along with the slides for microsphere testing, a standard H&E stained slide of the same tumour was also prepared, as shown in figure 5-25. The H in H&E stands for Hematoxylin, a basophilic stain that imparts a deep blue or purple colour to structures such as cell nuclei and ribosomes. The E stands for Eosin, which is an acidic stain that imparts a red colour to structures including collagen fibres, red blood cells, muscle filaments and mitochondria (Cormack 1993). The darkest regions of the picture show tumour regions stained blue, surrounded by fibrous connective tissue, stained lighter red. Figure 5-26: 10 um slice of sample 1 tissue mounted on slide and tested with control microspheres. a) transmitted light, no filter, 100 ms exposure; b) fluorescence micrograph of same, fluorescein filter, 2000 ms exposure The panel of micrographs shown in figure 5-26 (a and b) shows a single site on a 10 um thick slice of tissue sample 1 mounted on a slide. These images depict a sample that was prepared with the control microspheres. Despite the thin section of only 10 m, the tissue still generated significant fluorescence. There is little indication of microspheres that have bound to the structure, either as the discrete dark points of characteristic size in the incident light image, nor in the fluorescence image.

PAGE 136

120 The panel of images in figure 5-27 (a through d) shows a slide mounted 10 m slice of the same tissue as in figure 5-26, treated with folate immobilized microspheres. Panels a and b show transmitted light images seen through (Zeiss) filter sets 01 and 10 respectively. Panel c shows an image that combines the transmitted light and fluroescent light excitation, and panel d shows the fluorescence image only (both c and d also used filter set 10). The microspheres that bound to the tissue slice are clearly recognizable in all the imagesas dark masses in a and b and as green fluorescent spots in c and d. The degree of fluorescence intensity seen in the fluorescence excitation panels, particularly for the fluorescence-only image d, where the background brightness (noise) is lower, verifies the black spots as microspheres as opposed to some confound. Figure 5-28 shows the control experiment (tissue treated with control microspheres s19) to accompany the result of figure 5-27. The difference in number of retained microspheres is clearly recognizable between the two figures 5-27 and 5-28. Tissue slides were examined as collected and then frozen for future reference by wrapping in aluminum foil and freezing at o C. When slides prepared from tissue sample 1 were unfrozen and tested 48 h after acquisition it was found that the tissue sample mounted on the slide had apparently degraded and dried. When these samples were tested according to the same protocol used for other slides, it was found that even in the degraded condition the tissue was still able to specifically bind microspheres (data not shown).

PAGE 137

121 d c b a Figure 5-27: Panel of 10 um section mounted tissue from sample 1, treated with folate immobilized microspheres S19, 10x. a) transmitted light image, no filter; b) transmitted light image filter f10; c) fluorescence and transmitted light combined; d) fluorescent light only; all images 1000 ms exposure time.

PAGE 138

122 Figure 5-28: Panel of 10 um section mounted tissue from sample 1, treated with control microspheres s19, 10x. a) transmitted light image, no filter; b) transmitted light image filter f10; c) fluorescence and transmitted light combined; d) fluorescent light only: note tissue autofluorescence.

PAGE 139

123 5.5.2 Tissue Sample 2 It was found from the experiments with tissue sample 1 that the fluorescence microscopy data that could be gathered from direct observation of thick tissue sections was inconclusive due to the low quality of data. Slide-mounted 10 m sections were again obtained, and treated with both immobilized and control microspheres, with substantially the same results as for tissue sample 1. 5.5.3 Results with Tissue Samples The tissue samples showed significantly greater specific binding than the cell lines, at least on the slide mounted sections that were amenable to counting. The data appear very clear, as seen in the images presented. The fresh bulk tissue samples were difficult to work with and yielded no conclusive data because it was impossible to clearly distinguish microspheres from features. The counted microspheres from the tissue slices were normalized according to the area of tissue in the frame (the full frame is approximately 1.5 mm3). This eliminated variance that would be caused by smaller and larger pieces of tissue being imaged. The numbers presented signify the number of microspheres per normalized unit area of tissue slice. Only samples for which the microspheres and the area could be determined with reasonable certainty were used, leaving relatively small sample numbers (n=8 or 9 for each data set presented). It was found that the folate-immobilized microspheres had a significantly greater retention rate on the tumourous tissue slices, with an average value of 41.3 a standard deviation of 40.5, vs the control microspheres 3.4 3.4. This resulted in a p-value of 0.037. These results clearly indicate that statistical significance is met at 95 %

PAGE 140

124 confidence level. Furthermore, the numbers also show that there is a meaningful real-world difference between the two, in spite of the disturbingly high variance in the data sets. The high variance is suspected to arise because all parameters cannot be held constant over testing that spans days and weeks, when samples could degrade in storage, with possibly altered rates of adherence. The results are held to remain valid because test samples and controls were always paired and run together. Testing with normal tissue slices indicated that the folate-immobilized microspheres also had a higher rate of adhesion to the normal tissue than did the control microspheres, as expected. The critical result from the normal tissue tested was that the average normalized number of microspheres per unit area was 116 a standard deviation of 64.2 for the folate-immobilized microspheres. This compares to the average number of 41.3 for the folate immobilized microspheres tested on tumourous tissue. This clearly indicates a false positive signal being generated, at a p-value of 0.016. Explanations for this could include that the normal tissue tested had a true higher-than-expected folate receptor concentration, or that there is non-specific binding being generated by the folate-immobilized microsphereswhich is more likely. 5.6 Microsphere Recovery Experiments Folate-immobilized microspheres that were found by fluorescence microscopy to bind to cell lines and/or tissue samples were recovered as detailed in the application plan. Recovery was attempted by procedures detailed in the methods section, but the microspheres proved to have very strong binding to the tissue/cells. A 1 vol % acetic acid solution (2.6 pH) was proved to have almost no effect in dislodging bound microspheres. Stronger acetic acid solutions also had little effect, up to 5 vol % (2.2 pH),

PAGE 141

125 which represented the same acetic acid concentration as found in vinegar. Further products investigated to dislodge the microspheres included trypsin, acetylcysteine (a mucolytic agent), and Coca Cola (2.4 pH). None of these dislodged the microspheres. The only method of retrieval that proved effective was the use of a permanent magnet. The magnet was placed inside a plastic bag, brought into contact with the sample, and removed with the microspheres attached. The microspheres were removed from the magnet by inverting the bag. This method requires close contact of the magnet with the samplethe range of the technique appears to be only a few millimeters.

PAGE 142

CHAPTER 6 DYE DOPED SILICA PARTICLES 6.1 Introduction Silica particles can be formed in nanometer sizes, in a highly repeatable fashion. The silica surface is known to be relatively inert, but is amenable to a large body of known chemistry. Dye doped silica (DDS) particles were used in this research to model many functions of microsphere surface chemistry. Nanoparticles have a very high specific surface area, so that surface chemistry is very effective, and easier to characterize, simply because there is more surface. Advantages of DDS nanoparticles are that fluorescence intensity is high, and the large number of particles permits ready detection by fluorescence microscopy. The smaller size, combined with low polydispersity permits the ready application of absorbance and fluorescence spectrometry to test surface chemistry. 6.2 Background 6.2.1 Stber Process for Producing Nanoparticles The Stber process is a popular and convenient process of producing nanoparticles. It involves base catalyzed condensation of Tetraethylorthosilicate (TEOS) to form silica nanoparticles (Stober et al. 1968). Variation of physical parameters (stirring) and chemistry of solvent (water content) allows some size variation. This process can be applied repeatedly to form core-shell morphologies with desired chemistries while retaining a high degree of monodispersity (Hall et al. 2000). The 126

PAGE 143

127 addition of dye molecules allows the formation of fluorescent silica nanoparticles (FSNPs) (Vanblaaderen and Vrij 1992). 6.2.2 Nanoparticles applications Nanoparticles (NPs) have some properties that are an obvious product of their size that make them very useful. These include the large amount of surface area available, the large number of particles, and the small size of particles permitting ready access to small detail structures. Surface area scales as inverse square of size: meaning that as particles get smaller the specific surface area increases by the square of the size decrease factori.e. halving particle size will yield a fourfold increase in specific surface area. This large amount of available surface area provides the opportunity to do lots of surface chemistry. Since the surface of a particle is often the most convenient place to attach species of interest, the implication of a nanoparticle is that it can carry a large surface payload. Relative to a microsphere of 10 m diameter, a 100 nm diameter NP has 10,000 times the surface area available per unit mass of particles. Interesting applications for nanoparticles include many sensing and imaging methods that benefit from large numbers of particles to generate a significant response. Nanoparticles can be surface modified to generate specific binding, so that they attach to a known target. This type of scheme is used with magnetic particles for cell separation (Pankhurst et al. 2003), and with imaging techniques, where the particles can be optimized to provide contrast or response for various imaging modalities (Lanza et al. 2002; Flores et al. 2003; Kircher et al. 2003; Moffat et al. 2003).

PAGE 144

128 6.3 Materials and Methods 6.3.1 Materials Tetraethyl orthosilicate (TEOS, Aldrich 131903, [78-10-4]), Aminopropyl trimethoxy silane (APTS, Aldrich 281778, [13822-56-5]), Ammonium hydroxide (NH4OH 25-28 wt% aqueous, Aldrich 338818, [1336-21-6]), Ethylene triamine (triethylamine / Et3N, Aldrich 13,206-3, [121-44-2]), N-hydroxy succinimide (NHS, Aldrich 13,067-2 [6066-82-6]), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC, Aldrich 16,142-2, [25952-53-8]), Dimethyl sulfoxide (DMSO, Acros 12779 [67-68-5]), Fluorescein isothiocyanate (FITC, Molecular Probes F-143), folic acid (Fisher Bioreagents BP251925, [59-30-8]), ethanol (EtOH, various sources, [64-17-5]) and dimethylformamide (DMF, Aldrich 481343, [68-12-2]) were all used as purchased form the suppliers. All glassware was rigorously cleaned with HF and soap solution. De-ionized (DI) water was obtained from a Millipore system. 6.3.2 Methods 6.3.2.1 Method of conjugating FITC fluorophores to APTS Fluorescein isothiocyanate (FITC, see figure 4-2) is the isothiocyanate form of fluorescein, a fluorophore designed to be readily amine reactive for easy bioconjugation. APTS contains a primary amine group and reacts with FITC to form a stable thiourea linkage when the two are combined in ethanol, as is illustrated briefly in figure 6-1. Adapting a method of (Vanblaaderen and Vrij 1992), 138 mg APTS and 10.5 mg FITC were combined in 2 ml absolute ethanol under a dry nitrogen atmosphere and with light protection, and stirred magnetically for 24 h. This conjugate was stored refrigerated until used.

PAGE 145

129 FluoresceinNCS SiCH3CH2OOCH2CH3OCH2CH3CH2CH2CH2NH2 SiCH3CH2OOCH2CH3OCH2CH3CH2CH2CH2NHCNSFluoresceinH Figure 6-1: Conjugation of fluorescein isothiocyanate to APTS monomer. The isothiocyanate group on the fluorescein readily reacts with the terminal amine of the APTS. 6.3.2.2 Method of synthesizing fluorescent silica nanoparticles The basic method of synthesizing the silica nanoparticles was based on the Stber method (Stober et al. 1968), modified as described by Vanblaaderen and Vrij (Vanblaaderen and Vrij 1992) to incorporate the fluorophores FITC by covalent coupling into the silica network. The formation of the silica network with APTS is illustrated in figure 6-2. A 100 ml round bottom flask was arranged to provide magnetic stirring and was insulated from light. 2.4 ml NH4OH were combined with 30 ml absolute ethanol. After 5 minutes stirring, 1.2 ml TEOS was added. After 5 more minutes of stirring, 100 l of the FITC-APTS conjugate were added. After 24 h stirring, 245 l APTS were added to form a pure silica outer shell, and stirring was continued for another 24 h. An aminated surface was produced to enable surface immobilization chemistry by substituting 125 l each of APTS and TEOS for the 245 l TEOS to form the outer shell.

PAGE 146

130 OSiOOOSiCH2CH2CH2NH2SiSiSiOOOOOOOSiSiSi SiCH3CH2OOCH2CH3OCH2CH3OCH2CH3 SiCH3CH2OOCH2CH3OCH2CH3CH2CH2CH2NH2 TEOSAPTSin basic alcohol solutioncondensation reaction forms network shown and ethanol Figure 6-2: Formation of silica structure by base-catalyzed condensation, incorporating APTS. In the case of FITC conjugated APTS the fluorophore is attached as per figure 6-1 across the amine group of the APTS. The DDS particles were stored in the ethanol/ammonia mixture as this helped to prevent agglomeration. Washing occurred prior to surface immobilization and testing procedures, and involved centrifuging the particles in a high-speed centrifuge (>5,000 rpm), and replacing the supernatant with DI water, then redispersing by vortexing (30 s) and sonicating (5 min) alternately until the mixture appeared homogeneous. This was repeated 8 to 10 times. 6.3.2.3 Method of immobilizing folic acid onto the DDS nanoparticles Immobilization was performed by means of water-soluble carbodiimide (WSC) coupling chemistry. Two solutions were prepared separately and combined by stirring together for 20 min. Water washed, aminated DDS nanoparticles were suspended in DMSO: 20 mg in 1.0 ml, and activated with 20 l ethylene triamine to prepare solution 1.

PAGE 147

131 Solution 2 contained 1.0 ml 10 mM folic acid in DMSO, and 1.5 ml each of 15 nM NHS and 75 nM EDC in DI water. Each solution was stirred for 5 min, then the combined solutions were stirred together for 20 min, after which the nanoparticles were centrifuged and washed in DI water 5 times. These samples were stored away from light until use. 6.3.2.4 Transmission electron microscopy TEM images were acquired on a Hitachi H-7000 microscope. Samples were prepared by dropping DDS nanoparticles in ethanol onto a carbon coated copper grid. 6.3.2.5 Scanning electron microscopy Electron microscopy was performed on JEOL microscopes, either a 6400 or a 6335CF FEG-SEM. The samples were cast from either directly from native solution or from an alcohol solution onto SEM stubs and carbon coated for 2 min. 6.3.2.6 Zeta potential measurement Zeta Potential was measured by a Brookhaven Zeta Plus zeta potential analyzer and particle sizer. All measurements were made in PBS 7.4 pH. Samples were prepared at a concentration of approximately 1 mg/ml. 6.3.2.7 Light scattering particle size measurement Size data was gathered using a Coulter LS 230 and a Coulter IS 13 320 light scattering particle size analyzer. In each case the machine was equipped with a small volume fluid module. Particles could be used directly from their native solution, or as washed in water. A silica optical model was built for the particles and verified against SEM and TEM measurements.

PAGE 148

132 6.3.2.8 UV-visible absorption spectroscopy Absorption spectra were acquired using a Simadzu UV 2401-PC UV-Vis recording spectrometer. Water based samples were tested in PMMA cuvettes, while some solvent samples required the use of quartz cuvettes. 6.3.2.9 Fluorescence spectrometry Fluorescence spectra were obtained using a Fluorolog Tau-3 spectrofluorometer (Jobin Yvon Spex Instruments, S.A. Inc), capable of both steady state and lifetime measurements. Samples were prepared in DI water. 6.3.2.10 Confocal microscopy Confocal microscopy experiments utilized an Olympus IX81 microscope with Fluoroview software to assess dye distribution within the microspheres. Images were acquired using a default 20x objective, as well as a 60x oil-immersion objective (PLAPO 60XO). In each case optical and fluorescence images were acquired concurrently, with excitation provided by a 488nm argon laser for fluorescein labeled samples. 6.3.2.11 Cell experiments The cell lines and their treatments were identical to that described in chapter 4. The treatment of the samples was also similar. Briefly, 100 l of a 1 wt/vol % suspension of FSNPs was placed in a cell well with 2 ml PBS 7.4 pH. The cover slide with the cells grown on it was dipped into the cell well, and swirled for 60 s. The cover slide was then dipped in three successive rinses of 2 ml PBS 7.4 pH, in each of which it was also swirled for 60 s. The cover slide was then examined under the fluorescence microscope (Zeiss Axioplan 2, with Zeiss filter set #10 for fluorescein). Images were captured using Zeiss Axiovision 3 software that could also provide sizing information through the bundled scaling package.

PAGE 149

133 For experiments that involved the continued incubation of the cells after the application of the FSNPs, the suspension of FSNPs used was sterilized under UV light in the biological hood for 60 min. It had been previously determined that the UV lamp in the biological hood did not induce a significant loss of fluorescence from the FSNPs. 6.4 Results and Discussion The nanoparticles produced were characterized for size, fluorescence and surface chemistry. Since the aim within this research was to utilize the particles for a model system, control particles were made when appropriate to ensure that the effect investigated was truly due to a processing or treatment difference between the test sample and the control. 6.4.1 Size of FSNPs Figure 6-3 shows a TEM micrograph of the FSNPs. The particle size on the micrograph agrees well with the data acquired by the light scattering size measurement shown in figure 6-4, and with SEM pictures obtained. SEM micrographs verified the findings on nanoparticle size. These measurements were also confirmed by scanning electron microscopy (data not shown due to poor image quality, possibly due to charging). It was found that when larger size particles were detected by the sizing methods, it could be attributed to agglomeration. The native particles in basic alcohol solution did not agglomerate readily, but the washed particles in DI water agglomerated. Agglomeration behaviour was exacerbated by amination of the particles. The immobilization of folic acid to the surface of the FSNPs reduced the tendency for agglomeration displayed by aminated particles.

PAGE 150

134 Figure 6-3: TEM of dye doped silica nanoparticles, sample FITC2. 0.11100246810121416 Vol average size = 138 nmStd Deviation = 34 nm run 1 run 2 run 3Volume %Size (diameter) in um Figure 6-4: Coulter LS 230 graph showing size distribution of FITC2 FSNPs

PAGE 151

135 6.4.2 Fluorescence of the FSNPs The FSNPs were synthesized with a fluorescent core and a silica protective layer. In the case of aminated particles, the silica protective layer was composed half of APTS. Based on a size of 135 nm, and assuming that there was no loss of dye into solution, there should be approximately 2,500 FITC molecules contained in each FSNP. The small size of the FSNP permits the silica to be largely translucent as regards the fluorescence process, permitting excitation of, and fluorescence emission from each molecule. The outer silica (or aminated silica) shell protects the fluorophores from direct contact with the microenvironment. This is important as contact between the fluorophore and the environment can lead to oxidative degradationincreasing the rate of photobleaching. Lifetime measurements taken on a Fluorolog Tau-3 spectrofluorometer (data not shown) show the existence of two distinct lifetime states for the fluorophores. These lifetime states could be correlated to two microenvironments within the nanoparticlea hydrated silica shell, and a less hydrated core (Santra et al. 2004) It was confirmed by direct visual assessment and fluorescence microscopy that the particles were strongly fluorescent (using a handheld fluorescent lamp), enabling detection by fluorescence microscopy of sub milligram amounts of nanoparticles at 10 x magnification. 6.4.3 Folic Acid Immobilized FSNPs The FSNPs had folic acid immobilized to the surface via the free amine groups on the surface. The particles were extensively washed after immobilization to ensure that there was true covalent bonding of folate to the surface rather than a physical adsorption effect. The folate immobilized FSNPs were characterized by UV-Vis absorbance

PAGE 152

136 spectrometry (all in DI water medium), by comparing the absorption spectra of control particles from the same batch with the immobilized particles and folic acid. This data is presented in non-normalized graphical form in figure 6-5 below. Table 6-1 shows relevant peaks associated with the spectra shown. The absorption maxima appeared at 281 nm and 355 nm for the pure folic acid sample. The folate immobilized FSNPs showed absorption maxima at 280 nm with shoulders at 355 nm and 490 nm. The 490 nm shoulder is due to excitation maximum of the fluorescein molecules with which the FSNPs are doped, and appears identically on the control FSNPs. The control particles did not show any peaks or shoulders other than the fluorescein absorption, and a particle scattering based peak around 250nm. The presence of folate was thus confirmed on the immobilized FSNPs through the 280 nm absorption maxima and the 355 nm shoulder. Zeta potential analysis was also carried out on the FSNPs. Four types of FSNP surfaces were examined using a Brookhaven zeta plus zeta potential analyzer and particle sizer; all were tested at a concentration of approximately 1.0 mg/ml in 7.4 pH PBS. Pure silica NPs showed the most negative zeta potential value at .3 mV. The FITC doped nanoparticlesFSNPs, yielded a less negative zeta potential of .2 mV, indicating the presence of the amine groups from excess APTS that the FITC molecules had been conjugated to and were integrated into the structure. This is supported by the zeta potential of .2 mV measured for APTS/TEOS post-coated particles, which had an aminated surface. The measurement for folic acid immobilized FSNPs (folic acid was immobilized onto the same aminated surface as just detailed) showed a zeta potential of 6.1 mV. The immobilization of folic acid occurs across the -carboxyl of the glutamic

PAGE 153

137 acid end of folic acid, to an amine group on the particle. This consumes an amine group on the particle, and adds an aromatic amine with folic acid on the PABA end of the molecule, as well as a second carboxylate group on the glutamic acid end. The observed improved solubility of folate immobilized FSNPs may be due to the formation of a zwitterionic structure using amine and carboxylic acids. Figure 6-5: UV-Vis absorption spectra of folate immobilized FSNPs (top), 50 uM folic acid in solution (middle) and control FSNPs (bottom)

PAGE 154

138 Table 6-1: UV-absorption spectroscopy instrument output of absorption maxima for folic acid assay on FITC5 FSNPs whose spectral curves are shown in figure 6-3 Relevant Peaks for 2003 05 07b UV spec on Shimadzu UV 2401PC 200 to 600nm scan, slit width 1nm, scan interval 0.5nm FITC5 coupled to folic acid (yellow curve) Point P/V Wavelength nm. Abs. 1 Max 284.5 0.295 2 Max 252 0.298 3 Min 260 0.286 50 um folic acid in water (green curve) Point P/V Wavelength nm. Abs. 1 Max 344.5 0.149 2 Max 281.5 0.351 4 Min 339.5 0.148 FITC5 control sample (red curve) Point P/V Wavelength nm. Abs. 1 Max 252.5 0.208 4 Min 242 0.127 Note: peaks that were considered to be noise were omitted from this listing 6.4.4 Cell Experiments Cell experiments that were conducted using FSNPs demonstrated the same response in each case as those involving microspheres, as regards specific binding, but with a higher background of non-specifically bound nanospheres. Figure 6-6: Panel of FSNPs specifically bound to tumorous cells. Left image shows transmitted light image, right panel shows fluorescence image, same scale.

PAGE 155

139 Folate-immobilized FSNPs were suspended in water and added to a cell wells seeded with NCI-H23 human lung adenocarcinoma cells known to overexpress folate receptors (Franklin WA 1994). After several washes, fluorescence microscopy was performed to determine if folate immobilized FSNPs had bound specifically to the tumorous cells. Figure 6-6 shows strong evidence of specific binding. The subconfluent cell culture permits observation of individual cells, showing that each individual cell had been clearly labeled with the fluorescent nanoparticles. Figure 6-7: Optical (left panels) and fluorescence (right panels) confocal images of folate-immobilized FSNPs on BALB/3T3 fibroblasts. Bottom panels at higher magnification.

PAGE 156

140 Another line of cancerous cells, BALB/3T3 fibroblasts, was labeled with folate immobilized FSNPs and examined by confocal microscopy as shown in figure 6-7. It was found that these cells also generated strong specific binding of the nanoparticles. Additionally, confocal microscopy using Z-plane imaging demonstrated internalization of nanoparticles (and/or nanoparticle clusters). Control experiments using non folate-immobilized FSNPs showed that the controls generated a much lower degree of binding to cells. This suggests that the nanoparticles have a degree of non-specific binding to cells and a high degree of folate mediated specific binding. Folate-immobilized and control FSNPs are shown in figure 6-8, on NCI-H23 cells. In this particular image, the exposure time of the control image was longer, so the intensity is not directly comparable, but is actually significantly larger than is visually apparent. The greater coverage of the cells by folate-immobilized particles can be clearly seen. Figure 6-8: NCI-H23 cells treated with FSNPs. Left panel shows the control non folate immobilized sample, right panel shows folate-immobilized particles.

PAGE 157

CHAPTER 7 CONCLUSIONS AND FUTURE WORK 7.1 Magnetic Microsphere Preparation and Characterization We were able to produce novel microspheres that were superparamagnetic to enable magnetic separation and retrieval, surface functionalized to enable receptor binding, and dye labeled or dye releasing. This fulfilled the design requirements for the application that were defined in chapter 3. We found that standard methods could be adapted to produce these microspheres. Suspension polymerization was selected as the method of preparation that best suited the design requirements generated for the application in chapter 3. Microspheres doped with superparamagnetic iron oxide were produced, that enabled magnetic separation of the microspheres from suspension. The required surface chemistry was generated by copolymerization with an aminated monomer. The size range of microspheres produced was adjusted through the stirring rate during polymerization. Microspheres with volume average sizes of under 10 m were produced. Future work on microsphere preparation would aim at producing more monodisperse particle size distributions. This would decrease the light scattering by particles and permit easier characterization of the product by spectrophotometric methods. Activated swelling processes may offer the best chances of producing monodisperse microspheres with the properties and chemistry desired. Seed microspheres would be swelled with a monomer mixture that included an aminated 141

PAGE 158

142 monomer. In situ precipitation of iron oxide would be a viable method for incorporation of magnetically responsive material. 7.2 Ligand Immobilization Folic acid was attached to microspheres with preservation of receptor recognition. Standard carbodiimide-mediated bioconjugation techniques were adequate to immobilize folic acid onto the microspheres via carboxyl groups on the glutamic acid end of folic acid. Attachment of folic acid onto the microspheres was difficult to characterize, so a fluorescent silica nanoparticle model system with analogous surface chemistry was utilized to demonstrate immobilization effectiveness. It was shown by means of UV-absorption spectroscopy that the nanoparticles had folate immobilized onto their surface. Future work with ligand immobilization would build easily upon the research done, because the chemistry utilized is general and applicable to a wide range of ligands. This presents a ready template for building further applications. 7.3 Microsphere Labeling Incorporation of dye into microspheres by both swelling and covalent binding was investigated, and it was determined that the two methods offer fundamentally different advantages. Improved spatial resolution that is offered by covalently incorporated dye molecules. Spatial resolution is critical in development work, as direct observation of a system allows perception of interactions that are not obvious when only a quantification is elicited as a response. The swelling method permits removal of the dye payload from the microspheres, enabling more sensitive fluorometric methods for ratiometric assessment of recovered microspheres. This is because the microspheres themselves interfere with spectrofluorometry by light scattering effects that increase noise to signal ratio. Qualitative development work favours covalent labeling, while quantitative

PAGE 159

143 evaluation (the ultimate goal) favours dye loading by swelling and subsequent dye recovery. The present research focused on providing a proof of concept rather than optimizing applicational parameters. Future work would focus more on the swelling method of dye incorporation. This would require optimization of dye loading and recovery to correlate with the ratios of microspheres recovered. 7.4 Cell Line Testing Binding of microspheres to cells using the immobilized folic acid was demonstrated. Cell lines that were known to express folate receptors at elevated rate were used for testing. Most results came from NCI-H23 human lung adenocarcinoma cells. Retention rates of folate-immobilized microspheres on cell lines expressing folate receptors were demonstrated to be statistically significantly greater than for control microspheres, based on pooled data of all microspheres tested on the cell line. The p-value generated was 0.0074. This demonstrated the true disease state (true positive) side of the results matrixgiving the result shown in figure 7-1. Normal human dermal fibroblasts were used as the normal cell line providing the disease negative side of the results matrix. The control cell line proved to retain similar numbers of folate-immobilized microspheres as the cancerous cell line generating a p-value of 0.90 when testing for significance between the NCI-H23 cell line and the normal human dermal fibroblast cell line with folate-immobilized microspheres. This shows that the system in its present state yields false positives, being not able to distinguish between the cancerous and normal cell line. This could be due to non-specific binding to cells that do not

PAGE 160

144 express folate receptors, or it could be due to the control cell lines expressing folate receptors. This issue must be addressed in future research. CELL LINECancerousNormalTruepositiveFalsepositiveFalse negativeTrue negativeabcdMICROSPHEREScontrolsfolate immob. Figure 7-1: Results matrix for testing on cell lines. The expected outcome for a system working perfectly would be a positive response only for a, with negatives in b,c and d. Future work on cell line testing should pursue several points that provided difficulty in the present study. Interference caused by cell autofluorescence (the same applies for tissue samples) can be addressed by the use of a confocal microscope for spatial analysis. The confocal microscope uses a grating and can collect a narrower, more sharply defined wavelength range. The fluorescence microscope uses filters that collect an area under the curve, which allows more background and autofluorescence relative to the signal. Collecting at the precise target wavelength would permit better signal discrimination. The cell lines that were used in this study were selected based on literature references and supplier specifications, and of course by availability. The cell lines best suited to future research would be KB cells as a disease positive, and CHO (Chinese

PAGE 161

145 hamster ovary) cells as a control line, along with the already initiated oral squamous carcinoma cell line for tissue-specific cell testing. 7.5 Tissue Testing Tissue testing with both tumour tissue and normal control tissue was performed. Tumourous tissue was freshly excised human oral squamous cell carcinoma tissue. Control tissue was normal neck epithelial tissue. It was found that the folate-immobilized microspheres were retained by tumorous tissue, while the control microspheres were not, generating a p-value of 0.037 when comparing the two groups for difference. The difference in values of retained microspheres was about an order of magnitude for the two groups, demonstrating a meaningful difference for application purposes. It was found that folate immobilized microspheres were also retained by the normal tissue samples, at an even higher rate than by the cancerous tissue sample. This results in the same results matrix for tissue testing as shown in figure 7-1 for the cell line testing. Future work with tissue should include characterization of the tissue for receptors. Results generated with folate-immobilized microspheres could be compared to a map of folate receptors resident on the tissue. Such a map could be made by staining the tissue with immunohistochemical agents sensitive to folate receptors. Unfortunately, the antibody available for this application is MOV18, which is sensitive only to the alpha receptor, while the beta receptor may be the primary receptor expressed by HNSCC. Immunohistochemical solutions should be pursued, since the desirability of working with a completely characterized system is indisputable. 7.6 Microsphere Retrieval It was found that the microspheres could be retrieved magnetically from their bound state on tissue and cells. It was found that the acid wash that was envisioned was

PAGE 162

146 not successful. The magnetic retrieval method is adequate for the requirements of study, and could be made to work in an application as envisioned. Future work on microsphere retrieval can take several directions. The magnetically retrieved microspheres can be imaged by electron microscopy to determine if tissue or cells are being removed with the microspheres allowing more insight into the nature of the binding that is taking place between folate-immobilized microspheres and cells/tissue. If the binding that is taking place is indeed the desired folate receptor to folate binding then a release solution with increased salt concentrations and/or with a high folate content could aid in causing release from the receptors. Another approach to improving retrieval would involve microsphere size: increasing microsphere size would reduce the specific surface area available for binding reactions while concurrently increasing the per-microsphere magnetic content enabling easier magnetic retrieval.

PAGE 163

APPENDIX A POLYMERIZATION RECORDS FOR SELECTED SAMPLES THAT APPEAR IN THE MANUSCRIPT This appendix provides detailed records on the polymerization procedures and materials for samples that are detailed in the manuscript body. The volume of records prohibits inclusion of all sample records for the polymerizations performed. The coding system is sequential and by type. Each polymerization may have one or more parameters that were varied between individual samples. In this case details are provided for each sample. Dxxx polymerizations are dispersion polymerizations. ASxx polymerizations are activated swellings. For each activated swelling batch there are seed particles that were produced by dispersion polymerization. These were almost always pure styrene particles with no crosslinker content. Sxx polymerizations are suspension polymerizations. This procedure was selected and optimized for the application. List of polymerization records in Appendix A 1. D009 2. D013 3. D030 4. AS05 5. AS08 6. S11 147

PAGE 164

148 7. S12 8. S18 9. S19 10. S21 11. S22 12. S23 13. S24 14. S25

PAGE 165

149 Polymerization Records Identifier: Dispersion 9: #1 through 3 Date polymerized: 11/8 Monomers: ST/DEA Solvent: Methanol/Water Composition (polymers): 69/30/1 ST/DEA/DVB Component % (w/v or v/v) Source amount used Monomer 1 Styrene Acros 13279-0010, lot B0070337, 69% (vol) distilled 10/18, opened 6/98 690ul Monomer 2 DEA, Aldrich 40,898-0, lot 13126DG 30% (vol) distilled 10/19, opened '96 300ul Crosslinker DVB, Sigma 1% (vol) NaOH washed 10/23 10ul Initiator AIBN, Sigma, 11/00 0.3% of whole rexlzd in MeOH 11/6 30mg Other Procedure Made batch of monomers, then dissolved AIBN in this. Solvent phase:Methanol/Water: 70/30 components %(w/v or v/v) source amount used Primary solvent Methanol, Fisher A452-4 70% (vol) Shelf, 3/00 Co-solvent Water 30% (vol) distilled Stabilizer 1 PVP 40: Sigma lot 270-0062 3% (vol) dissolved in methanol, then filtered 0.2um 3g for 100ml solvent stock solution Co-stabilizer Surfactant Other Procedure Made stock solution of solvent: 70/30, added to monomers, 9ml. Vortexed and Nitrogen purged. Polymerization Procedure: Gas Purged Apparatus: Rotating air incubator Time in: 1230 Time out: 1200 11/9 for 1&2 Polymerization time: 23.5h for 1&2 Polymerization Temp: 55C or Temp Profile: Agitation: rate: 50=4rpm type: rotation

PAGE 166

150 Post Polymerization: Observations on samples: The substance produced was milky, sort of vicous, with some collection of 'skin' at interfacial surfaces. Cleaning media: distilled water Time:1h Other notes: Results: Characterization: light microscope showed definite presence of microspheres though no size range could be established. A good dilution for viewing appears to be approximately .05% microsphere content (theoretical). We used 0.5ml of solution directly from stirring in distilled water (sample in 100ml water), and added that to 10ml water.

PAGE 167

151 Polymerization Records Identifier: Dispersion 13: #1 through 6 Date polymerized: 11/27/00 Monomers: ST/DEA/DVB Solvent: Methanol/ Water w. ~1.75% Fe2O3 / PVP40 Composition (polymers): 69/30/1 ST/DEA/DVB Component % (w/v or v/v) Source amount used Monomer 1 Styrene Acros 13279-0010, lot B0070337, 69% (vol) distilled 11/14 #1-2: 690ul #3-6: 345ul Monomer 2 DEA, Aldrich 40,898-0, lot 13126DG 30% (vol) distilled 11/14, opened '96 #1-2: 300ul #3-6: 150ul Crosslinker DVB, Sigma 1% (vol) NaOH washed 10/23 #1-2: 10ul #3-6: 5ul Initiator AIBN, Sigma, 11/00 0.3% of whole rexlzd in MeOH 11/15 #1-2: 30mg #3-6: 15mg Other Procedure Made 1 large batch of monomers, then dissolved AIBN in this. Afterwards divided into individual vials: 1ml into 1,2 0.5ml into 3-6. Solvent phase: Methanol/Water: 70/30, premixed from 11/16 with 3% PVP40 for 1, 3; stock solution of 70/30 MeOH/water with 1.5%PVP made 11/19 for #5. Made 70/30 with half of water replaced with Fe2O3 (3.5% by wt), 3% PVP40 for #2,4; same with 1.5% PVP40 for #6 components %(w/v or v/v) source amount used Primary solvent Methanol, Fisher A452-4 70% (vol) Shelf, 3/00 Co-solvent Water 30% (vol) distilled Stabilizer 1 PVP 40: Sigma lot 270-0062 3% (w/v) 1.5% (w/v) dissolved in methanol, then filtered 0.2um 3%, for 1-4 1.5% for 5,6 Co-stabilizer Surfactant Other 20mM Na2HPO4 20mM in the iron oxide solution as stabilizer Nakato shelf #2,4 have 15% of solvent phase, #6 has 7.5%

PAGE 168

152 Procedure Used stock solution made 11/19 for dispersion 012 for 1 and 3. Separate mixture of same date for 5. Always 10ml reaction mix. 1 and 2 were 90/10 solvent/monomers and #3-6 were 95/5 solvent/monomers. Samples 1-4 were made with 3% stabilizer (relative to entire mass), while 5 and 6 were made with 1.5% PVP40. Polymerization Procedure: Gas Purged Apparatus: Rotating air incubator Time in:1900 Time out:1930 11/28 Polymerization time:24h Polymerization Temp: 65 or Temp Profile: Agitation: rate: 50=4rpm type: rotation Post Polymerization: Observations on samples: Before polymerization the iron oxide containing samples already appeared to have a slightly grainy texture. After polymerization the non-iron oxide samples had the characteristic appearance (mily slightly viscous, coconut milk smell), while the iron oxide containing samples were completely agglomerated black balls within a relatively white fluid. No interaction at all between the two! Cleaning media: distilled water Time: 1h Other notes: Results: Characterization:

PAGE 169

153 Polymerization Records Identifier: Dispersion 30: #1-5 Date polymerized: 6/26/01 Monomers: ST/DEA/DVB Solvent: Ethanol/water, 80/20 to 90/10 Composition (polymers): 69/30/1 ST/DEA/DVB, [M] = 5% for all samples Component % (w/v or v/v) Source amount used Monomer 1 Styrene Acros 13279-0010, lot B00500448, 69% (vol) distilled 4/11/01 1380ul Monomer 2 DEA, Aldrich 40,898-0, lot 13126DG 30% (vol) distilled 1/16/01, from bottle of '97 600ul Crosslinker DVB, Sigma D-0916 lot 50K3652 1% (vol) NaOH washed 01/16/01 20mg Initiator AIBN, Sigma, 11/00 3% of monomer rexlzd in MeOH 6/01 60mg Other Procedure Made 1 large batch of monomer: 2ml, then dissolved AIBN in this. Afterwards divided into individual screw top vials: 250ul in each. Solvent phase: Ethanol/Water: var from 80/20 to 90/10 components %(w/v or v/v) source amount used Primary solvent Ethanol, Fisher unknown, denatured 80-90 vol% Shelf, 3/00 1: 80/20 2: 82.5/17.5 3: 85/15 4: 87.5/12.5 5: 90/10 Co-solvent Water 20-30 vol% distilled see above Stabilizer 1 PVP40 Stock from 2/6 12% / [M] = 0.030g / sample with 250ul M Co-stabilizer Surfactant Other Procedure Polymerization Procedure: Gas Purged

PAGE 170

154 Apparatus: Water shaker bath: Haake SWB25 Time in:1600 Time out: 1330 Polymerization time:21.5h Polymerization Temp: 60C or Temp Profile: Agitation: rate: n=75 type: shaking vials horizontal inside 50ml centrifuge vials Post Polymerization: Observations on samples: Cleaning media: distilled water Time:1 h Other notes: Results: The samples all had no agglomerations very nice appearance. The higher numbered samples (less water content in solvent) all had a stronger residual smell of monomer may have to either increase time for polymerization or this may be the limit on the window of opportunity for polymerization by dispersion. Characterization:

PAGE 171

155 Polymerization Records: activated swelling Identifier: AS05 Date polymerized: 2/18/02 Composition: 50/50 ST/DVB Seed particles: ST Composition 100 % ST Batch D052_3 Size data Vol avg size = 1.24 um 0.16 sd, PDI = 1.05 Other Procedure prepared 60 mg of particles in 1.66 ml water Swelling emulsion: components %(w/v or v/v) source amount used Primary solvent water DI 15 ml swelling agent DBP [84-74-2] Acros 16600 lot B0072055 200 ul surfactant / em SDS [151-21-3] 0.25 wt % Aldrich 85,192-2 38 mg Other Procedure made emulsion of water and SDS, then sonicated 10 min, added DBP and sonicated again, then combined with seed particles, sonicated a few mintues and left stirring 24 h, in refrigerator at 4 C Monomer phase components %(w/v or v/v) source amount used Primary solvent water 15 ml Monomer 1 DVB 50 vol % Washed 10/25 320 ul Monomer 2 ST 50 vol % distilled 10/29 320 ul Crosslinker Initiator BPO 40 mg surfactant / em SDS [151-21-3] 0.25 wt % Aldrich 85,192-2 38 mg Other PVP40 PVA, 85-146K MW, 88% hydr 10 wt % sol 10 wt % sol PVP 40 Airvol 523 1.5 ml in 1/2 1.5 ml in 1/2 Procedure 15 ml water + SDS, sonicated together. Initiator was dissolved in monomers seperately used slight excess of monomers due to extra transfer step. Then sonicated (whole) to disperse as emulsion, then mixed (stirring magn) for 24 h with emulsion containing swollen seed particles. Added 3 ml PVP40 (10%) solution as stabilizer, stirred 15 min, put into 2 40 ml vials, sonicated, put on to polymerize at 70 C.

PAGE 172

156 Polymerization Procedure: Gas Purged Apparatus: SWB 25 Time in:1230 Time out:1230 Polymerization time:24 h Polymerization Temp: 70 C or Temp Profile: Agitation: rate: 100 cpm type: shaking Post Polymerization: Observations on samples: milky, with not much agglomerated mass Cleaning media: filled vials with water and sonicated 20 min Time: Other notes: Results: appeared fine Characterization: Coulter DLS; 1: Vol avg size = 2.72 +/0.70, PDI 1.1 2: vol avg size 2.74 +/0.72 PDI 1.11

PAGE 173

157 Polymerization Records: activated swelling Identifier: AS08 Date polymerized: 6/10/02 Composition: HEMA / ST / EGDMA Seed particles: ST Composition 100 % ST Batch D061_2, measured at 2.5 wt % dry ptcles washed Size data Vol avg size = 1.43 um 0.19 sd, PDI = 1.053 Other D061_2 washed with DI water 3x Procedure 6.4 ml solution used to get 160 mg solids. Swelling emulsion: components %(w/v or v/v) source amount used Primary solvent water DI 15.6 ml to retain conc swelling agent DBP [84-74-2] Acros 16600 lot B0072055 160 ul = 1:1 DBP:seeds surfactant / em SDS [151-21-3] 0.25 wt % Aldrich 86,201-0 50 mg Other Procedure made emulsion of water and SDS, vortexed, then sonicated 10 min, repeated; added DBP, sonicated and vortexed again; combined with seed particles, sonicated briefly and left stirring 24 h on stir plate. Monomer phase components %(w/v or v/v) source amount used Primary solvent water 20 ml Monomer 1 HEMA 1/3 from refrig 400 ul Monomer 2 ST 1/3 distilled 10/29 400 ul Crosslinker EGDMA 1/3 from refrig 400 ul Initiator BPO from refrig 60 mg surfactant / em SDS [151-21-3] 0.25 wt % Aldrich 85,192-2 50 mg Other PVP40 10 wt % sol PVP 40 1.5 ml in each Procedure 20 ml water + SDS, sonicated together. Initiator was dissolved in monomers seperately. Sonicated with water mix to make emulsion, added activated particles, mixed (stirring magn) for 24 h. Added 3 ml PVP40 (10%) solution as stabilizer, stirred 15 min, put into 2 40 ml vials, sonicated, put on to polymerize at 70 C. Polymerization Procedure: Gas Purged

PAGE 174

158 Apparatus: SWB 25 Time in:1600 Time out:1600 Polymerization time:24 h Polymerization Temp: 70 C or Temp Profile: Agitation: rate: 100 cpm type: shaking Post Polymerization: Observations on samples: milky, with some agglomeration at bottom Cleaning media: filled vials with water and sonicated 60 min Time: Other notes: Results: appeared fine Characterization: trimodal size distribution by Coulter w number avg size of 2.5 um for water washed samples

PAGE 175

159 Suspension Polymerization Record Identifier S11 Date started 5/22/03 Monomers used magn doped MMA (PDMS laced) and maleic anhydride Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Airvol 523 lot 03011959, Mw 85-146k stock 2 % solution, filtered: 25 um paper Other Other NaNO3 sol. phase polym inhibitor 0.01 w/v % = 10 mg Other Processing: Heated solution in 3 neck flask w stirrer and thermometer, used premade Monomer phase: MMA (magn doped) and maleic anhydride Component Amount Seeds / dopant: magnetite: oleate and pdms coated 5.8 wt % before pdms doping prob 5 wt % magnetite. resusp in MMA from chloroform Monomer 1: MMA: distilled 2/03 4 ml, magn doped, 80 vol % Monomer 2: Maleic Anhydride Dajac 7579 1000 mg ground from birquettes Monomer 3: Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 375 mg, 7.5 w/v % Processing: Injected w 25G needle at 500 rpm stirring Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: 500 rmp mechanical Gas purge: Temp: 75 C Time: 1600 5/22 -> 2200 5/22 = 6 h Observations: looked brownish reddish no obvious agglomerations Post processing: washed in water Characterization: 210 um vol avg, with some small fraction.

PAGE 176

160 Suspension Polymerization Record Identifier S18 Date started 9/25/03 Monomers used magn doped MMA (PDMS laced) and AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Airvol 523 lot 03011959, Mw 85-146k stock 2 % solution, unfiltered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 65 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 2/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 1000 mg as solid form Monomer 3: Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1020 9/25 -> Observations: Post processing: Characterization:

PAGE 177

161 Suspension Polymerization Record Identifier S19 Date started 10/22/03 Monomers used magn doped MMA (PDMS laced) and FITC-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 65 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 2/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 1000 mg as solid form, heated to mix with fitc conjugated AEMHS (20 mg each) made 9/03 Monomer 3: Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1200 10/24 -> 1600 10/24 Observations: Post processing: Characterization:

PAGE 178

162 Suspension Polymerization Record Identifier S21 Date started 1/13/04 Monomers used magn doped MMA (PDMS laced) and FITC-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 70 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 2/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 900 mg as solid form, heated to mix with fitc conjugated aemh Monomer 3: AEMHS conj with FITC 2mg FITC in 100 mg AEMHS, conj in 2ml EtOH, solvent evap off at red press % T Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1045 1/13 ->1715 1/13 = 6.5 h Observations: latte colour. Post processing: Characterization:

PAGE 179

163 Suspension Polymerization Record Identifier S22 Date started 1/14/04 Monomers used magn doped MMA (PDMS laced) and FITC-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 70 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 7/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 900 mg as solid form, heated to mix with fitc conjugated aemh Monomer 3: AEMHS conj with FITC 5mg FITC in 100 mg AEMHS, conj in 2ml EtOH, solvent evap off at red press % T Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1200 1/14 ->1845 1/14 = 6.7 h Observations: latte colour. Post processing: Characterization:

PAGE 180

164 Suspension Polymerization Record Identifier S23 Date started 1/15/04 Monomers used magn doped MMA (PDMS laced) and FITC-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 70 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 7/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 900 mg as solid form, heated to mix with fitc conjugated aemh Monomer 3: AEMHS conj with FITC 10mg FITC in 100 mg AEMHS, conj in 2ml EtOH, solvent evap off at red press & T Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 0900 1/15 ->1600 = 7 h Observations: latte colour. Post processing: Characterization:

PAGE 181

165 Suspension Polymerization Record Identifier S24 Date started 1/16/04 Monomers used magn doped MMA (PDMS laced) and FITC-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 70 C Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA), 6/03 Monomer 1: MMA: distilled 7/03 4 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 900 mg as solid form, heated to mix with fitc conjugated aemh Monomer 3: AEMHS conj with FITC 20mg FITC in 100 mg AEMHS, conj in 2ml EtOH, solvent evap off at red press & T Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 125 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 125 ul, 2.5 v/v % Processing: Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1215 1/16 ->2045 = 8.5 h Observations: latte colour. Post processing: Characterization:

PAGE 182

166 Suspension Polymerization Record Identifier S25 Date started 2/20/04 Monomers used magn doped MMA (PDMS laced) and Texas RedX-conjugated AEMHS Suspension medium Water: 2 % aqu PVA Protocol Shim / Kim adapted protocol Suspension phase: Water: 2 % aqu PVA sol: heated Component Amounts Medium: water 100 ml Stabilizer: PVA 88% hydrol, [9002-89-5] Celvol 823 used 2 w/v %, not filtered Other Other Other Processing: Heated solution in 3 neck flask w stirrer and thermometer 70 C; Monomer phase: MMA (magn doped) and AEMHS Component Amount Seeds / dopant: magnetite: oleate and pdms coated 6 % pdms coated MMA (dil w MMA) Monomer 1: MMA: distilled 7/03 2 ml, magn doped, 80 vol % Monomer 2: AEMHS, Acros 357810250, [2420-94-2]) 300 mg as solid form, heated to mix with fitc conjugated aemh Monomer 3: AEMHS conj with FITC 5 mg Texas RedX in 200 mg AEMHS, conj overnight in 2ml THF (stirring), solvent evap off on ice Initiator: AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1] recrystallized 2/10/03 62.5 mg, 2.5 w/v % Crosslinker: DVB, Sigma D-0916 lot 50K3652, [1321-74-0], NaOH washed 10/29/02 63 ul, 2.5 v/v % Processing: Monomer conjugate mixture turned intense salmon colour, with slight lumps apparent. Was able to dissolve apparently evenly in monomers by mixing with warm suspension fluid. Injected w large gauge needle, stirring with homogenizer on setting low 4 Polymerization: Apparatus: 3 necked vessel, w heating and stirring Agitation: homogenizer, 4 Gas purge: Temp: 75 C Time: 1200-1800 = 6 h Observations: turned bright salmon colour in fluid

PAGE 183

167 Post processing: Drained fluid off. Suspension fluid was obvious pinkish colour definitely absorbed some dye. When settling on magnetic separator, there was also a layer of white-pink polymer? probably some MMA and dye polymerized without magnetite.

PAGE 184

APPENDIX B CELL LINE DATA The cell lines utilized are listed below; NCI-H23; ATCC code CRL-5800 Human lung adenocarcinoma cells BALB/3T3; ATCC code CCL-163 Mouse fibroblasts cells NHDF from Clonetics Corporation of Cambrex Life Sciences Normal human dermal fibroblasts SCC-9, ATCC code CRL-1629 Oral squamous cell carcinoma 168

PAGE 185

169 ATCC Number: CRL-5800 Price: $179.00 Designation: NCI-H23 [H23] Depositors: AF Gazdar JD Minna Biosafety Level : 1 Shipped: frozen Medium & Serum: See Propagation Growth Properties: adherent Organism: Homo sapiens (human) Morphology: epithelial Tissue: lung; adenocarcinoma; non-small cell lung cancer Permits/Forms: In addition to the MTA mentioned above, other ATCC and/or regulatory permits may be required for the transfer of this ATCC material. Anyone purchasing ATCC material is ultimately responsible for obtaining the permits. Please click here for information regarding the specific requirements for shipment to your location. Related Cell Culture Products Comments: This line was derived from a lung cancer obtained from a patient prior to therapy. The cells carry the K-ras 12 mutation, and there is a mutation in codon 246 (ATC -> ATG, isoleucine -> methionine) of the p53 gene. There is expression of C-myc, L-myc, v-src, v-abl, v-erb B, c-raf 1, Ha-ras, Ki-ras and N-ras RNAs. The cells express heterogeneous mRNA expression for PDGF A and B chain, transforming growth factor alpha and beta and the epidermal growth factor receptor (EGFR). NCI-H23 exhibits a high degree of c-myc DNA amplification (20fold) but no detectable amplification of c-myc RNA. The cells stain positive for keratins 5+8 and 18 and vimentin but are negative for neurofilament. NCI-H23 cells are L-dopa decarboxylase-negative.

PAGE 186

170 They have a reported colony forming efficiency of 9.7% in soft agarose. For additional cell line information see: Data from NCI Human Tumor Cell Lines Database The line is available with the following restrictions: 1. This cell line was deposited at the ATCC by Dr. A. Gazdar and Dr. J. Minna and is provided for research purposes only. Neither the cell line nor products derived from it may be sold or used for commercial purposes. Nor can the cells be distributed to third parties for purposes of sale, or producing for sale, cells or their products. The cells are provided as service to the research community. They are provided without warranty of merchantability or fitness for a particular purpose or any other warranty, expressed or implied. 2. Any proposed commercial use of the these cells, or their products must first be negotiated with the University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd., Dallas, Texas 75235. Telephone (214) 699-8056, FAX (214) 688-7233. Oncogene: myc +; src +; abl +; erb +; ras +; sis DNA Profile (STR): Amelogenin: X CSF1PO: 10 D13S317: 12 D16S539: 11 D5S818: 12,13 D7S820: 9,10 TH01: 6 TPOX: 8,9 vWA: 16,17 Age: 51 years Gender: from male organism(s) Ethnicity: Black Propagation: ATCC medium: RPMI 1640 medium with 2 mM L-glutamine adjusted to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, and 1.0 mM sodium pyruvate, 90%; fetal bovine serum, 10% Temperature: 37.0 C Subculturing: Remove spent medium, and add fresh 0.25% trypsin, 0.03% EDTA solution for 2 to 3 minutes at room temperature. Remove the trypsin and incubate the flask at 37C for 5 to 10 minutes or until the cells detach. Add fresh medium, aspirate and dispense into new flasks. Subcultivation Ratio: A subcultivation ratio of 1:3 to 1:6 is recommended Medium Every 2 to 3 days

PAGE 187

171 Renewal: Freeze Medium: Culture medium, 95%; DMSO, 5% Doubling Time: 38 hrs Related Products: Recommended medium (without the additional supplements or serum described under ATCC Medium) ATCC No: 30-2001 recommended serum ATCC No: 30-2020 References: 1805 : Little CD et al. Amplification and expression of the c-myc oncogene in human lung cancer cell lines. Nature 306: 194-196, 1983. PubMed: 6646201 1806 : Takahashi T et al. p53: A frequent target for genetic abnormalities in lung cancer. Science 246: 491-494, 1989. PubMed: 2554494 22403 : Mitsudomi T et al. p53 gene mutations in non-small-cell lung cancer cell lines and their correlation with the presence of ras mutations and clinical features. Oncogene 7: 171-180, 1992. PubMed: 1311061 22434 : Brower M et al. Growth of cell lines and clinical specimens of human non-small cell lung cancer in a serum-free defined medium. Cancer Res. 46: 798-806, 1986. PubMed: 3940644 22465 : Broers JL et al. Spontaneous changes in intermediate filament protein expression patterns in lung cancer cell lines. J. Cell Sci. 91: 91-108, 1988. PubMed: 2473086 22868 : Forsberg K et al. Expression of functional PDGF beta receptors in a human large-cell lungcarcinoma cell line. Int. J. Cancer 53: 556-560, 1993. PubMed: 8382192 23036 : Gazdar AF et al. Establishment of continuous, clonable cultures of small-cell carcinoma of lung which have amine precursor uptake and decarboxylation cell properties. Cancer Res. 40: 3502-3507, 1980. PubMed: 6108156 23570 : NCI-Navy Medical Oncology Branch Cell Line Supplement. J. Cell. Biochem. suppl. 24: 1996. 24389 : Lung Cancer 4: 155-161, 1988.

PAGE 188

172 ATCC Number: CCL-163 Price: $179.00 Designation: BALB/3T3 clone A31 Depositors: S Aaronson Biosafety Level : 1 Shipped: frozen Medium & Serum: See Propagation Growth Properties: adherent Organism: Mus musculus (mouse) Morphology: fibroblast Tissue: embryo; fibroblast Permits/Forms: In addition to the MTA mentioned above, other ATCC and/or regulatory permits may be required for the transfer of this ATCC material. Anyone purchasing ATCC material is ultimately responsible for obtaining the permits. Please click here for information regarding the specific requirements for shipment to your location. Related Cell Culture Products Comments: The BALB/3T3 clone A31 is one of several cell lines (see ATCC CCL-164, BALB/3T12) developed by S.A. Aaronson and G.T. Todaro in 1968 from disaggregated 14to 17-day-old BALB/c mouse embryos. [22708] The cells are extremely sensitive to contact inhibition of cell division, grow at a high dilution, exhibit a low saturation density and are highly susceptible to transformation in tissue culture by the oncogenic DNA virus SV40 and murine sarcoma virus. [26022] The serum used is calf serum, NOT fetal calf serum. The depositor recommends calf serum because fetal calf serum causes transformation and loss of contact inhibition. The serum initially employed and found satisfactory was from the Colorado Serum Co. Tested and found negative for ectromelia virus (mousepox).

PAGE 189

173 Virus Susceptibility: Herpes simplex virus; Vesicular stomatitis Indiana virus Virus Resist: Poliovirus 1 Tumorigenic: No, the cells were not tumorigenic in immunosuppressed mice, but did form colonies in semisolid medium. Reverse Transcript: negative Cytogenetic Analysis: modal number = 78; range = 62 to 109. The stemline number is hypotetraploid. Most of the cells contained only telocentric or acrocentric chromosomes. Strain: BALB/c Age: embryo; 14 to 17 day gestation Propagation: ATCC medium: Dulbecco's modified Eagle's medium with 4 mM L-glutamine adjusted to contain 1.5 g/L sodium bicarbonate and 4.5 g/L glucose, 90%; bovine calf serum, 10% Temperature: 37.0 C Subculturing: Never allow cultures to become completely confluent before subculture. 1. Remove and discard culture medium. 2. Briefly rinse the cell layer with 0.25% (w/v) Trypsin0.53 mM EDTA solution to remove all traces of serum that contains trypsin inhibitor. 3. Add 2.0 to 3.0 ml of Trypsin-EDTA solution to flask and observe cells under an inverted microscope until cell layer is dispersed (usually within 5 to 15 minutes). Note: To avoid clumping do not agitate the cells by hitting or shaking the flask while waiting for the cells to detach. Cells that are difficult to detach may be placed at 37C to facilitate dispersal. 4. Add 6.0 to 8.0 ml of complete growth medium and aspirate cells by gently pipetting. 5. Add appropriate aliquots of the cell suspension to new culture vessels. 6. Incubate cultures at 37C. Subcultivation Ratio: For 60mm plates, use an inoculum of 3 X 10(5) cells per plate and subculture every 3 days. Medium Renewal: Twice per week Freeze Medium: Complete growth medium supplemented with 5% (v/v) DMSO Related Products: Recommended medium (without the additional supplements or serum described under ATCC Medium) ATCC No: 30-2002 recommended serum ATCC No: 30-2030 References: 22708 : Aaronson SA Todaro GJ Development of 3T3-like lines from Balb-c

PAGE 190

174 mouse embryo cultures: transformation susceptibility to SV40. J. Cell. Physiol. 72: 141-148, 1968. PubMed: 4301006 26022 : Todaro GJ Aaronson SA Properties of clonal lines of murine sarcoma virus transformed Balb-3T3 cells. Virology 38: 174-202, 1969. PubMed: 4306523 26023 : Aaronson SA Todaro GJ Basis for the acquisition of malignant potential by mouse cells cultivated in vitro. Science 162: 1024-1026, 1968. PubMed: 4301647 26024 : Jainchill JL Todaro GJ Stimulation of cell growth in vitro by serum with and without growth factor. Relation to contact inhibition and viral transformation. Exp. Cell Res. 59: 137-146, 1970. PubMed: 4194429 32554 : Thompson SA et al. COOH-terminal extended recombinant amphiregulin with bioactivity comparable with naturally derived growth factor. J. Biol. Chem. 271: 17927-17931, 1996. PubMed: 8663535 32907 : Anderson MT et al. Simultaneous fluorescence-activated cell sorter analysis of two distinct transcriptional elements within a single cell using engineered green fluorescent proteins. Proc. Natl. Acad. Sci. USA 93: 85088511, 1996. PubMed: 8710900

PAGE 191

175 NHDF Dermal Fibroblasts, Adult All cells are performance tested and test negative for HIV-I, hepatitis-B & C, mycoplasma, bacteria, yeast and fungi. A certificate of analysis (CoA) is provided for each cell strain purchased. Cell Specifications: Clonetics cells are guaranteed to perform as indicated when used with Clonetics media and reagents. Typical Properties Sterility : Neg. Mycoplasma : Neg. Hepatitis B by PCR : Neg. Hepatitis C by PCR : Neg. HIV-1 by PCR : Neg. Questions? Call the e-Commerce Customer Service Desk and refer to this product. Technical Questions? Contact Technical Service, click here Order placement on our web site is available to customers in the US and Canada only. Pricing shown is applicable to US and Canadian customers only. Other customers please refer to the Contact Us section of our web site for the local distributor in your area. To order this product place the quantity you wish to order in the QTY field and click the shopping cart icon. When finished shopping, click Your Shopping Cart to view the contents of your cart and proceed to checkout. For further information n placing orders on our site refer to o Sales Order Placement FAQs

PAGE 192

176 Cell Lines ATCC Number: CRL-1629 Price: $224.00 Designation: SCC-9 Depositors: JG Rheinwald Biosafety Level : 1 Shipped: frozen Medium & Serum: See Propagation Growth Properties: adherent Organism: Homo sapiens (human) Tissue: tongue; squamous cell carcinoma Cellular Products: epidermal keratins; low levels of involucrin Permits/Forms: In addition to the MTA mentioned above, other ATCC and/or regulatory permits may be required for the transfer of this ATCC material. Anyone purchasing ATCC material is ultimately responsible for obtaining the permits. Please click here for information regarding the specific requirements for shipment to your location. Comments: Growth of SCC-9 is enhanced by using a feeder layer of Xirradiated STO cells (ATCC 56-X). The cells do not grow well in semi-solid medium. Tumorigenic: yes, Tumors developed within 21 days at 100% frequency (5/5) in nude mice inoculated subcutaneously with 10(7) cells. Age: 25 years Gender: male Propagation: ATCC complete growth medium: A 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F12 medium containing 1.2 g/L sodium bicarbonate, 2.5 mM L-glutamine, 15 mM HEPES and 0.5 mM sodium pyruvate supplemented with 400 ng/ml hydrocortisone, 90%; fetal bovine serum, 10% Temperature: 37.0 C

PAGE 193

177 Subculturing: Remove medium, and rinse with 0.25% trypsin, 0.03% EDTA solution. Remove the solution and add an additional 1 to 2 ml of trypsin-EDTA solution. Allow the flask to sit at room temperature (or at 37C) until the cells detach. Add fresh culture medium, aspirate and dispense into new culture flasks. Freeze Medium: culture medium 95%; DMSO, 5% Related Products: Recommended medium (without the additional supplements or serum described under ATCC Medium) ATCC 30-2006 recommended serum ATCC 30-2020 feeder layer cells ATCC 56-X References: 23039 : Rheinwald JG Beckett MA Tumorigenic keratinocyte lines requiring anchorage and fibroblast support cultures from human squamous cell carcinomas. Cancer Res. 41: 1657-1663, 1981. PubMed: 7214336

PAGE 194

APPENDIX C MAGNETIC SEPARATION DEMONSTRATION The file below contains a real-time movie (.avi format) demonstrating the magnetic separation of microspheres (S16), from media. The microspheres were suspended in water, inside a 12 mm thick centrifuge tube. The magnet utilized was an ordinary permanent magnet or rare-earth composition. s16doublemagnetdemo.avi 178

PAGE 195

LIST OF REFERENCES Antony, A. C. (1996). "Folate receptors." Annual Review of Nutrition 16: 501-521. Arshady, R. (1992). "Suspension, emulsion and dispersion polymerization: A methodological survey." Colloid and Polymer Science 270(717-732). Auclair, P. L., and K. Rasmussen (2002). Oral cavity and major and minor salivary glands. The Cancer Handbook M. Allison ed. London, UK, Nature Publishing Group. 1: 491-502. Barthelmy, D. (2004), "Mineralogy Database." http://webmineral.com/data/Maghemite.shtml February 19, 2004. Biosciences. (2004), "Raman and Rayleigh scattering." Amersham Biosciences. http://www5.amershambiosciences.com/aptrix/upp00919.nsf/Content/DrugScr+CyDye+Fluors+introduction%5CDrugScr+CyDye+Fluors+Overview+of+Fluorescence%5CDrugScr+CyDye+Fluors+Overview+of+Fluorescence+Photophysics%5CDrugScr+CyDye+Fluors+Overview+of+Fluorescence+Photophysics+Raman+and+Rayleigh+scattering February, 2004. Campbell, M. K. (1991). Biochemistry Philadelphia, Saunders College Publishing. Campbell, P. N., and A. D. Smith (1994). Biochemistry Illustrated Edinburgh, Churchill Livingstone. Canto, M. T., A. M. Horowitz, and W. L. Child (2002). "Views of oral cancer prevention and early detection: Maryland physicians." Oral Oncology 38(4): 373-377. Cormack, D. H. (1993). Essential Histology Philadelphia, J. B. Lippincott Company. Epstein, J., and C. Scully (1997). "Assessing the patient at risk for oral squamous cell carcinoma." Spec Care Dentist 17(4): 120-128. Espinosa-Mansilla, A., I. Duran-Meras, and F. Salinas (1998). "High-performance liquid chromatographic fluorometric determination of glyoxal, methylglyoxal, and diacetyl in urine by prederivatization to pteridinic rings." Analytical Biochemistry 255(2): 263-273. 179

PAGE 196

180 Field, M. J., and M. K. Jeffcoat (1995). "Dental Education at the Crossroads a Report by the Instituteof-Medicine." Journal of the American Dental Association 126(2): 191-195. Flores, A. B., L. A. Robles, M. O. Arias, and J. A. Ascencio (2003). "Small metal nanoparticle recognition using digital image analysis and high resolution electron microscopy." Micron 34(2): 109-118. Franklin, W. A., M. Waintrub, D. Edwards, K. Christensen, P. Prendegrast, J. Woods, P. A. Bunn, and J. F. Kolhouse (1994). "New Anti-Lung-Cancer Antibody Cluster-12 Reacts with Human Folate Receptors Present on Adenocarcinoma." International Journal of Cancer : 89-95. Franklin WA, W. M., Edwards D, Christensen K, Prendegrast P, Woods J, Bunn PA, Kolhouse JF (1994). "New anti-lung-cancer antibody cluster 12 reacts with human folate receptors present on adenocarcinoma." Int J Cancer Suppl 8: 89-95. Friedlander, P. F. (2003). "The use of genetic markers in the clinical care of patients with head and neck cancer." Archives of Otolaryngological Head and Neck Surgery 129: 363-366. Galia, M., F. Svec, and J. M. J. Frechet (1994). "Monodisperse polymer beads as packing material for highperformance liquid-chromatography effect of divinylbenzene content on the porous and chromatographic properties of poly-nstyrene-co-divinylbenzene) beads prepared in presence of linear polystyrene as a porogen." Journal of Polymer Science Part a-Polymer Chemistry 32(11): 2169-2175. Grimes, D. A., and K. F. Schulz (2002). "Uses and abuses of screening tests." Lancet 359(9309): 881-884. Hall, S. R., S. A. Davis, and S. Mann (2000). "Cocondensation of organosilica hybrid shells on nanoparticle templates: a direct synthetic route to functionalized core-shell colloids." Langmuir 16(3): 1454-1456. Haugland, R. P. (2003). Handbook of fluorescent probes and research products Eugene, OR, Molecular Probes. Herman, B. (1998). Fluorescence microscopy New York, Springer. Hermanson, G. T. (1996). Bioconjugate techniques San Diego, Academic Press.

PAGE 197

181 Herrero, R., X. Castellsague, M. Pawlita, J. Lissowska, F. Kee, P. Balaram, T. Rajkumar, H. Sridhar, B. Rose, J. Pintos, L. Fernandez, A. Idris, M. J. Sanchez, A. Nieto, R. Talamini, A. Tavani, F. X. Bosch, U. Reidel, P. J. F. Snijders, C. J. L. M. Meijer, R. Viscidi, N. Munoz, and S. Franceschi (2003). "Human Papillomavirus and Oral Cancer: The International Agency for Research on Cancer Multicenter Study." J Natl Cancer Inst 95(23): 1772-1783. Kim, J. W., J. W. Shim, J. H. Bae, S. H. Han, H. K. Kim, I. S. Chang, H. H. Kang, and K. D. Suh (2002). "Titanium dioxide/poly(methyl methacrylate) composite microspheres prepared by in situ suspension polymerization and their ability to protect against UV rays." Colloid and Polymer Science 280(6): 584-588. Kircher, M. F., U. Mahmood, R. S. King, R. Weissleder, and L. Josephson (2003). "A multimodal nanoparticle for preoperative magnetic resonance imaging and intraoperative optical brain tumor delineation." Cancer Res 63(23): 8122-8125. Kranz, D., T. Patrick, K. Brigle, M. Spinella, and E. Roy (1995). "Conjugates of folate and anti-T-cell-receptor antibodies specifically target folate-receptor-positive tumor cells for lysis." Proc Natl Acad Sci 92(20): 9057-9061. Lanza, G., D. Abendschein, X. Yu, P. Winter, K. Karukstis, M. Scott, R. Fuhrhop, D. Scherrer, and S. Wickline (2002). "Molecular imaging and targeted drug delivery with a novel, ligand-directed paramagnetic nanoparticle technology." Acad Radiol. Suppl 2: S330-331. Leamon, C. P., and P. S. Low (1991). "Delivery of macromolecules into living cells a method that exploits folate receptor endocytosis." Proceedings of the National Academy of Sciences of the United States of America 88(13): 5572-5576. Leamy, P. J. (2003). "Preparation, characterization and in vitro testing of poly (lactide-co-glycolide) and dextran magnetic microspheres for in vivo applications." Ph.D. Dissertation, Materials Science and Engineering University of Florida, Gainesville. Leckey, A. (1997). "Active microspheres for use in the treatment of hepatic tumors." Masters Thesis, Materials Science and Engineering University of Florida, Gainesville. Low, P. S., C. P. Leamon, J. A. Reddy, M. A. Green, C. Mathies, M. J. Turk, D. J. Walters, J. Lu, R. J. Lee, and M. Kennedy (2001). "Folate-mediated delivery of therapeutic and imaging agents to cancer tissue in vivo." British Journal of Pharmacology 134: 178P. Mashberg, A., and A. Samit (1995). "Early diagnosis of asymptomatic oral and oropharyngeal squamous cancers." Ca-a Cancer Journal for Clinicians 45(6): 328-351.

PAGE 198

182 Moffat, B., G. Reddy, P. McConville, D. Hall, T. Chenevert, R. Kopelman, M. Philbert, R. Weissleder, A. Rehemtulla, and B. Ross (2003). "A novel polyacrylamide magnetic nanoparticle contrast agent for molecular imaging using MRI." Mol Imaging. 2(4): 324-332. Pande, P., S. Soni, J. Kaur, S. Agarwal, M. Mathur, N. K. Shukla, and R. Ralhan (2002). "Prognostic factors in betel and tobacco related oral cancer." Oral Oncology 38(5): 491-499. Pankhurst, Q. A., J. Connolly, S. K. Jones, and J. Dobson (2003). "Applications of magnetic nanoparticles in biomedicine." Journal of Physics D-Applied Physics 36(13): R167-R181. Reddy, J. A., and P. S. Low (1998). "Folate-mediated targeting of therapeutic and imaging agents to cancers." Critical Reviews in Therapeutic Drug Carrier Systems 15(6): 587-627. Robineau, M., and D. Zins (1995). "Surfactant-coated particles in magnetic fluids. Characterization and study of themal stability under inert atmosphere." Annales De Chimie-Science Des Materiaux 20(6): 327-333. Ross, J. F., P. K. Chaudhuri, and M. Ratnam (1994). "Differential regulation of folate receptor isoforms in normal and malignant-tissues in-vivo and in established cell-lines physiological and clinical implications." Cancer 73(9): 2432-2443. Santra, S., B. Liesenfeld, D. Chatel, C. D. Batich, W. Tan, and R. A. Mericle (2004). "Development of novel folate immobilized fluorescent silica nanoparticles for sensitive detection of folate receptors in neoplastic cells." Journal of Nanoscience and Nanotechnology accepted for publication. Shim, J. W., J. W. Kim, S. H. Han, I. S. Chang, H. K. Kim, H. H. Kang, O. S. Lee, and K. D. Suh (2002). "Zinc oxide/polymethylmethacrylate composite microspheres by in situ suspension polymerization and their morphological study." Colloids and Surfaces a-Physicochemical and Engineering Aspects 207(1-3): 105-111. Sigma (2002), "Aldrich Polymer Products CD-Catalog and Reference Guide." Aldrich Chemical Co. www.sigma-aldrich.com August, 2002. Sivakumar, M., and K. P. Rao (2002). "Synthesis, characterization, and in vitro release of ibuprofen from poly(MMA-HEMA) copolymeric core-shell hydrogel microspheres for biomedical applications." Journal of Applied Polymer Science 83(14): 3045-3054. Sperling, L. H. (1992). Introduction to Physical Polymer Science New York, John Wiley & Sons.

PAGE 199

183 Stober, W., A. Fink, and E. Bohn (1968). "Controlled Growth of Monodisperse Silica Spheres in Micron Size Range." Journal of Colloid and Interface Science 26(1): 62-&. Suh, J. R., A. K. Herbig, and P. J. Stover (2001). "New perspectives on folate catabolism." Annual Review of Nutrition 21: 255-282. Tabor, M. P., R. H. Brakenhoff, H. J. Ruijter-Schippers, J. E. van der Wal, G. B. Snow, C. R. Leemans, and B. J. M. Braakhuis (2002). "Multiple head and neck tumors frequently originate from a single preneoplastic lesion." American Journal of Pathology 161(3): 1051-1060. Tanojo, H., H. E. Junginger, and H. E. Bodde (1997). "Influence of pH on the intensity and stability of the fluorescence of p-aminobenzoic acid in aqueous solutions." European Journal of Pharmaceutical Sciences 5(1): 31-35. Tsien, R. Y. (1998). "The green fluorescent protein." Annual Review of Biochemistry 67(1): 509-544. Tuncel, A. (1999). "Electron Microscopic Observation of Uniform Macroporous Particles. II. Effect of DVB Concentration." Journal of Applied Polymer Science 71: 2291-2302. Tuncel, A., and H. Cicek (2000). "2-Hydroxypropylmethacrylate based monoand bifunctional gel beads prepared by suspension polymerization." Polymer International 49(6): 485-494. Tuncel, A., M. Tuncel, H. Cicek, and O. Fidanboy (2002a). "2-hydroxyethylmethacrylate carrying uniform porous particles: preparation and electron microscopy." Polymer International 51(1): 75-84. Tuncel, A., M. Tuncel, B. Ergun, C. Alagoz, and T. Bahar (2002b). "Carboxyl carrying-large uniform latex particles." Colloids and Surfaces a-Physicochemical and Engineering Aspects 197(1-3): 79-94. Tuncel, A., M. Tuncel, and B. Salih (1999). "Electron Microscopic Observation of Uniform Macroporous Particles. I. Effect of Seed Latex Type and Diluent." Journal of Applied Polymer Science 71: 2271-2290. Tuncel, A., E. Unsal, and H. Cicek (2000). "pH-sensitive uniform gel beads for DNA adsorption." Journal of Applied Polymer Science 77(14): 3154-3161. Ugelstad, J. (1978). "Swelling capacity of aqueous dispersion of oligomer and polymer substances and mixtures thereof." Macromolecular Chemistry and Physics 179: 815.

PAGE 200

184 Ugelstad, J., T. Ellingsen, A. Berge, and O. B. Helgee (1988). "Process for preparing magnetic polymer particles." United States Patent 4,774,265 Ugelstad, J., J. H. Kaggerad, F. K. Hansen, and A. Berge (1979). "Absorption of Low Molecular Weight Compounds in Aqueous Dispersions of Polymer-Oligomer Particles, 2." Macromolecular Chemistry and Physics 180: 737. Vanblaaderen, A., and A. Vrij (1992). "Synthesis and Characterization of Colloidal Dispersions of Fluorescent, Monodisperse Silica Spheres." Langmuir 8(12): 2921-2931. Voltairas, P. A., D. I. Fotiadis, and L. K. Michalis (2002). "Hydrodynamics of magnetic drug targeting." Journal of Biomechanics 35(6): 813-821. Wang, Q. C., K. Hosoya, F. Svec, and J. M. J. Frechet (1992). "Polymeric porogens used in the preparation of novel monodispersed macroporous polymeric separation media for high-performance liquid-chromatography." Analytical Chemistry 64(11): 1232-1238. Wang, Q. C., F. Svec, and J. M. J. Frechet (1994). "Fine control of the porous structure and chromatographic properties of monodisperse macroporous poly(styrenecodivinylbenzene) beads prepared using polymer porogens." Journal of Polymer Science Part a-Polymer Chemistry 32(13): 2577-2588. Wang, S., and P. S. Low (1998). "Folate-mediated targeting of antineoplastic drags, imaging agents, and nucleic acids to cancer cells." Journal of Controlled Release 53(1-3): 39-48. Wasilewski, P., and G. Kletetschka (1999). "Lodestone: natures only permanent magnet what it is and how it gets charged." Geophysical Research Letters 26(15): 2275-2278. Weitman, S. D., R. H. Lark, L. R. Coney, D. W. Fort, V. Frasca, V. R. Zurawski, and B. A. Kamen (1992). "Distribution of the folate receptor Gp38 in normal and malignant-cell lines and tissues." Cancer Research 52(12): 3396-3401. Zhang, Y., N. Kohler, and M. Q. Zhang (2002). "Surface modification of superparamagnetic magnetite nanoparticles and their intracellular uptake." Biomaterials 23(7): 1553-1561.

PAGE 201

BIOGRAPHICAL SKETCH Bernd Liesenfeld was born in Frankfurt, in what was then formally the Federal Republic of Germany, commonly West Germany, in 1970. He attended high school in New York City and received his International Baccalaureate from the United Nations International School in 1988. In 1992 he received his Bachelor of Science degree from the University of Vermont, in Burlington, VT, in engineering management, specializing in mechanical engineering. He received his Graduate Diploma in Materials Science with honours from Monash University in Melbourne Australia in 1994, and went on to pursue his PhD at the University of Florida in Gainesville Florida. 185


Permanent Link: http://ufdc.ufl.edu/UFE0004401/00001

Material Information

Title: Superparamagnetic Folate-Immobilized Dye Labeled Microspheres for Oral Cancer Screening
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004401:00001

Permanent Link: http://ufdc.ufl.edu/UFE0004401/00001

Material Information

Title: Superparamagnetic Folate-Immobilized Dye Labeled Microspheres for Oral Cancer Screening
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004401:00001


This item has the following downloads:


Full Text











SUPERPARAMAGNETIC FOLATE-IMMOBILIZED DYE LABELED
MICROSPHERES FOR ORAL CANCER SCREENING














By

BERND LIESENFELD


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


2004

































Copyright 2004

by

Bemd Liesenfeld
















ACKNOWLEDGMENTS

I would like to thank all the people that contributed to this work, either directly by

helping to do the work or indirectly by supporting me.

Many thanks go to the members of my committee: my advisor Dr Chris Batich,

my external member Dr Ken Wagener, and my departmental members Dr Paul Holloway,

Dr Abbas Zaman and Dr Ken Anusavice who have contributed wonderful academic

examples as much as splendid technical help, and helped me find my own direction.

Additional thanks goes to Dr Karl Soderholm, who agreed to be a substitute member for

my defense at a late stage, and Dr Chiayi Shen, who substituted for Dr Soderholm with

almost no notice. Dr Shen, like each of my members, made an impressively careful,

thoughtful and insightful reading of my dissertation with minimal notice.

Perhaps more than that of anyone the always willing and generous help, of

Swadeshmukul Santra has enabled many research successes. Also deeply helpful was

Jon Dobson for providing expert advice and background information on magnetics.

Patrick Leamy was extremely helpful as a labmate, and the quality of the research

he left behind helped to enable many subsequent research successes.

Special thanks go to the graduate and undergraduate students who participated on

my proj ects in some very meaningful ways. Special thanks among these go to Cindy

Rau-Zink for coaxing any number of cell lines through the experiments needed. The

twins--Jompo Moloye and Taili Thula--helped prevent the cell lab from imploding

under its own weight. Bradley Willenberg and Mike Tollon were responsible for much









help on MAIC machinery and the machinery in our own labs. David Chatel was one of a

line of French summer students who have all contributed beyond their years of education.

Other primary participants include Rekha Nair who sacrificed many hours counting cells

and microspheres. JP Bullivant graciously trained any number of students in the

magnetite production process that we helped Pat Leamy to develop. Other participants

on the proj ect that I would like to thank include Leland Black, Vasana Maneeratana, and

members of the Batich research group that have provided excellent technical help: Albina

Mikhailova, Nakato Kibuyaga and the balance of the group.

Many other people have helped to provide support in ways that deserve

recognition. Gill Brubaker has provided tireless training and support for maj or

instrumentation at the Particle Science Engineering Research Center. Kevin Powers and

Gary Scheiffele at the same center also provided important assistance. Eric, Wayne, and

Erik at MAIC have each provided valuable instrumental assistance and guidance.

Jennifer Wrighton has provided critical administrative support that enables the

functioning of all the polymer research groups.

I would also like to thank my personal supporters, especially Kelly Rooney and

my parents who have been very patiently supportive and positive. Additional thanks go

to my triathlon and cycling teams--the Tri-Gators and Team Florida--which provided

sporting mental regeneration.





















TABLE OF CONTENTS


page


ACKNOWLEDGMENT S ................. ................. iii........ ....


LIST OF TABLES ................ ...............x............ ....


LI ST OF FIGURE S .............. .................... xi


AB S TRAC T ......_ ................. ............_........x


CHAPTER


1 INTRODUCTION ................. ...............1.......... ......


2 BACKGROUND .............. ...............3.....


2.1 M agneti sm ................. ........... ...............3.......
2.1.1 Types of Magnetism ................. ...............4............ ...
2.1.1.1 Diamagnetism ............... ..............
2.1.1.2 Paramagnetism .............. ...............6.....
2. 1.1.3 Ferromagnetism ........._ ....... __ ...............6.
2. 1.1.4 Ferrimagnetism ............ ....... ...............8.
2.1.1.5 Antiferromagnetism .............. ...............9.....
2. 1.2 Domain Size Effects ...._. ......_._._ .......__. ...........1
2. 1.2. 1 Single domains ........._.. ....___......... ...........1
2. 1.2.2 Superparamagnetism ........._. ........_._._ ........ ............1
2.1.3 Some Applications of Magnetic Particles ........._. ..... ...._._.........12
2. 1.3.1 Hyperthermic treatments ................ ...............13........... ..
2. 1.3.2 Contrast media. ................. .. ......... ...............14. ...
2.1.3.3 Particle guidance by magnetic forces............... ...............14.
2.2 Cancer .................. .............. .. ...............17......
2.2.1 General Cancer Background .............. ...............17....
2.2.2 Oral Cancer ................. ........... ...............18......
2.3 Folic Acid and Receptor Targeting ................. ...............21...............
2.3.1 Folic Acid ................. ...............21................
2.3.2 Folate Receptors ................. ....___ .....__ ............2
2.4 Fluorescence .............. .... ...............23..
2.4.1 The Fluorescence Process .............. ...............24....
2.4.2 Fluorescence Techniques ............... .. .. ...............25
2.5 Polymerization Methods for Producing Microspheres .............. ..................27











2.5.1 Emulsion Polymerization .............. ...............29....
2.5.2 Soapless Emulsion Polymerization .............. ..... ............... 3
2.5.3 Dispersion Polymerization .............. ...............30....
2.5.4 Precipitation Polymerization ................. ............. ....._.. .......3 1
2.5.5 Suspension Polymerization .............. ...............3 1....


3 PROPOSED STRATEGY AND DESIGN REQUIREMENTS ................. ...............33

3.1 Appropriate Uses of Screening Tests............... ...............33.
3.2 Proposed Testing Procedure .............. ...............35....
3.2.1 Components ................. ...............35.................
3.2.2 Procedure ................. ........... ...............35.......
3.3 Design Parameters for Microspheres ................. .........___.....__......37
3.3.1 Microsphere Size Considerations ......____ ........_ ...............37
3.3.2 Ligand Immobilization ......___ ........__ ....___ ............3
3.3.3 Magnetic Guidance............... ...............39
3.3.4 Dye labeling .............. ...............39....

4 MATERIALS AND METHODS .............. ...............41....


4.1 Magnetic Material for Microspheres .............. ...............41....
4.1.1 Materials used in Iron Oxide Preparation............... ..............4
4. 1.2 Magnetite Production and Treatment .........__ ....... ..............41
4.1.2.1 Method of iron oxide precipitation............... .............4
4. 1.2.2 Method of coating iron oxide .....__ ................ ................. .4
4.1.3 Characterization of Iron Oxide. ................ .............. ...............44
4.2 Microsphere Polymerization and Characterization............... ............4
4.2.1 Incorporation of Fluorescent Dye Into Functional Monomer .............44
4.2.1.1 Materials for fluorescent dye incorporation into functional
m onom er............... .... ... .... .. .. .. .........4
4.2. 1.2 Method of conjugating fluorescent dye to functional monomer .....45
4.2.2 Suspension Polymerization Procedure .............. ....................4
4.2.2.1 Materials for suspension polymerization .............. ....................48
4.2.2.2 Method of suspension polymerization .............. .....................4
4.2.3 Microsphere Post-Polymerization Processing .........._..._. .................50
4.2.4 Microsphere Dye Loading by Swelling and Solvent Evaporation.......5 1
4.2.5 Microsphere Characterization .............. ...............52....
4.2.5.1 Coulter sizing .............. ...............53....
4.2.5.2 Light microscopy............... ...............5
4.2.5.3 Zeta potential analysis........................ .. .........5
4.2.5.4 Scanning electron microscopy (SEM) and energy dispersive
spectroscopy (ED S) .........._......... ......... .......__ .......... 5
4.2.5.5 Inductively coupled plasma spectroscopy (ICP) ................... ..........54
4.2.5.6 X-Ray powder diffraction (XRD) .............. ....................5
4.2.6 Microsphere Fluorescence Properties .............. .....................5
4.2.6.1 Sample preparation................ ..............5
4.2.6.2 UV-Visible absorbance spectroscopy .............. ....................5











4.2.6.3 Fluorescence spectrometry ............... ...............57....
4.2.6.4 Confocal microscopy............... ...............5
4.2.6.5 Fluorescence microscopy ....................... ...............58
4.2.7 Preparation of Micro spheres for Cell Work ................. ... ............... ..59
4.2.8 Preparation of Microspheres by Dispersion Polymerization. ...............59
4.2.8.1 Method of dispersion polymerization .............. ....................5
4.2.8.2 Incorporation of magnetic species................... ..............6
4.2.9 Preparation of Microspheres by Activated Swelling. ................... ........61
4.3 Immobilization of Folic Acid onto Microspheres............... ..............6
4.3.1 Folic Acid Immobilization Procedure .............. .....................6
4.3.1.1 Materials for folic acid immobilization................. ............. 6
4.3.1.2 Method of folic acid immobilization onto microspheres ................64
4.3.2 Characterization of Folic Acid Immobilized Microspheres ...............65
4.3.2.1 UV-Visible spectroscopy .............. ...............66....
4.3.2.2 Fluorospectrometry .............. ...............66....
4.3.2.3 Brookhaven zeta plus .............. ...............66....
4-4 Cell Testing .............. ...............66....
4.4.1 C ell Lines .. ....................... .................................6
4.4.1.1 Malignant cell line CRL-5800 / NCI-H23 human epithelial lung
adenocarcinom a ............ .... ............. .............. ... ... ... .......6
4.4.1.2 Secondary testing cell line CCL-163 / BALB/3T3 clone A31 mouse
fibroblasts ................... ......... ... .. ........... ..................6
4.4. 1.3 Control cell line NHDF: normal human adult fibroblasts ...............67
4.4.1.4 Oral squamous cell carcinoma cell line SCC-9................ ...............68
4.4.2 Cell Culture Procedures................ ... ................6
4.4.3 Cell Culture Preparation and Testing Procedure ................. ...............68
4.4.3.1 Cells seeded onto multi-well plates............... ...............68.
4.4.3.2 Cells seeded onto coverslips .............. .... .... .... .............6
4.4.3.3 Cell experiments for microsphere specific binding.........................69
4.4.3.4 Fluorescently labeled microspheres .............. .....................7
4.4.3.5 Image treatment and analysis .............. ...............70....
4.4.3.6 Micro sphere recovery experiments .............. ....................7
4.5 Tissue Testing ............... ... .. ........... .. .......... ...........7
4.5.1 Institutional Review Board (IRB) Approval .............. ....................71
4.5.2 Tissue Preparation ................ ...............72........... ....
4.5.2. 1 Fresh tissue samples ................ ...............72.............
4.5.2.2 Snap-frozen tissue samples .............. ... ...............72..
4.5.3 Sample Treatment for Testing with Microspheres .........._... ..............73
4.5.3.1 Unmounted tissue testing procedure .............. ....................7
4.5.3.2 Slide mounted tissue testing procedure............._ .........._ .....73
4.5.4 Microscopy of Prepared Tissue Samples ............. ..........._ .......74
4.5.5 Magnetic Recovery of Microspheres From Tissue Samples ................75


5 RE SULT S AND DI SCU SSION ............... ...............7


5.1 M icrosphere Synthesis............... ...............7
5.1.1 Dispersion Polymerized Samples ................. ....._.._.............. ....7












5.1.1.1 D013 dispersion polymerization with ferrofluid ..........................77
5.1.1.2 DO30 dispersion polymerization with iron oxide precipitated in
situ .............. ... ...............78..
5.1.2 Activated Swelling .............. ...............84....
5.1.3 Suspension Polymerization ................ ...............88...
5.1.3.1 Magnetic dopant characterization .............. ....... ...............8
5.1.3.2 Suspension polymerization methods and incorporation of iron
oxide ............ _.. .... ..... .............8
5.1.3.3 M icrosphere size control .............. ...............91....
5.1.3.4 Particle morphology .............. ...............94....
5.2 Folic Acid Immobilization ......___ .........__ .... ............9
5.2.1 UV-Visible Spectrophotometry ........................... ........._.._. ...95
5.2.2 Fluorescence Spectrometry .............. ...............96....
5.2.3 Zeta Potential Measurement ................. ................................96
5.3 M icrosphere Labeling .............. ... .. .. ....... .. ...............98.....
5.3.1 Dye Loading Vs. Covalent Coupling of Dye ................. ................. .98
5.3.2 Dye Loading by Swelling ................. ............ ............... 99.....
5.3.3 Covalent Coupling of Dye to Microspheres ................. ........_._. ....103
5.3.3.1 Microsphere fluorescence .............. ...............104....
5.3.3.2 Confocal microscopy............... ..............10
5.3.3.3 Dye content optimization .............. ...............105....
5.4 Cell Line Testing...................................... ........0
5.4.1 Initial Testing with Cell Line NCI-H23 ........._._...... .._._...........110
5.4. 1.1 Initial testing results ........._._... .. ....__. .. ...._._.. .........10
5.4. 1.2 Determination of desirable microsphere size from testing............ 11 1
5.4.2 Non-labeled Microspheres ................. ........................__....11
5.4.3 Labeled Microspheres ........._._........__. ....__. .............1
5.4.4 Malignant Cell Lines Testing .......__. ......... ._ .. ......._._... ....113
5.4.4. 1 NCI-H23 human lung adenocarcinoma cell line ........................114
5.4.4.2 BALB/3T3 mouse fibroblast cell line............__ ..........___.....115
5.4.4.3 SCC-9 oral squamous cell carcinoma cell line..............._._. ..........115
5.4.5 Control Cell Line Testing ......___ ....... .......__.........16
5.5 Tissue Testing ............ ..... ._ ...............117..
5.5.1 Tissue Sample 1................. ........ ............11
5.5.1.1 Fresh tissue testing with sample 1 ................. ..................1
5.5.1.2 Mounted tissue testing with sample 1 ............_.. .........__......117
5.5.2 Tissue Sample 2................ ...............123..
5.5.3 Results with Tissue Samples ......____ ..... ... ..........__......123
5.6 Microsphere Recovery Experiments ....._.__._ ..... ... .__. .. ...._._.........124


6 DYE DOPED SILICA PARTICLES ......__....._.__._ ......._._. ............2

6.1 Introduction ........._.___..... ._ __ ...............126....
6.2 Background ........._..... .... ...... ._. ... ...._._.... ..........12
6.2.1 Stoiber Process for Producing Nanoparticles..........._...._ ............_.._. 126
6.2.2 Nanoparticles applications .............. ...............127....
6.3 Materials and Methods. .........._..._. ......._ ....._.. ...........2











6.3.1 M materials ................. ...............128................
6.3.2 Methods ............... .. .... ............ .. ......... .. .................2
6.3.2.1 Method of conjugating FITC fluorophores to APTS ........._.......128
6.3.2.2 Method of synthesizing fluorescent silica nanoparticles.............129
6.3.2.3 Method of immobilizing folic acid onto the DDS
nanop arti cl e s ................ .......... ...............130 ....
6.3.2.4 Transmission electron microscopy ................. ......................131
6.3.2.5 Scanning electron microscopy .............. ...............131....
6.3.2.6 Zeta potential measurement .............. ...............13 1..
6.3.2.7 Light scattering particle size measurement ................ ...............131
6.3.2.8 UV-visible absorption spectroscopy .............. .....................3
6.3.2.9 Fluorescence spectrometry ................. ................ ......... .132
6.3.2.10 Confocal microscopy............... ..............13
6.3.2.11 Cell experiments............... ..............13
6.4 Results and Discussion .............. ...............133....
6.4.1 Size of FSNPs.................. ..............13
6.4.2 Fluore science of the F SNP s ................ ...............135.............
6.4.3 Folic Acid Immobilized FSNPs .............. ...............135....
6.4.4 Cell Experiments ................ ...............138................


7 CONCLUSIONS AND FUTURE WORK ................. ...............................14


7.1 Magneti c Mi cro sphere Prep arati on and Characteri zati on ................... ...........14 1
7.2 Ligand Immobilization............... ............14
7.3 Microsphere Labeling ................. ......... ...............142 .....
7.4 Cell Line Testing............... ...............143
7.5 Tissue Testing ................. ...............145...............
7.6 Microsphere Retrieval ................. ...............145................


APPENDIX

A POLYMERIZATION RECORDS FOR SELECTED SAMPLES THAT APPEAR
INT THE MANUS CRIPT ............ ..... ._ ...............147..


B CELL LINE DATA ............ ..... ._ ...............168..


C MAGNETIC SEPARATION DEMONSTRATION ......____ ..... .. ...__...........178

LIST OF REFERENCES ............ ...... ..__ ...............179..


BIOGRAPHICAL SKETCH ............_...... .__ ...............185...

















LIST OF TABLES


Table pg

2-1 Units used for magnetic quantities .............. ...............5.....___ ....

5-1 Suspension Polymerization Series Results ................. ............__.......__......92

5-2 Zeta potential measurement on folate-immobilized and control sl9 microspheres.
Averages presented are for the three runs depicted, each consisting of 10 cycles...98

5-3 Polymer solubility parameter values for selected solvents, presented in common
form of (cal/cm3) 1, HOt in SI units. ............. ...............100....

5-4 Group Molar Contribution calculation of Hildebrand polymer solubility parameter
for AEMH monomer unit ..........___...... ...............102......

5-5 Swelling of microspheres by selected solvents and solvent-dye compatibility .....102

5-6 FITC concentrations in sample batches prepared to optimize dye content and
fluorescence yield ........... ..... .._ ...............107..

5-7 Initial cell experiments statistical evaluation ........... ..... ...............111

5-8 Results showing normalized counts of microspheres per unit area for control and
immobilized S18 microspheres on NCI-H23 cells............... ............... ...14

6-1 UV-absorption spectroscopy instrument output of absorption maxima for folic acid
assay on FITC5 FSNPs whose spectral curves are shown in figure 6-3 ................138



















LIST OF FIGURES


Figure pg

2-1 A lodestone with nails and magnetite fragments attached ............... .............. .4

2-2 Diamagnetic response ................. ...............5...............

2-3 Paramagnetic response curve ..........._. ....._... .. ...............6....

2-4 Hysteresis loop ........._.___..... .___ ...............8.....

2-5 Inverse spinel structure of magnetite showing A and B lattice positions ........._......9

2-6 Energy minimization by domain walls .............. ...............10....

2-7 Domain wall transitions ............_...... .__ ...............11...

2-8 Schematic for drug delivery system ...._. ......_._._ .......__. ...........1

2-9 Squamous cell carcinoma of the tongue ........._._.......___ ........___.......1

2-10 Carcinoma of tongue ........._._ ...... .... ...............20..

2-11 Mechanism of receptor mediated endocytosis used to target anti-cancer drugs to
tumourous cells. ............. ...............22.....

2-12 Structure of Folic Acid............... ...............22..

2-13 Jablonski diagram showing energy states for a fluorescence process ...................24

2-14 Polyatomic molecule spectra showing excitation and emission intensity
equivalence .............. ...............26....

2-15 Normalized fluorescence emission spectra of fluorescein (FL),
tetramethylrhodamine (TMR) and Texas Red (TR) dyes ........._._..........._.......28

2-16 fluoro probes hybridized to human metaphase chromosomes .............. .............29

3-1 Results matrix for disease state vs. test result............... ...............34.

3-2 Steps of the proposed testing strategy .......................__ ......__.........3











3-3 Size considerations for microspheres............... ..............3

4-1 Methyl methacrylate (MMA) structure ................. ...............43........... ...

4-2 Fluorescein Isothiocyanate (FITC) structure .............. ...............45....

4-3 Texas Red-X (TR) structure................. ..............4

4-4 Aminoethyl methacrylate hydrochloride salt (AEMHS) structure ................... .....46

4-5 Mechanism of FITC conjugation to AEMH monomer.. ..........__... .............. ..46

4-6 Mechanism of Texas Red-X conjugation to AEMH monomer .............................47

4-7 Suspension polymerization setup using mechanical stirrer and heating mantle....50

4-8 Magnetic separator apparatus .............. ...............51....

4-9 Spectra for Zeiss filter set 10 .............. ...............58....

4-10 Folic acid structure............... ...............6

4-11 Schematic of folic acid immobilization onto microspheres............... ..............6

4-12 Schematic detailing the carbodiimide mediated coupling of a carboxyl group
to an amine to form an amide linkage ................. ...............65..............

4-13 Tissue slice mounted on slide being rinsed as part of testing procedure ...............74

5-1 Sample D013-4; ST-co-DEA particles dispersion polymerized in ferrofluid .......77

5-2 EDS spectra of sample D013-4 showing strong iron peaks............... ................78

5-3 Diethyl aminoethyl methacrylate (DEA) structure ................. .......................79

5-4 Samples d03 0_3 ................. ...............80..___ .....

5-5 XRD spectrum from sample d030_4m. ............. ...............81.....

5-6 XRD data listing from spectrum shown in figure 5-5 for sample d030_4m. ........82

5-7 D009 dispersion polymerized microspheres. .............. ...............83....

5-8 Seed particles D052 for activated swelling made by dispersion polymerization
of styrene ........... _... ......... ...............85.....

5-9 Size graph for sample ASO5 and its seeds D052_3 .............. .....................8

5-10 SEM micrograph of sample ASO5 .............. ...............86....










5-11 Sample ASO8, polymerized from equal parts ST/HEMA/EGDMA ....................87

5-12 SQUID magnetometer magnetic hysteresis curve for uncoated iron oxide...........89

5-13 TiO2 (10 w/v %) doped PMMA particles produced by Shim's suspension
polymerization process applied for most samples produced in this study............. 94

5-14 Sample SO4 showing particles formed by magnetite doped suspension
polym erization .............. ...............95....

5-15 UV-Visible absorption spectrum. ............. ...............97.....

5-16 S19 particles dispersed on slide at 10x ............. ...............104....

5-17 S19 microspheres, 40x fluorescent image showing fluorescence intensity
inhomogeneity within microspheres. .............. ...............104....

5-18 Confocal image of S19 microspheres at 60x. ................ .......... ...............106

5-19 Light intensity plot for figure 5-18. ............. ...............106....

5-20 Confocal microscope image of sample S25 with Texas Red dye ................... .....109

5-21 NCI-H23 cell line tested with S11 control microspheres (left) and S11 folate
immobilized microspheres (right) ................. ...............111................

5-22 Volume average size graphs for sample S11 ................ ......... ................1 12

5-23 BALB/3T3 mouse fibroblast cell line with folate-immobilized sl9
microspheres ........... ..... ._ ...............116...

5-24 Fresh tissue sample I with S19 immobilized microspheres .............. .... ........._..118

5-25 H&E stained 10 um section of squamous cell carcinoma tumour of
tongue/neck region from tissue sample 1............... ...............118...

5-26 10 um slice of sample 1 tissue mounted on slide and tested with control
microspheres ................. ...............119...._ ._ ......

5-27 Panel of 10 um section mounted tissue from sample 1, treated with folate-
immobilized microspheres sl9, 10x .............. ...............121....

5-28 Panel of 10 um section mounted tissue from sample 1, treated with control
micro spheres sl 9, 10x............... ...............122..

6-1 Conjugation of fluorescein isothiocyanate to APTS monomer. ................... .......129

6-2 Formation of silica structure by base-catalyzed condensation, incorporating
APT S ................ ...............13. 0...............










6-3 TEM of dye doped silica nanoparticles, sample FITC2. ............. ...................134

6-4 Coulter LS 230 graph showing size distribution of FITC2 FSNPs ................... .. 134

6-5 UV-Vis absorption spectra of folate immobilized FSNPs (top), 50 uM folic
acid in solution (middle) and control FSNPs (bottom) ................. ................ ..137

6-6 Panel ofFSNPs specifically bound to tumorous cells. ............. ....................138

6-7 Optical (left panels) and fluorescence (right panels) confocal images of folate-
immobilized FSNPs on BALB/3T3 fibroblasts .........___.........._._ .............139

6-8 NCI-H23 cells treated with FSNPs ................ ...............140..............

7-1 Results matrix for testing on cell lines............... ...............144.
















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

SUPERPARAMAGNETIC FOLATE-IMMOBILIZED DYE LABELED
MICROSPHERES FOR ORAL CANCER SCREENING

By

Bernd Liesenfeld

May 2004

Chair: Christopher D. Batich
Major Department: Materials Science and Enginering

A design concept is presented and developed for a screening test for oral cancer.

The application is based on generating specific binding between microspheres and

receptors known to be expressed specifically on malignant cells. Quantifieation of the

test is derived from a ratiometric determination of test microspheres immobilized with

folate against control microspheres.

Microspheres were suspension copolymerized polymethyl methacrylate and

aminoethyl methacrylate, and were doped with superparamagnetic iron oxide to permit

magnetic separation of microspheres from testing suspension. Magnetic separation was

demonstrated. Specific binding was provided by folic acid that was immobilized on the

microsphere surface by carbodiimide chemistry. Microsphere labeling was performed by

covalent bonding of fluorophores to monomers prior to polymerization, permitting spatial

imaging of microspheres by fluorescence microscopy.










Testing of specific binding of folate to tumorous cell lines was performed using

cell lines known to overexpress folate receptors. Cell lines used included NCI-H23

human lung adenocarcinoma, with controls provided by normal human dermal

fibroblasts. It was found that the folate-immobilized microspheres were preferentially

retained by the tumourous cell line, relative to control microspheres (p = 0.0074). There

was no significant difference between the retention of folate-immobilized microspheres

by the cancerous cell line as compared to the control cell line (p = 0.90) as determined by

pooled data.

Testing of specific binding to relevant tissue was performed using excised oral

cancer tissue that had been frozen and sectioned onto slides. It was found that the folate

immobilized microspheres were retained by the cancerous tissue at a higher rate than the

control microspheres (p = 0.037). Controls performed with normal tissue shows that the

folate-immobilized microspheres were retained by normal tissue at a higher rate than the

cancerous tissue. Both cell line data and tissue data show false positive responses, which

may be due to non-specific binding of folate-immobilized microspheres to samples.















CHAPTER 1
INTTRODUCTION

The incidence of oral cancer is about 30,000 newly diagnosed cases annually for

the U.S. meaning that roughly one in 10,000 people are affected. Approximately 90 %

of these are squamous cell carcinomas (SCC). (Tabor et al. 2002; Herrero et al. 2003),

Epidemiological data shows that smokers are at much higher risk for SCC than the

general population, and that excessive alcohol use exacerbates the risk. It is estimated

that up to 90 % of oral cancers are associated with these risk factors. Oral SCC often

manifest as painless and innocent appearing keratinized ulcerations, so that the initial

lesions typically remain undiagnosed until the malignancy metastasizes. When a

metastasized malignancy is found, the prognosis is quite poor, and the treatment route has

high morbidity. If the initial lesion were diagnosed prior to metastasis, the prognosis

would be greatly improved, and the treatment routes available would be less morbid. The

combination of readily identifiable risk groups, and improved outcomes for earlier

detection make oral cancer an ideal application for a screening test.

A novel detection system is developed that could be applied to oral cancer for

screening purposes. The primary thrust of the dissertation is the development and

characterization of the microsphere system that acts as the reporter particle. Emphasis is

also placed on testing performed to verify functionality of the system. Much of the

technical development work for this system is not unique to the ligand and target chosen

to suit the particular application. Once regard is taken for the peculiarities and specific

handling requirements unique to a receptor-ligand binding system, the technology can be










applied in a more general manner. It is believed that this type of screening system could

be useful for numerous conditions.

The structure of the dissertation is to present the background on the application

and relevant fields of study pertaining to the research, followed by a detailed description

of the envisioned application system that is ultimately the goal of the research group to

develop. The technical requirements of the applications are in turn used to develop the

design parameters of the particles. The materials and methods section provides intimate

detail on how the product was synthesized, characterized, processed and tested. A results

and discussion section is provided that details the findings and interprets them as relevant

to the research. A largely independent chapter on dye doped silica particles is inserted

prior to the conclusions. This chapter discusses synthesis and characterization as well as

some experiments conducted with fluorescent silica nanoparticles. This system was

applied as a model for the chemistry of the microspheres in situations where physical

characteristics of the microspheres proved to be confounding to important

characterizations. The silica nanoparticles had analogous surface chemistry to the

microspheres. Conclusions are based on the findings from the microsphere-based

research, supplemented by information that could only be derived using the model

sy stem.















CHAPTER 2
BACKGROUND

2.1 Magnetism

Magnesia (in the Thessaly region of Greece) lent its name to a line of gentle, wise

and just centaurs the Magnetes, mythically descended from the Magnesian Mares that

birthed the first centaurs. Eventually the word passed into more modern language as

'magnates', at one point describing landowners or medieval noblemen, but in modern

times referring to a "great man" or one that is particularly important or influential in

some field--particularly business. Our word Magnetism is also descended from the

Thessalian inhabitants the Magnates, due to the local abundance of lodestone--a

magnetized form of the mineral magnetite (Fe304). Lodestone (or its equivalent

loadstone) has magnetic polarity, enabling the Chinese mathematician Shen-Kua (1030 -

1093 AD) to build the first recorded navigational compass by using the permanently

magnetic lodestone to magnetize a soft iron compass needle (since the magnetic response

of the soft iron needle would fade over time a lodestone had to be carried to remagnetize

the compass needle occasionally).

While magnetite is relatively abundant, lodestone is not as common. Lodestone

turns out to be an intimate mixture of magnetite and maghemite that is typically found

very close to the surface--typically in volcanic regions (Wasilewski and Kletetschka

1999). As rocks of the proper composition cool, they are magnetized by the earth's

magnetic field, and 'freeze in' that magnetic orientation. This permits researchers to










study deposits of magnetized rock for a historical recording of shifts in the earth' s

magnetic field, or to use these patterns of shifts to date geological structures.
















Figure 2-1: A lodestone with nails and magnetite fragments attached. Sourced:
Moskowitz (1991), "Hitchhiker's Guide to Magnetism",
http://www.geo.umn. edu/org s/irm/hg2m/hg2m~index .html, February 2004

2.1.1 Types of Magnetism

There are Hyve maj or groups of magnetic materials, classified by their type of

magnetic behaviour--their response to an applied Hield. These Hyve are diamagnetism,

paramagnetism, ferromagnetism, ferrimagnetism and antiferromagnetism. Of these, the

first two do not show collective magnetic interactions nly the last three have long-

range magnetic order. The materials that are generally considered as 'magnetic' are

ferromagnets and ferrimagnets--all other groups have relatively weak magnetic

properties as expressed by the materials' bulk susceptibility: X. The bulk susceptibility X

is the slope of the M vs. H curve, and is used as a measure of the strength of a materials

magnetic response. Some basic magnetic terms and their units are detailed in table 2-1.









Table 2-1: Units used for magnetic quantities
Term Magnetic quantity SI units CGS units
B Magnetic Induction T (tesla) G (gauss)
H Applied field A/m (ampere / meter) Oe (oersted)
M Magnetization A/m (ampere / meter) G (gauss), emu/cm3

In terms of equations (using SI unit formulations) this yields:
B = Clo (H + M) (equation 2-1)
M = XH (equation 2-2)
Where C1o is the permeability of free space.

Detailed below are the types of magnetic responses in terms of the magnetization

vs. applied field curves that are generated (the response diagrams are not drawn to scale

and indicate only the trend of behaviour, not the magnitude relative to other curves).

Some texts show B vs. H while others show M vs. H curves. As far as illustrating the

relevant trends, these are largely the same.


M


slope~ = H


Figure 2-2: Diamagnetic response

2.1.1.1 Diamagnetism

All matter is diamagnetically responsive, but the effect is very weak and can be

completely masked by stronger magnetic properties (ferro- and ferri-magnetism).

Diamagnetism arises from the non-cooperative behaviour of orbiting electrons when

exposed to a magnetic field, and is not temperature dependent. As seen in figure 2-2,

diamagnetic materials have a small negative magnetic susceptibility: X = 10 5, so that the










magnetic response opposes the applied magnetic Hield. Superconductors are a special

class of diamagnetic materials that have a susceptibility X -1.





slope= x








Figure 2-3: Paramagnetic response

2.1.1.2 Paramagnetism

Some of the atoms or ions in this class of materials have net magnetic moments

due to unpaired electrons in partially filled orbitals. For paramagnetic materials the

susceptibility X is a very small positive value of approximately 10-3 10- The

magnetization of paramagnets is aligned parallel to an applied magnetic Hield. The form

of the paramagnetic magnetic response curve is shown in Eigure 2-3.

2.1.1.3 Ferromagnetism

Ferromagnets are the most widely recognized group of magnetic materials.

Susceptibility values are large and positive: X = 50 10,000. This is due to strong

interactions by electronic exchange forces (a quantum mechanical phenomenon that

arises from the relative orientation of the spins of electrons) resulting in parallel or

antiparallel alignment of magnetic moments. These materials are capable of being

magnetized up to their saturation magnetization (the maximum induced magnetic

moment), by relatively weak applied Hields. At temperatures above their Curie

temperature, the magnetic behaviour changes to paramagnetic (and the remnant










magnetization goes to zero). Ferromagnets exhibit the magnetic property of hysteresis,

and are able to retain a certain magnetization when the applied field is removed. Those

ferromagnets that retain a large percentage of their saturation magnetization when the

applied field is removed are called hard magnets, whereas those that lose most of their

magnetization are termed soft magnets. This is illustrated in figure 2-4, which shows a

hysteresis curve with relevant parts of the curve labeled. Starting from the origin, the

magnetization curve slopes upwards as a filled H is applied at an initial susceptibility of

Xo, eventually reaching saturation magnetization Ms at point A. Reducing and reversing

the applied field the magnetization does not follow the same curve again this effect

being called hysteresis. The curve proceeds along past point B and at zero applied field

retains a magnetization Mr the remnant magnetization. This is the amount of

magnetization that the material will retain when not in a magnetic field the permanent

magnetization of the material. Following the curve to point D reaches the intersection

with zero magnetization at applied field -He representing the coercive force He, also

known as the material's coercivity that is the amount of field (in the opposite direction)

required to demagnetize the material. Further increases in the applied field to point E

generate a magnetization in the negative direction up to the saturation point -Ms, reached

at point E. Any further increases in field magnitude over E (negative direction) or A

(positive direction) can produce no further increase in magnetization past the saturation

points this is the state of the material being fully aligned in its magnetic state.

Reversing the curve again produces the same type of curve through points F (the negative

remnant magnetization) and G--the coercivity. Further reversals will traverse the path

around the loop indicated by ABCDEFGA... This loop is called the hysteresis loop.










M B

























Figure 2-4: Hysteresis loop. This shows the magnetic response of ferromagnetic and
ferrimagnetic materials below their Curie temperatures

2.1.1.4 Ferrimagnetism

Ferrimagnets are ionic compounds that display both ferromagnetic and

antiferromagnetic properties, and exhibit a small positive magnetic susceptibility that

increases with temperature. Macroscopically, ferrimagnets behave in much the same way

as ferromagnets, exhibiting hysteresis and saturation in their magnetic curves.

Magnetite is perhaps the best known ferrimagnet, with an inverse spinel crystal

structure. For ferrimagnets, there are A and B type ions that occupy different lattice sites

(figure 2-5)-fr magnetite the structural formula resolves from the molecular Fe304 to

[Fe3+]A [Fe3+, Fe2+]B 04. The trivalent ferric ions occupy tetrahedral sites where they are

surrounded by four oxygens, while the divalent ferrous ions occupy octahedral sites and










are surrounded by six oxygens. The A sublattice spins are antiparallel to the B sublattice

spins and with the negative AB exchange interaction, the B site ferrous ions contribute

the net magnetic moment of magnetite.










.. A--- tetraedrlF
t'.'. A. -" A-sit
ocaeda Fe

MoIwt (1985)


Fgr2-:Ines spnlsrcueo antt hoigAadBltiepstos
Sorcd Msoiz(91," itchkrsGiet ants"
htp/wwgoun~d/rsimhImh~~ne~tl Feray20









aniprall rrngmntofmgnti dple ta I---is-- eqa n opposite. r









2.1.2 Domain Size Effects

Magnetic domains in ferrimagnetic materials are much larger than atomic

dimensions, but small in a macroscopic sense, existing at a scale of single to 100s of Clm.

Domains can be the result of particle or grain size, or they can form to minimize energy

within a single crystal. Figure 2-6 illustrates the process of domain wall formation as it

minimizes the system energy.

Domain Formation


++++I -- I++ +-










Single Multidomain
Domain
Total Energy = Magnetostatic Energy +
Wall Energy

Figure 2-6: Energy minimization by domain wall. Sourced: Moskowitz (1991),
"Hitchhiker's Guide to Magnetism",
http://www.geo.umn. edu/org s/irm/hg2m/hg2m~index .html, February 2004

Magnetostatic energy (the energy associated with the surface charge distribution)

is minimized by subdividing a single domain into multiple domains. The surface charges

at the ends of a domain form a demagnetizing field. When a domain is split into two, the

magnetostatic energy is reduced by almost half. The demagnetizing field is reduced by

bringing opposite charges closer together. The formation of each new domain wall

requires a transition between the alignments of the spins. A gradual transition between

spins results in a wide wall (Figure 2-7a) that minimizes the exchange energy. A sharp









transition (Figure 2-7b) between spin states minimizes the anisotropy energy. The

balance between these competing forces produces domain wall widths of ~ 100 nm.



(a) 'n wide wall 1




thin wall





Figure 2-7: Domain wall transitions. Sourced: Moskowitz (1991), "Hitchhiker's Guide
to Magnetism", http ://www.geo.umn.edu/orgs/irm/hg2m/hg2mrindehtl
February 2004

2.1.2.1 Single domains

Below a certain grain size, domain walls are no longer energetically feasible, and

the grain (or particle) will contain a single domain (SD) that will be uniformly

magnetized to saturation magnetization. It is an energetically difficult process to rotate

the magnetization of an SD. SD grains are therefore magnetically hard (high coercivity

and remnance). For magnetite the transition from multiple domains to SD is around 80

nm (Leamy 2003).

2.1.2.2 Superparamagnetism

When an SD particle is small enough, and/or temperature is high enough, then the

thermal energy can overcome the anisotropy energy that separates opposite magnetization

states, and spontaneous reversal of magnetization occurs. Remanence and coercivity thus

become zero, and the grain / particle becomes superparamagnetic. A superparamagnetic

particle will have zero net magnetic moment when no field is applied. Under the










application of a field, a net statistical alignment of the magnetic moments will occur.

This behaviour is similar to paramagnetism, but on a scale of perhaps 105 atoms rather

than for a single atom, generating a much higher susceptibility. The transition size of a

particle from SD behaviour to superparamagnetic behaviour has been calculated (see

equation 2-3) as between 22 and 23 nm, for spherical pure magnetite at room

temperature.

1/t = fo exp ((-Kvv)/(kT)) (equation 2-3)
where
fo = frequency factor (19OgsC-1)
Kv = anisotropy constant
v = particle volume
k = Boltzmann constant
T = absolute temperature

The practical value of particles with superparamagnetic properties is that they will

show a strong magnetization in response to an applied field allowing the generation of a

significant motile force. When the applied field is removed, no remnant magnetization

remains, and the particles have no drive to agglomerate. This type of response is easy to

verify on macroscopic samples with equipment no more complex than a magnet and

some superparamagnetic particles.

2.1.3 Some Applications of Magnetic Particles

There are a large number of applications that exploit the magnetic properties of

materials, beginning most obviously with the magnetic compass needle long ago, to such

mundane current items as refrigerator magnets. Many high tech applications such as

magnetic storage media (cassette tapes through hard drives), and medical imaging

technologies exploit the magnetic properties of fine particulates. Detailed below are









samples of biomedical applications that utilize magnetic particles (mostly

superparamagnetic).

2.1.3.1 Hyperthermic treatments

Hyperthermia is a clinical treatment used to combat tumours that involves the

artificial elevation of temperature in the body. Mild hyperthermia (under 420 C) is used

to stimulate the immune system of the body, while higher temperatures (about 450 C) are

applied to attempt to destroy specific (cancerous) cell populations. The mechanism of

hyperthermic treatment is to introduce coated magnetic particles to the site in question

and heat them through application of an alternating current (AC) field. This method is

called inductive heating, and provides a way to heat only the tissue in immediate contact

with the magnetic particles. Targeting the particles to the desired tissue thus represents

the key to successful hyperthermia applications.

The underlying physical phenomena that produce heating vary by the class of

magnetic material utilized, but are usually quoted as a specific absorption rate (SAR) so

that comparisons are viable. Ferromagnetic or ferrimagnetic (FM) particles rely on

hysteretic losses to generate heat. Quite high magnetic fields can be required to achieve

saturation magnetization higher than generally considered clinically viable for humans.

Superparamagnetic (SPM) particles generate heat when a field is applied and removed by

relaxation either of the particles themselves when in fluid (Brownian rotation,) or by

rotation of the atomic magnetic moments within the particles (Neel relaxation). The

latter process occurs at higher frequencies, and is the only process available if the

particles are not free to rotate physically. Owing to the lower applied field (well within

limits considered safe for human clinical use) required to generate heating in SPM










particles, these are the primary avenue of development in the Hield (Pankhurst et al.

2003).

2.1.3.2 Contrast media

Paramagnetic and superparamagnetic particles have been applied as contrast

agents for magnetic resonance imaging (MRI) applications, since they alter the relaxation

times of the tissue in which they are resident. This occurs through the particles'

enhancement of the local magnetic Hield strength due to the influence of their own

induced magnetic Hields, and by their creation of Hield inhomogeneities. In order to

obtain specific image enhancements for features of interest, the key is to optimize the

targeting of the particles to the relevant tissue or features, since the contrast enhancement

effects are very short range.

2.1.3.3 Particle guidance by magnetic forces

There are many variations on this concept that are in current clinical use. The

common thread through these is that there are two components to the applications that are

critical to their success. The first is that the particles are responsive to a magnetic field -

such that they can be retained at some site of interest (a target organ perhaps, or a

collection site). The second is a chemical modification of the particle to carry a payload

- this might be a label or drug or antigen that will either facilitate attachment of the

particle of interest to the target or that will provide a therapeutic or analytical service.

Applications that can be categorized under this heading include magnetic separations and

magnetically targeted drug delivery.

To make particle guidance applications feasible, the motive force on particles

needs to be sufficient to move the particles over some distance in a viscous fluid. Fluids

are often aqueous or blood, and may be flowing at some rate vw (velocity of water) such










that a Av = vm vw where vin represents the magnetic particle velocity (zero if particles

are being retained in a fixed position by a fixed magnet). The hydrodynamic drag force

Fd then equates to equation 2-4:

Fd = 6 n: r Rm Av (equation 2-4)

In the above equation r represents the viscosity of the medium, and Rm

represents the radius of the magnetic particle. This equation shows the force that must be

generated between the applied field and the magnetic particle to retain the particle in a

flowing medium. The magnetic force Fm generated by a particle is given by equation 2-5

below:

Fm = Van AZ V(1/ B H) (equation 2-5)

Where Vm is the volume of the magnetic particle (if the magnetic core is coated,

or a dispersion of magnetic nanoparticles exists within a microparticle then the magnetic

particle volume may not be equal to the hydrodynamic volume for the particle). The term

V(1/ B H) represents the differential of the magnetostatic field energy density,

indicating the direction of magnetic force.

Magnetic separations include several applications such as magnetic cell sorting

(magnetic particles are coated with a reagent that selectively attaches to target cells,

permitting the separation of bound cells) and immunoassays (methods that employ

antibodies to measure analyte concentrations). These applications rely on the magnetic

component of the particle to enable recovery from a flow system with the efficiency of

the process being proportional to the flow rate at which the particles can still be

recovered. Targeting can be tailored by careful selection of the species on the particles










that generate specific binding allowing multiple targets to be accessed in a single

separation.




blood vessel
Magnetic nanoparticles










Figure 2-8: Schematic for drug delivery system. A magnet external to the body retains
magnetic carriers at target site

Drug delivery is another application to benefit from magnetic carriers to permit

more localized delivery. Localized drug delivery is desirable for many drugs such as

anti-inflammatory agents or chemotherapeutics that are toxic in the doses required if

administered systemically. Localized targeting permits an overall lower drug dose to

yield a higher local concentration, reducing side-effects. Drug deliver applications are

typically designed around magnetite or maghemite magnetic particles either with a

magnetic core or with magnetic nanoparticles dispersed within a polymer microparticle.

Biocompatible polymers often used include polyvinyl alcohol (PVA) and dextran. The

drug payload can be either attached to the surface of the particle or stored inside if the

particle. The magnetic drug carrier particles are typically introduced into the body as a

ferrofluid and then detained at the target region by means of a strong applied field as

shown in figure 2-8. Hydrodynamic investigations of blood flow and magnetic retention

suggest that successful magnetic drug delivery is more likely for regions of slower blood









flow where a magnet can be placed in close proximity (Voltairas et al. 2002; Pankhurst et

al. 2003).

2.2 Cancer

2.2.1 General Cancer Background

Cancer as a disease has been known to society for a very long time, with sources

as far back as 2500 BC recording surgical treatments in ancient Egypt. It is believed that

the ancient Egyptians were able to distinguish malignant from benign tumours tumourr

meaning literally a 'swelling' or 'lump'). Only malignant tumours are termed cancerous,

their cells invading the basement membrane to spread to other parts of the body in a

process called metastasis. Benign tumours do not invade other parts of the body, though

they may continue to grow, which can still be a problem in a constrained setting (i.e. the

brain).

The term cancer is credited to Hippocrates (~430 360 BC), who noted the

crablike form of a tumour ("karkinoma" being Greek for crab). The simplest modern

definition of cancer is provided by the American Cancer Society (ACS): "cancer is a

group of diseases characterized by uncontrolled growth and spread of abnormal cells".

Cancerous cells growing out of control are unable to recognize the signals that cause

normal cells to stop growing at their natural boundary, to specialize to perform some

useful function, or to stop replicating and die.

Cancers are classified by two characteristics: their histological type (categorized

by tissue type wherein the cancer originated) and location (primary site of origination).

There are Hyve maj or categories of cancers: carcinoma, leukemia, lymphoma, myeloma

and sarcoma. Carcinomas are malignancies of epithelial tissue that have a tendency to

metastasize to other parts of the body, and account for 80 to 90 percent of all cancer










cases. There are two maj or types of carcinoma, categorized by site of development.

Adenocarcinomas develop within an organ or gland, while squamous cell carcinoma

(squamous meaning flat or scale-like) develop within the squamous epithelium.

Sarcomas are cancers of supportive or connective tissues, i.e. bone, muscle, tendon,

cartilage, muscle and fat. Myelomas originate in plasma cells of bone marrow.

Leukemias are also cancers of bone marrow specifically as relates to the production of

white blood cells. Lymphomas are cancers of the lymphatic system arising in lymphoid

tissue or lymph nodes.

2.2.2 Oral Cancer

In the U.S. head and neck squamous cell carcinomas (HNSCC, see figure 2-9)

represent approximately 4 % of all newly diagnosed cancers, or about 30,000 new cases

yearly (Tabor et al. 2002). Over 90 % of cancers of the oral mucosa and the lips

vermilion are squamous cell carcinomas--these are what is classically referred to as 'oral

cancer' (Auclair and Rasmussen 2002). HNSCC is a serious health problem with a well

defined at-risk population. Heavy alcohol indulgence acts as an independent

multiplicative factor for the primary risk factor: tobacco use (Mashberg and Samit 1995).

Smokeless tobacco products (such as 'dip', 'chew' or tobacco pouches) may represent an

even greater health challenge than smoked tobacco. In some Asian countries, where

chewing of betel leaves supplements or substitutes for use of tobacco products, the oral

cancer rate soars to 40 % of new cancers (Pande et al. 2002).

Efficient diagnosis of oral cancers presents a number of difficulties: the hidden

nature of many sites in the oral environment, the often innocuous appearance of the

lesions, a lack of pain or clear lumps to guide the clinician (figure 2-9 and 2-10), and the









abundance of similar-appearing harmless ulcerations. Detection frequently occurs only

when the tumour has metastasized ften as a lump in the neck. At this later stage the

prognosis is quite poor, worse than for breast, cervical, or prostate cancers (Canto et al.

2002)


















Figure 2-9: Oral squamous cell carcinoma. The small, granular red lesion shown by the
arrow was totally asymptomatic and was only slightly thickened. Sourced:
Bristol Biomedical Image Archive (2002),
http://www.bri sbio.ac.uk/ROAD S/subj ect-
listing/carcinomasquamouscell .html, April 2004.

Clinicians were found, in a survey (Mashberg and Samit 1995), to underestimate

the incidence of the disease, and hence not routinely examine as carefully as possible for

it. Examination is complicated by the innocuous appearance of the lesions, which is not

easily discriminated under the lighting conditions available (head lamps, pen lights).

Staining with toluidine blue provides a sound method for detection, but suffers the

confound that all ulcerations accept stain, requiring a subsequent confirming treatment

some weeks later (Epstein and Scully 1997).

The above factors combine to make late diagnosis the more typical situation, as

confirmed by a recent perspective (Friedlander 2003). Friedlander concludes that "early









detection may be the most important strategy to improve overall survival for patients with

head and neck cancer. Currently one third of patients present with HNSCC present with

early-stage disease, whereas two thirds present with late-stage disease.... Contributing

factors to late presentation ... include lack of a definite symptom complex..."




















Figure 2-10: Squamous cell carcinoma of the tongue. Lesions can be painless and
indistinct in spite of being in advanced stages. Sourced: Ghorayeb (2004)
"Pictures of Tongue Cancer", http://www.ghorayeb.com/TongueCancer.html
April 2004.

At the late stage the simple excision of the offending mass is no longer sufficient

treatment and radiation based therapy is indicated in order to prevent further metastases,

and/or to combat metastases that have already occurred. Radiation based therapy in the

region of the oral cavity is highly undesirable even when successful since a common side

effect is the inactivation of the radiation-sensitive salivary glands. A deficit in salival

production brings a host of complications to the patient, including reduced taste

perception, increased tooth decay rates, and drymouth--all contributing to reduced

quality of life. Calls have been made for research into improved detection methods

(Field and Jeffcoat 1995); if a malignancy could be detected before it has a chance to









metastasize, then a purely local surgical intervention should provide a complete and Einal

treatment [the continued existence of risk factors notwithstanding].

2.3 Folic Acid and Receptor Targeting

2.3.1 Folic Acid

Folic acid (also known as vitamin B9 Or B,), is needed by cells in the synthesis of

DNA nucleotides and it cannot be synthesized by the cell. Cell populations that are

growing rapidly (i.e. tumorous growths) require more folic acid, and upregulate the

production of folic acid receptors (Antony 1996; Suh et al. 2001). The resultant

overexpression of folate receptors is exploited in cancer cell targeting by attaching folate

to genes, contrast agents and drugs. An example is shown in Eigure 2-11, where an

anticancer agent linked to folate makes use of the folate receptor in two ways: first by

selecting the proper cell by attaching to the folate receptor, and then gaining entry into

the cell via receptor mediated endocytosis in a 'Troj an Horse' strategy.

Folic acid, as shown in Eigure 2-12, is composed of three molecular fragments:

(left to right) 6 methylpteridin, P amino benzoic acid (PABA) and glutamic acid. The

methylpteridin end of the molecule is the biochemically active moiety that participates in

the folic acid cycle (Campbell 1991; Campbell and Smith 1994) and through which the

receptor binding occurs. In order to preserve receptor function, folate conjugation has to

occur at the y-carboxylate group of the glutamic acid fragment (Leamon and Low 1991).

Tritium labeled folate attached via the y-carboxyl group demonstrated the same rate of

endocytosis as free folic acid, while attachment via the amine group of the methylpteridin

fragment abrogated the receptor binding (Kranz et al. 1995; Wang and Low 1998).



































Figure 2-11:. Mechanism of receptor mediated endocytosis used to target anti-cancer
drugs to tumourous cells. Sourced: Purdue News (2002), "Researchers
developing technology to outsmart metastasized cancers",
http://www. purdue. edu/UNS/html4ever/0206 19. Low.Endocyte. html, April
2004




II II




Figure 2-12: Structure of Folic Acid

2.3.2 Folate Receptors

Folate receptors exist in a number of isoforms: FR-u, FR-P, FR-y and a soluble

form (sFBP: soluble folate binding protein). Of these, only the first two are relevant to

targeting applications. FR-a was recognized first as a tumour marker, with cancers of the


Conlugate


gg









ovaries, kidneys, uterus, testes, brain and colon, as well as adenocarcinomas of the lungs

being found to prominently overexpress the a isoform of the folate receptor. Significant

expression of FR-a on nonmalignant cells occurs only in the kidneys, lungs, choroid

plexus and placenta, where in all but the latter case the receptors were situated on the

apical membrane surface. The primary folate transport mechanism for normal cells is

thought to occur via the reduced folate carrier a plasma membrane transport protein that

reacts preferentially to reduced folate but has no capacity to transport folate-drug

conjugates, since it is not anchored to the cell membrane as are the FR-a and FR-P (both

FR-a and FR-P are glycosyl phosphatidylinositol linked membrane proteins). FR-P is

overexpressed in many cancers of the breast, brain, mesenchymal tissue, testes, head and

neck (squamous cell) and hematopoietic cells granulocyticc lineage). FR-P has a tenfold

lower folate binding affinity than FR-u, with KD ~ 10-9M and 10-10M respectively (Reddy

and Low 1998; Wang and Low 1998).

Estimates on folate receptor overexpression rates in malignant tissues are quite

variable, and may not distinguish well between heterogeneous cell populations that are

grossly part of the same complex. Literature indicates that the expression of folate

receptors for squamous cell carcinoma of the oral mucosa exceeds the rate of expression

on otherwise identical, non-malignant tissues by >20 fold (Ross et al. 1994)

2.4 Fluorescence

Histology is the study of the microscopic structure of plant and animal tissue. It is

typically performed by staining tissues to highlight features of interest, or increase

contrast between features. The use of dyes permits the more sensitive detection of

features. The use of fluorescent dyes permits further increases in sensitivity that can









enable visualization of subcellular features. Fluorescence microscopy makes use of

fluorescent dyes, in many cases conjugated to specific binding agents, to generate images

that uniquely highlight specific features.








8,:







Figure 2-13: Jablonski diagram showing energy states for a fluorescence process.
Reprinted with permission from Molecular Probes, Haugland, R. "Handbook
of fluorescent probes and research products", web edition,
http://www.prob es.com/handb ook/Higure s/0664.html, January 2004

2.4.1 The Fluorescence Process

Fluorescence is found mainly in heterocyclics or polyaromatic hydrocarbons. It is

a three stage process depicted in the Jablonski diagram of Eigure 2-13 (Haugland 2003).

An energy input in the form of a photon of light huEX excites an electron from the So

ground state to the S1' excited singlet state energy level (1). The excited singlet state

exists for a short amount of time (1 10 nanoseconds) before decaying into the relaxed

singlet excited state S1 (2). When the electron returns to the ground state a photon of

energy hUEM is emitted (3). The energy dissipation between S1' and S1 (huEX hUEM) is

the Stokes shift the difference between the energy of the photon causing excitation

hUEX and the photon emitted hUENI. The emitted photon is at a lower energy and thus at a

longer wavelength. This phenomenon is partly responsible for the sensitivity that can be









achieved by fluorescence techniques, since it allows the separate detection of emission

electrons against a low background since the excitation signal is at a different

wavelength that can be removed (i.e. by optical filters). Another factor that contributes to

the high sensitivity achievable is fluorophores are not consumed in fluorescence and can

be repeatedly excited to generate many thousands of detectable photons. This is because

the excitation emission process is cyclical, and the excited electron returns to the

ground state via path (3) in figure 2-13 unless photobleaching occurs (essentially an

oxidative degradation process that can occur in the excited singlet state Si'). For a

polyatomic molecule in solution, the discrete energy level transitions depicted in the

Jablonski diagram of figure 2-13 are replaced by broader energy spectra for excitation

and emission. These spectra are depicted in figure 2-14, which also shows the

equivalency between the emission and excitation intensities: excitation at intensity EX 1

results in emission at that same intensity, while excitation at EX 2 yields emission at EM

2. This same figure also illustrates that the fluorescence emission spectrum is largely

independent of the excitation wavelength.

2.4.2 Fluorescence Techniques

Fluorescence techniques enable large numbers of very interesting biological

characterizations and analytical processes making them big business for the scientific

community. There are sizable companies specializing in the production of fluorophores

(Haugland 2003), especially high value added products such as fluorophores conjugated

either to general reactive groups for easy applications in individual research, or

fluorophores labeled proteins (and nucleic acids, lipids, etc.) known to be important to

common research and diagnostic applications.











Excitation Emission an
spectrum EX 1 EM 1 spectrum I &




r, I
Wavlegt

Fiur -1:Poyaoicmleul peta hoig xittinan mison nest
eqivlec Rerned wit pemsso frmMlclrPoeHulnR









Finformain where identific mlcl pcr hwn xiation ofd itms egos rrec ionso intersismnsitoeby

ther fuoescvaence signls nte a t mcoopricscae.Fow cytom oetelrs moeas Hure luo ,resec

of inividaldoo patile (cels orscetl)ia stroe am, peearmitin qouantificationofable

groups wthin a largro poplaton, ut and only exitue a/6~ t asnle fixedr wavlenth

Fluoesne sanne ors pidenytiyp fluorescence sinasfrom schsorcs tas elropie hoesaisi

goaels spaialy ont aye mcosinopi saleo. Spetrfluorometer mearsurpe anaerage signal

coing from a bulk sample (such that the data acquired reatosebsthspcr of fneeti oigured 2-

14ei mlore than te discet neryans of firsoigurae 21. FlwcThmtr mae fluorescencespcrmtri

perap thevia most flexibele mofthese in astrments sine it poides qacntinuous rang of ae

excitawtion an lremissonwavelength that can bnye moitoed aThasie absoptio speectromte



coisg acommonu istrmpentuc that sue toh dt acquired smlr databuts nth include in thisr lis






of fluorescence instrumentation since it does not measure emission. The excitation










spectrum of a fluorophores species is essentially the same as its absorption spectrum, but

the absorption spectrum can be influenced by particulate phenomena such as small

particle based scattering.

Interesting applications of fluorescence detection in molecular biology include the

use of green fluorescent protein (Tsien 1998) as a tag for gene expression, as well as

fluorescence in situ hybridization (FISH) for gene mapping, etc. (Haugland 2003).

Systems have been designed that enable four to five different fluorescent signals to be

resolved with optical filters only more when linked to interferometers and sophisticated

software. Multicolor labeling experiments involve the introduction of two or more

probes for resolution or monitoring of different obj ects or reactions. This technique is

widely applied in FISH, flow cytometry, DNA sequencing and fluorescence microscopy

(Herman 1998). Selecting dyes with narrow bandwidths and maximum spectral

separation maximizes resolution. Figure 2-15 shows the emission spectra of three dyes

commonly used in fluorescence microscopy that have good spectral separation:

fluorescein, tetramethylrhodamine and texas red. Figure 2-16 shows an image created for

a human metaphase chromosome hybridization experiment, using red and green

fluorophores.

2.5 Polymerization Methods for Producing Microspheres

The definition of polymerization types is predominantly phenomenologically

based. The methods of interest are heterogeneous processes leading to the formation of

small particulates (up to 1 or 2 mm in size) by means of addition polymerization of

vinyl-type monomers. The initial reagents for all systems include a monomer (or

mixture of monomers) to be polymerized, an initiator and a polymerization medium (a

liquid phase). In many cases there is a stabilizer added to the mixture--which may be









called an emulsifier in some systems. Other additions to the mixture include a

crosslinker if the final product is desired to be insoluble, as well as dopants or additives

to impart desirable properties (such as magnetic response, colour or radio-opacity) or

chemistries (i.e. surface chemistries that facilitate binding reactions for labeling or

targeting) to the particles.




0..TMR lTA










I I I


soo sno soo05
Wavelength (nrn)

Figure 2-15: Normalized fluorescence emission spectra of fluorescein (FL),
tetramethylrhodamine (TMR) and Texas Red (TR) dyes. Reprinted with
permission from Molecular Probes, Haugland, R. "Handbook of fluorescent
probes and research products", web edition,
http://www.probes.com/handbook/figures/083.html, January 2004

There are four fundamentally different techniques commonly employed for the

formation of particulate polymers from monomers in solution. These are emulsion,

dispersion, suspension and precipitation polymerizations. These can be distinguished by

several factors, including the size range of particles each can generate, the physical

process applied (including reactor design), and the manner in which polymerization takes

place within the system. Solvent evaporation is also a popular process, but this is not a










polymerization process, since polymer is dissolved in solvent and reshaped rather than

monomers being polymerized. There are also instances where the above processes are

performed with seed particles.



















Figure 2-16: fluoro probes hybridized to human metaphase chromosomes. Reprinted
with permission from Molecular Probes, Haugland, R. "Handbook of
fluorescent probes and research products", web edition,
http://www.prob es.com/handb ook/figure s/0674.html, January 2004

2.5.1 Emulsion Polymerization

Emulsion polymerization is known to produce the smallest particle sizes of the

methods mentioned typically in the range from approximately 50 nm to just below 1

Clm. The monomer is only slightly soluble or insoluble in the medium, but a surfactant is

added to allow emulsification. An initiator that is soluble in the medium, but not the

monomer is used. The monomer exists in the medium as both droplets and as surfactant

- enabled micelles of nanometer size. Initiated oligomers eventually produce stabilized

nuclei, and become the loci of polymerization for the system. The original micelle size

does not determine the eventual particle size produced since the initiator is not soluble in

the monomer. The product is formed by the stabilized micelles containing radical -









bearing oligomers, and these tend to grow to largely the same size, so that nanospheres of

low polydispersity can be formed.

2.5.2 Soapless Emulsion Polymerization

Soapless or emulsifier free emulsion polymerization is very descriptively

named; in the absence of micelle formation, nucleation occurs by precipitation of radical

bearing oligomers and macromolecules. These coalesce since they are not stabilized in

any way. Literature indicates that growing particles eventually are somewhat stabilized

by the charges on their chain end groups forming electrostatic charges. Typically, the

product size is slightly larger than for emulsion polymerized particles (up to micron size),

and the initial monomer concentration is slightly lower.

2.5.3 Dispersion Polymerization

A dispersion polymerization utilizes an initially homogeneous mixture of medium

and stabilizer, in which the monomer and initiator are soluble. Phase separation of the

growing oligomers / polymer radicals occurs because the medium is designed to be a

poor solvent for the polymer forming the nuclei of the primary particles. The locus of

polymerization for this type of system lies in the precipitated nuclei, as they swell with

monomer from the homogeneous medium / monomer mixture. This mechanism leads to

spherical particles that range in size from approximately 100 nm to 10 Clm. The stabilizer

is a critical component in this system without it the particles tend to coagulate during

formation. Stabilizers for dispersion polymerizations are polymer compounds with low

solubility in the medium and good affinity for the polymer particles formed. Particle size

control is affected by identity and concentration of the stabilizer, as well as by control of

the polymerization temperature and adjustment of the medium to alter the degree of

polymerization at which the particles precipitate out of solution. The medium is often an










aqueous alcohol mixture (with the solubility parameter adjusted by means of water

content) tailored to compatibilize the solubility parameter of the medium with the

monomer. It is possible to produce particles with very low polydispersity via dispersion

polymerization.

2.5.4 Precipitation Polymerization

Precipitation polymerization is initially like dispersion polymerization in that the

reaction mixture is a homogeneous solution. The important difference is that once the

particles precipitate out of solution, they are not swollen polymerization takes place

entirely within the continuous phase. The precipitated nuclei coagulate to form irregular

particles within a size range that is similar to dispersion polymerization at approximately

100 nm to 100 Clm.

2.5.5 Suspension Polymerization

Suspension polymerizations are generally capable of forming a product that is

from a few Clm to a few mm in size, with significant polydispersity in the eventual size of

the formed microspheres. An initiator that is soluble in monomer is used, and the

resulting monomer / initiator mixture is insoluble in the polymerization medium. The

monomer mixture is inj ected into medium, where the small droplets that form essentially

comprise 'microreactors'. The reaction can be seen as a small-scale bulk polymerization

within the droplet, with excellent heat transfer characteristics due to the large amount of

surface area available. If the monomer mixture is prepared with a diluent, then the

droplet can be seen as a small-scale solution polymerization.

The size of particles formed in suspension polymerization is controlled

predominantly by stirring and monomer injection. Anything that acts to create smaller










droplets initially or that acts to break up existing droplets leads to the formation of

smaller particulates by suspension polymerization. Other factors that influence the size

of particles produced include the monomer to suspension medium ratio, the stabilizer

concentration and the viscosities of monomers and medium. The control of a suspension

polymerization is based largely on empirical observation and iterative solutions due to the

complicated interaction of these parameters, making it as much an art as a science.

The morphology of particles produced by suspension polymerization relates to the

swelling reaction of the polymer to the monomer. Particles composed of polymer that is

swellable / soluble in its monomer tend to produce a smooth surface. A rough surface is

produced on particles that are not swellable by their monomer. Particle porosity can be

tailored by the addition of suitable monomer diluent (causing porosity during particle

formation), and control of the amount of crosslinking (to retain porosity on removal of

diluent) .















CHAPTER 3
PROPOSED STRATEGY AND DESIGN REQUIREMENTS

3.1 Appropriate Uses of Screening Tests

Factors that should be considered in the application of a screening test include

socioethical as well as scientific parameters (Grimes and Schulz 2002). Whereas

screening in modern society is correctly viewed to be a valuable medical tool, it is

important to note that not all screening tests that are technologically feasible would

provide any statistical benefit to the population. A good screening test should be

sensitive, applicable to a defined at-risk group, and it must hold reward for both the

patient and the community. Both the individual and the community must see benefits in

terms of improved overall health (statistically), and in terms of reduced expenditures for

health care.

Screening must be understood to have ethical implications that arise from the

fundamental difference between a diagnostic and a screening test. Diagnostic medical

tests are performed in response to a patient complaint. Screening is performed on a

predominantly healthy population. The negatives associated with screening tests are the

expense involved, patient inconvenience / morbidity (i.e. colonoscopy), and the

possibility of false positive results (which bring fiscal and emotional consequences).

This leaves aside the possibility of false negatives, which at least leave the patient in

substantially the same position as before being screened. The positive aspects of

screening must outweigh these negatives, by improving general health and reducing total

healthcare costs.









Screening test sensitivity and specifieity are critical parameters, as is the disease

prevalence in the population being screened. The positive predictive value of a screening

test is equal to the number of true positives divided by the number of total positive test

results (see the treatment matrix of Eigure 3-1, and equations 3-1 through 3-3). The

higher the number of disease-negative subj ects that are tested, the lower the positive

predictive value of a test. Put simply, the positive predictive value of any given test is

higher when the incidence of said disease is higher in the tested population. This leads to

the conclusion that a screening test is more appropriate for a condition for which an at-

risk population can be readily defined. In the case of screening tobacco users this criteria

is indeed met.

DISEASE
Positive Negative


True False
.1 positive positive



eld
False True
negative negative




Figure 3-1: Results matrix for disease state vs. test result

Specifieity = d / (b + d) (equation 3-1)
Sensitivity = a / (a + c) (equation 3-2)
Positive predictive value = a / (a + b) (equation 3-3)

The situation of false positive results is highly dependent on the disease. In some

cases a false positive diagnosis can have grave impact on a patient such as a screening

test for STDs when a false positive would not only traumatize the patient but have the










potential to wreck a marriage. A false positive robs the patient of their perceived health,

and diverts precious resources for the unnecessary treatment of healthy individuals -

highlighting the need for sensible application of any screening test within a suitable

environment.

A further requirement for the sensible application of a screening test is that an

earlier diagnosis improves the patient' s prognosis again a condition that is met with the

target group and condition, where early detection can lead to a simpler and more

successful treatment.

3.2 Proposed Testing Procedure

3.2.1 Components

The 'brick wall' structure in figure 3-2 represents epithelial cells, with normal

cells depicted as unfilled black rectangles and tumorous cells shown as containing cyan

(pink) ovoids.

Open circles = microspheres with dye A, no surface folate = controls

Closed circles = microspheres with dye B, folic acid coupled = sample

3.2.2 Procedure

Performed after a dental cleaning to limit confounds.

Step 1: Equal portions of microspheres A and B are administered as a suspension

in mouthwash and swished for 30 s to ensure proper distribution through the oral cavity.

Step 2: The subject expels the mouthwash of step 1, and rinses with water several

times to remove entrapped microspheres. Folate bound microspheres (B) should adhere

to tumorous cells by physically binding to folate receptors expressed by the cells.










O = microspheres with dye A, no surface folate
, = microspheres with dye B, folate immobilized

True positive True negative
Step 1) Add microspheres solution


O
O -
OO
tumorous cells
normal cells


O C
Or
S OO


Step 2) Remove unbound microspheres by wash


Step 3) Remove bound microspheres with acid
wash and capture magnetically


acid wash


O





Step 4) Analyze ratio of A to B microspheres


Ratio: CL

positive


5vs. O


negative


Figure 3-2: Steps of the proposed testing strategy










Step 3: An acid wash releases the receptor ligand binding, and a rinse

recaptures the microspheres. These can now be collected magnetically, and assayed

spectroscopically to determine the ratio of binding microspheres (B) to controls (A).

Since some portion of microspheres will always be retained in the oral cavity by

entrapment, the control particles are essential to prevent false positives. Entrapment

should retain equal amounts of A and B, and any excess of B will be due to receptor-

mediated binding.

3.3 Design Parameters for Microspheres

Microspheres synthesized for the application described above require certain

properties to accomplish the task. These properties, and the manner in which they can be

designed into the microspheres were optimized for the particular application described.

The design approach and the chemistry involved are very broadly applicable however,

with only minor modifications to the immobilized ligand and microsphere optimization

required to adapt the 'system' to other applications.

3.3.1 Microsphere Size Considerations

The 'ideal' size of microspheres for this application was initially unknown.

Known parameters that helped determine the boundary conditions are shown in the

diagram of figure 3-3.

The lower boundary is based on endocytosis. Folate receptors exist to conduct

folic acid into the cell, and have been shown to endocytose drug molecules (i.e. proteins)

and small (100 nm) particles (Wang and Low 1998). The upper size limit of particles

that are endocytosed represents the smallest particle that could be recovered thus the

smallest particle size that could be of interest to the application. The lower bound on









microsphere size was postulated as being on in the single Clm range--roughly from 2 5





Minimum size
to prevent
a~ 8 endocytosis ~ 2-5 um





Specific surface area


0.1 1.0 10 100 1000
log size in um


Figure 3-3: Size considerations for microspheres

The upper bound on size has several components. First, as the microspheres

increase in size the specific surface area decreases--yielding less area to immobilize the

binding agent on. Second, larger particles will be increasingly susceptible to drag-

induced shear stresses due to fluid movement near the surface. Both these phenomena

dictate that larger particles become increasingly difficult to retain via physical binding to

a surface. The upper boundary of microsphere sizes that could be retained by receptor

mediated binding is postulated as 80 100 Clm.

The size considerations represented a theoretical starting point for devising a

suitable polymerization process for microsphere production. It was postulated that if a

broad size distribution of microspheres in the above size range were produced, then the

microspheres in the 'correct' size range would be the same population that adhered to the

cells of interest. This empirical solution made a suspension polymerization process the









obvious choice, due to the broad distribution of particle sizes within the range of single to

~100 Clm that can be achieved.

3.3.2 Ligand Immobilization

The microspheres must be capable of having folic acid immobilized onto their

surface, with retention of receptor recognition. For folic acid this requires coupling via a

carboxylic acid residue on the glutamic acid end. To enable the required coupling

chemistry, a monomer with a terminal amine functional group that can couple to the

carboxylic acid group is polymerized into the microspheres.

3.3.3 Magnetic Guidance

The application requires sufficient magnetic loading to permit magnetic recapture

of the microspheres for analysis. The nature of the magnetically responsive species must

be superparamagnetic to avoid agglomeration of the microspheres. Viable approaches to

forming magnetically responsive microspheres include doping monomers with magnetite

prior to polymerization, polymerization in the presence of a ferrofluid, or in situ

precipitation of magnetite onto microspheres. For this application magnetite was doped

into the monomers prior to suspension polymerization, and the microspheres were

collected magnetically ensuring the exclusive retention of magnetically responsive

microspheres.

3.3.4 Dye labeling

This application requires the preparation of microspheres labeled with two readily

distinguishable dyes. The usefulness of the control particles would be diminished if the

dyes were not incorporated in the same manner, or if the dye identity influenced the

surface chemistry of the particles. Two approaches to dye labeling were investigated:

microsphere swelling and dye loading, and conjugation of dye to monomer prior to






40


microsphere polymerization. The dyes were chosen based on spectral separation and

fluorescence intensity, so that they could be both detectable and distinguishable.















CHAPTER 4
MATERIALS AND METHODS

4.1 Magnetic Material for Microspheres

The microspheres require a magnetic component to enable magnetic recapture and

separation. The magnetic component had to be superparamagnetic magnetite in order to

maximize the magnetic response and to avoid agglomeration of particles.

Superparamagnetic iron oxide was produced and coated with oleate and polydimethyl

siloxane (PDMS) to facilitate dispersion in the main monomer: methyl methacrylate

(MMA). This doped monomer was used in the suspension polymerization process in

exactly the same manner as pure monomer would have been.

4.1.1 Materials used in Iron Oxide Preparation

Oleic acid (Aldrich 36,425-5, [112-80-1]), polydimethyl siloxane (PDMS, 10 cs,

Hidls America PSO39, [63148-62-9]), ammonium hydroxide (25 % aqu solution, Acros

255210025, [1336-21-6]), cyclohexane (Aldrich 15,474-1, [110-82-7]), hydrochloric acid

(Acros 12463-0010, [7647-01-0]), chloroform (Fisher C297-4, [67-66-3]), ferric (iron II)

chloride tetrahydrate (Aldrich 22,029-9, [13478-10-9]) and ferrous (iron III) chloride

hexahydrate (Aldrich 20,792-6, [10025-77-1]) were all used as received. Methyl

methacrylate monomer (MMA, Aldrich M55909, [80-62-6]) was distilled at reduced

pressure to remove inhibitor.

4.1.2 Magnetite Production and Treatment

All magnetite used was produced in our own laboratory, in a process that has been

described extensively by Leamy (2003) and was originally adapted from methods of









Robineau and Zins (1995). Briefly, stable suspensions of iron oxide were formed by the

base catalyzed precipitation of iron chlorides. Ferrous and ferric chlorides were

dissolved in water and hydrochloric acid (HC1), and added to stirring ammonia solution

to induce precipitation of iron oxides. Iron oxide product was retained magnetically

while most of the aqueous supernatant was decanted. Oleic acid (10 wt % relative to the

oxide) in cyclohexane was added to the magnetic slurry to coat the iron oxide and render

it hydrophobic. Addition of methyl alcohol permits a solvent exchange to cyclohexane as

iron oxide falls to the bottom and the aqueous/alcohol phase can be decanted. The iron

oxide was air dried overnight and suspended in chloroform. PDMS (10 wt % relative to

the oxide) was added at this stage. The solvent was again evaporated off by air drying to

recover the solid product, which was dispersed in monomer.

4.1.2.1 Method of iron oxide precipitation

An aqueous solution of iron oxides was produced by combining 2.03 g ferrous

iron chloride tetrahydrate (FeCl2*4H20) with 4.88 g ferric iron chloride hexahydrate

(FeCl3*6H20) and 0.887 ml 37 % aqueous HCI in 20 ml distilled water. A 250 ml beaker

was filled with 8.3 ml aqueous 28-30 % NH40H in 155 ml ofDI water. This solution

was mechanically mixed for 5 min by a 4-blade stainless steel stirrer (PT1035 stirrer,

Kinematica GmBH, Germany), set to approximately 400 rpm. The iron oxide solution

was poured into the 250 ml beaker and left stirring for 10 min., after which stirring was

stopped and a 1 inch square rare earth magnet (NeFeB) under the bottom of the beaker

was used to collect the magnetite produced. The supernatant was decanted, leaving

approx 60 ml magnetite solution, which was stirred at low speed (approx. 200 rpm).









4.1.2.2 Method of coating iron oxide

The minimum effective oleic acid concentration had been determined previously

in cooperation with Leamy (2003), as 10 w/w %. 0.221 g oleate dissolved in 11.14 ml

cyclohexane was added to the slurry of iron oxide under stirring. The role of oleate is to

coat the iron oxide particles resident in the aqueous phase and render them hydrophobic,

thus forcing segregation into the cyclohexane phase. 15 ml of methanol was added after

stirring for 15 min, reducing the density of the aqueous phase to permit the ferrofluid in

cyclohexane to segregate to the bottom of the beaker more easily. The cyclohexane

ferrofluid was retained in the beaker through placement of the same rare earth magnet as

above, under the beaker, while the aqueous supernatant was decanted. 10 w/w %

(relative to the iron oxide content) polydimethyl siloxane (PDMS) was added directly into

the cyclohexane ferrofluid slurry under agitation. Air-drying overnight in a petri dish

evaporated off the cyclohexane. The coated magnetite was dispersed in methyl

methacrylate (MMA) monomer (see structure in figure 4-1) by sonicating (Branson 2510)

and vortexing (Fisher Vortex Genie 2) repeatedly, at a concentration of 6 w/v % (of iron

oxide, not including weights of coatings). This product was stored in the refrigerator

until further use.




H,C- C C -OCH,

CH,


Figure 4-1: Methyl methacrylate (MMA) structure









4.1.3 Characterization of Iron Oxide

The iron oxide produced was based on processes well characterized both in

literature and within our own labs (Leamy 2003). The magnetic properties were

established by Leamy using superparamagnetic quantum interference device (SQUID)

magnetometry. The physical size of the particles of iron oxide produced was

characterized by transmission electron microscopy (TEM) and by scanning electron

microscopy (SEM). The magnetic response was also anecdotally verifiable, as any

permanent magnet could strongly attract the microspheres.

4.2 Microsphere Polymerization and Characterization

The microspheres were prepared by a suspension polymerization process, using

doped monomers. The monomers were distilled as appropriate and processed to include

dopant species. The main backbone monomer, methyl methacrylate (MMA), had

magnetite dispersed in it as described above at a concentration of 6 w/v %. The

functional comonomer, aminoethyl methacrylate hydrochloride salt (AEMHS) was

conjugated to fluorescent species as described below in certain polymerizations. In some

instances, dye loading was performed by means of solvent swelling, in which case dye

loading occurred after polymerization and washing steps.

4.2.1 Incorporation of Fluorescent Dye Into Functional Monomer

Two separate strategies were applied to incorporate fluorescent dyes into the

microspheres. The strategy described here involved the covalent coupling of dyes

fluorescein isothiocyanate (FITC, structure shown in figure 4-2) or Texas Red-X

(structure shown in Eigure 4-3) to the functional copolymer utilized in the polymerization

reaction-aminoethyyl methacrylate hydrochloride salt (AEMHS, structure shown in

Eigure 4-4).










HIO h ,O


II







Figure 4-2: Fluorescein Isothiocyanate (FITC) structure



+



2Q O
3 liI
SSO NMH(CH2]5 --C -O-N





Figure 4-3: Texas Red-X (TR) structure

4.2.1.1 Materials for fluorescent dye incorporation into functional monomer

Fluorescein isothiocyanate (Molecular Probes F-143, Eugene OR), Texas Red-x

succinimidyl ester (Molecular Probes T6134, Eugene OR), absolute ethanol and 2

aminoethyl methacrylate hydrochloride salt (AEMH, Acros 357810250, [2420-94-2])

were all used as received.

4.2.1.2 Method of conjugating fluorescent dye to functional monomer

The dye-monomer conjugate was prepared by magnetically stirring (overnight) a

mixture of dye (FITC or Texas Red-X succinimidyl ester) and AEMH in solvent

(absolute ethanol for FITC and THF for Texas Red-X). The dye was conjugated to at









least 3.5x molar excess of monomer. The fluorophores were delivered with readily

reactive leaving groups so that conjugation to monomer was simple.




H,C= C -C-OCHCH,NH, HCI

CH,

Figure 4-4: Aminoethyl methacrylate hydrochloride salt (AEMHS) structure

The reaction of FITC with AEMH is illustrated schematically in figure 4-5. The

isothiocyanate group reacts with the primary amine on the monomer. The reaction occurs

by attack of the nucleophile on the electrophilic central carbon of the isothiocyanate

group. This forms a thiourea linkage between the FITC and AEMH with no leaving

group (Hermanson 1996).

AEMH -NH2+ S=C=N--FITC
monomer isothiocyanate
with primary fluorophore
amine


AEMH-N--C--N-FITC
H H

Isothiourea bond

Figure 4-5: Mechanism of FITC conjugation to AEMH monomer

The reaction of Texas Red-X with AEMH is shown schematically in figure 4-6.

The succinimidyl ester of the Texas Red reacts with the primary amine on the monomer

to form a stable amide bond, and the N-hydroxysuccinimide forms the leaving group

(Hermanson 1996).









Texas Red
C= O
AEMH-NH2 O

monomerO N O
with primary
ammne "
succinimidyl ester
fluorophore


O OH
AEMH-N-C= Texas RedO N O
H---_

Amide bond NHS leaving group

Figure 4-6: Mechanism of Texas Red-X conjugation to AEMH monomer

Experiments were performed to optimize the fluorescent dye concentration (using

FITC) by conjugating different amounts of dye to 100 mg AEMH. For regular

production of microspheres the dye was conjugated to either 20 or 100 mg AEMH. The

stirring mixture was protected from light to avoid photobleaching of the fluorophores.

The ethanol was removed either by rotovapor, or by flash freezing the monomer-dye

conjugate in liquid nitrogen and then drawing the solvent off at reduced pressure and

temperature. The monomer-dye conjugate was dispersed in 980 or 900 mg AEMH, to

bring the total amount of monomer to 1000 mg, and mixed at ~3 50 C to ensure that

mixing occurred intimately in the liquid phase. The AEMH monomer with dispersed dye

conjugate was used together with the magnetite doped MMA monomer in the suspension

polymerization recipe.









4.2.2 Suspension Polymerization Procedure

The suspension polymerization procedure was adapted from Shim and Kim who

devised a simple but effective suspension polymerization protocol to produce methyl

methacrylate particles incorporating inorganic oxides for sunscreen applications(Kim et

al. 2002; Shim et al. 2002).

4.2.2.1 Materials for suspension polymerization

Polyvinyl alcohol (PVA, 87-89 % hydrolyzed, 85 146,000 MW, Hoechst-

Celanese) was used as delivered. The originally applied Airvol 523 polymer was

supplanted (by the producer, who indicates that the new Celvol product is precisely the

same as the old Airvol equivalent) by CelVol 523 and CelVol 823--the latter being a

new easier-dissolving grade that was adapted for use in all polymerizations subsequent to

lot S19. Monomers were used as prepared above. Crosslinker divinyl benzene (DVB,

Sigma D-0916 lot 50K3 652, [1321-74-0]) was NaOH washed three times. Initiator Azo-

bis isobutyronitrile (AIBN, Aldrich 44,109-0, lot 02612HI [78-67-1]) was recrystallized

in methanol. Additional monomers aminoethyl methacrylate (2-aminethyl methacrylate

hydrochloride, AEMH or AEMHS, Acros AC357810250, [2420-94-2]), aminostyrene

(AmST, Acros AC30848, [1520-21-4]) and Maleic anhydride (MA, Daj ac 7579, [108-3 1-

6]) were used without further purification.

4.2.2.2 Method of suspension polymerization

The suspension medium was 100 ml of 2 w/v % aqueous polyvinyl alcohol (PVA)

solution, prepared by dissolving celvol 823 PVA (89 % hydrolyzed, 85-146,000 MW) in

distilled water and stirring mechanically (Caframo stirrer RZR1, Ontario Canada) at

moderate speed overnight at 700 C. This mixture was prepared in a 3 necked 300 ml

round bottom flask, with a Teflon stirring rod admitted through the central neck, and a









thermometer fitted for temperature monitoring, and the third neck capped with a rubber

stopper when not in use for reactant addition. Heating was provided by a heating mantle

controlled by a Variac power adjuster. The monomer mixture was prepared by

combining the monomers to be polymerized in a 20 ml glass vial. In most cases 80 90

v/v % 'main' monomer--always methyl methacrylate, was used with 10 20 w/v %

functional monomer maleicc anhydride, aminostyrene or aminoethyl methacrylate). An

additional 2.5 v/v% divinyl benzene (DVB) was added as crosslinking agent, as well as

2.5 w/v % azobisisobutyronitrile (AIBN) as initiator. These ingredients were combined

at mildly elevated temperature (~400 C) to prevent coagulation of AE1VH, and mixed

until homogeneously dispersed.

The suspension medium temperature was elevated to 750 C, and the stirrer was

switched for a homogenizer (Kinematica PT1035, Germany) in later suspension

polymerizations, with speed adjusted so that mild or no frothing of the mixture

occurred--a setting of 4 on the controller. The erratic electric supply of the building

required that an uninterruptible power supply with automatic voltage regulation (APC XL

series 1500 with voltage boost and trim functions to limit variation < 5 % in voltage

output) was installed to prevent the homogenizer from either stopping due to low voltage

or frothing the mixture excessively due to overvoltage. The monomer mixture was

injected through a large gauge stainless steel needle affixed to a syringe. After injection

the suspension mixture was left stirring for 4-8 h at 750 C, after which time the heating

mantle was switched off. The mixture was left stirring until it had cooled to near room

temperature, then decanted into 50 ml centrifuge vials for washing and storage. The

polymerization apparatus is shown in figure 4-7 below.









4.2.3 Microsphere Post-Polymerization Processing

The microspheres were recovered by magnetic separation and three washes with

distilled water. The washed microspheres were stored in solution, away from light (for

those microspheres that had either dye content or had folic acid immobilized onto them).

Magnetic separation was performed using custom produced magnetic separators (see

figure 4-8), which were largely the work ofLeamy (2003). These permitted efficient

separation of the microspheres from medium at high speed (seconds to minutes).

Microspheres that were not covalently dye labeled were ready for swelling with dye at

this stage.

























Figure 4-7: Suspension polymerization setup using mechanical stirrer and heating
mantle. Panel a shows the Caframo stirrer, with teflon paddles. Panel b
shows the Kinematica homogenizer.

Some microsphere batches were prepared with maleic anhydride, which

polymerizes across its inner n: bond (with free radical initiation). This ring is readily









reactive towards primary amines, forming an amide bond and a carboxylate group on the

other opened site. This carboxylate group can in turn react with another primary amine in

a condensation reaction to form a second amide bond.

The method of performing the maleic anhydride modification was to add excess

ethylene diamine (approx 4 ml for <1 g microspheres) to microspheres from which water

has been removed by magnetic separation (excess water was removed, but no further

drying was done), stir for 15 min, and wash repeatedly with DI water.

















Figure 4-8: Magnetic separator apparatus. The steel cylinder contains four magnets held
in place by a machined insert. Magnets are magnetized through their width:
opposite faces have like charges and adjacent faces have opposite charges.

4.2.4 Microsphere Dye Loading by Swelling and Solvent Evaporation

Solvent was chosen by assessing the microsphere swelling (in solvent), and dye

solubility (in solvent).

Microsphere swelling tests were performed using a Zeiss Axioplan 2 imaging

microscope. Microscope slides were affixed with double sided tape, and dried

microspheres were scattered at low concentration onto the tape. Analysis was performed

using a 10x objective, with 1300 x 1030 pixels used. The camera was operated in time

lapse mode at one frame taken every 12 seconds for 10 20 minutes. The time was









determined by a test sequence for ascertaining the time to ultimate degree of swelling.

The images were then sized using Zeiss Axiovision 3.1 software with scaling module.

The solubility of dye in solvent was assessed by adding 1 w/v % dye into solvent

and vortexing (30 s) and sonicating (5 min), repeating the process twice. Evaluation was

based on visual observations of precipitates / undissolved fractions of dye remaining in

solvent after mixing.

Triple washed microspheres were resuspended in dye/solvent solution by

vortexing (30 60 s) and sonicating (10 min) several times. The microspheres were left

in the dye solution overnight. Microspheres were collected magnetically, and supernatant

decanted, after which they were dried in a vacuum oven overnight at low temperature

(~400 C). Dried particles were water washed repeatedly until no further dye was visible

in the supernatant after washing. The microspheres were then washed 3 more times.

4.2.5 Microsphere Characterization

The microspheres produced were universally characterized for size, since the

suspension polymerization process has a significant potential for size variation. Several

standard techniques were applied. All batches produced were sized using a Coulter light

scattering particle size analyzer, which provides a size histogram of the particles tested.

Most batches of particles were also sized by light microscopy, and in some cases electron

microscopy was used, providing additional shape and surface data, including surface

morphology.

The microspheres were also characterized for fluorescence response when doped

with fluorophores. A set of experiments to provide some optimization data for

fluorophores concentration was also conducted. Evaluation was provided by

fluorescence spectroscopy.









4.2.5.1 Coulter sizing

Size histograms were generated using Coulter instruments: initially a Coulter LS

230, and then its successor machine, a Coulter LS 13 320. Both were equipped with

small volume fluid modules, where the particles were sized suspended in water. The

procedure for testing involved selecting (or building) the appropriate optical model--a

collection of physical constants including the real and imaginary components of the index

of refraction for the material composing the particles, and the index of refraction of the

medium (water in this case). After acquiring a background measurement, the particles to

be sized were loaded, and measured in three consecutive discrete runs. The newer

instrument provided an automatically calculated average for all runs performed. All tests

were done with stirring at the minimum rate (30 % for the LS 230, 20 % for the LS 13

320), which prevented the particles from settling during the 90 s runs. In the event of

inconsistent results from consecutive runs the measurements were repeated.

The optical model built for the methyl methacrylate based magnetite doped

particles was constructed to use a real index of refraction of 1.5, and an imaginary index

of refraction of 5. These values were determined from the Coulter manual description of

indices for materials with similar properties when no data for the exact material was

available. The variation in results with the indices selected was projected as minimal,

since the impact of the indices on the sizing calculations decreases as particle size

increases over 1 Clm.

4.2.5.2 Light microscopy

For light microscopy particles were typically suspended in ethanol solution, to

enable quicker drying times onto a cover slide. A very dilute suspension of particles was










placed on a slide, which was then viewed under the microscope. All microspheres were

viewed in transmitted light mode. The microspheres with fluorophores could also be

usefully imaged by fluorescence microscopy. In all cases images generated for

characterization were acquired using a Zeiss Axioplan 2 imaging microscope and the

attendant Zeiss Axiovision 3.1.2 software.

4.2.5.3 Zeta potential analysis

A Brookhaven Zeta Plus zeta potential analyzer and particle sizer was used to

acquire information on the surface charge of the microspheres, both before and after the

immobilization procedure. Microspheres were dilutely suspended in pH 7.4 PBS.

4.2.5.4 Scanning electron microscopy (SEM) and energy dispersive spectroscopy
(EDS)

Samples were prepared for electron microscopy by drying (either in vacuum oven

or by freeze dryer), and sputter coating for 1 min with carbon. Images were then

acquired with either a JEOL 6400 scanning electron microscope or a JEOL 6335CF field

emission scanning electron microscope. Both microscopes also featured EDS

attachments with LINK ISIS software capable of generating spectra on constituents

present and semi-quantitative analyses.

4.2.5.5 Inductively coupled plasma spectroscopy (ICP)

A Perkin Elmer Plasma 3200 IRL inductively coupled plasma spectroscopy (ICP)

system was used to determine the concentration of iron in the microspheres. This was

interesting because the iron oxide concentration is a primary factor responsible for the

strength of the magnetic response.

Samples were prepared by dissolving the microspheres to atomic constituents.

Dissolution occurred either in a solution of hydrofluoric acid (HF) and sulfuric acid










(H2SO4), Or in an acid mixture sometimes called "aqua regia" onsisting of three parts

hydrochloric acid and one part nitric acid. The samples were compared to Fe standards to

establish an iron concentration in the sample.

4.2.5.6 X-Ray powder diffraction (XRD)

XRD was performed using a Philips APD 3720 in MAIC. Samples were

prepared by washing microspheres and drying in a vacuum oven. The resulting

diffraction patterns were analyzed by comparison to known patterns for possible

compounds present, as referenced from the internal database.

4.2.6 Microsphere Fluorescence Properties

Optimization was carried out on a series of suspension polymerized batches of

microspheres composed of the standard 80/20 mix of methyl methacrylate and

aminoethyl methacrylate monomers. 2.5 wt % initiator and 2.5 vol % crosslinker were

added, also the standard amounts. A table in the results section (table 5-6) shows the

amount of fluorescein isothiocyanate added to each sample, as well as the molar ratio of

fluorophores to monomers. The fluorophores to monomer ratio was calculated by

applying the assumption that all fluorophores and monomer in the feedstock was

successfully integrated into the microspheres. This assumption is not valid, but it does

provide a consistent error. The absolute numbers are probably off by some tens of

percent, but the ratios between the samples should have the same error to them so that the

comparison between samples remains valid.

In addition to assessing the fluorescence generated by each dye concentration, the

polymerized samples were used to assess the effectiveness of the dye incorporation

process by leaching out dye through a DMSO solvent wash. The DMSO supernatant was









examined for dye content and for the nature of the dye species that had been leached out

of the microspheres.

Fluorescence intensity was examined for both microspheres in aqueous solution,

and for DMSO washed microspheres. UV-absorbance spectroscopy was used to

determine the concentration of dye in DMSO. Fluorescence spectroscopy was used to

determine the identity of the dye molecules: whether they were native fluorescein

isothiocyanate or if they were conjugated to monomer or oligomers.

4.2.6.1 Sample preparation

Four samples of FITC-AEMHS conjugate were prepared. Each amount x of

FITC was conjugated to 100 mg of AEMHS in 2 ml of absolute ethanol, and left stirring

for 48 h, while protected from light exposure. These solutions were made with 2, 5, 10

and 20 mg of FITC. The solutions were flash frozen in liquid nitrogen, and placed under

vacuum within an ice bath for 8 h to remove the solvent. The monomer-dye conjugate

that remained had 900 mg AEMHS monomer added to it under gentle

heating-approximately 400 C. The dye-conjugated monomer was dispersed within the

liquid pure monomer by both mechanical stirring and sonication. Suspension

polymerization was carried out as for other samples (this series of samples was labeled

S21 through S24, in order of increasing FITC concentration).

The polymerized samples S21 through S24 were collected into two 50 ml

centrifuge vials each. Each vial was washed three times with 10 ml distilled water, with

the particles being magnetically separated in magnetic separators previously constructed

in our labs (Leamy 2003). The particles were magnetically retained and the supernatant

was decanted, after which the particles were redispersed in new media by vortexing (1










min) and sonication (10 min). 5 ml of each triple washed samples was withdrawn by

pipette the water decanted. The samples were dispersed in 5 ml DMSO by vortexing and

sonicating as above, then magnetically separated and the supernatant was collected. This

process was repeated for one wash after the DMSO supernatant was clear to the naked

eye of any dye. All supernatant was saved for analysis.

4.2.6.2 UV-Visible absorbance spectroscopy

UV-Visible absorbance spectrometry was performed on a Shimadzu 2401PC UV-

Visible recording spectrophotometer, with 6-cell attachment. Experiments for

characterization were conducted using spectrum mode, while concentration experiments

were done in photometric mode. PMMA cuvettes were used for aqueous media, while

Polystyrene cuvettes were used for DMSO based organic solutions.

4.2.6.3 Fluorescence spectrometry

Fluorescence spectrometry experiments were performed on a Fluorolog Tau-3

(Jobin Yvon Spex Instruments, S.A. Inc) with a 450 W xenon lamp as the excitation

source. For characterization of FITC fluorescence, emission was monitored at the 515

nm emission peak. Excitation was provided at the 488 nm excitation peak. When

particle based scattering or raman scattering caused increased in noise beyond the

discernible signal, it was possible to utilize lower excitation bands of fluorescein at 350

nm to generate a signal that was lower in intensity, but improved with regard to

signal/noise ratio.

4.2.6.4 Confocal microscopy

Confocal microscopy experiments utilized an Olympus IX81 microscope with

Fluoroview software to assess dye distribution within the microspheres. Images were

acquired using a default 20x obj ective, as well as a 60x oil-immersion obj ective (PLAPO















































-- -- - ------ -- ---- ------7------ -----T --------




-----~-------- --~--- ~ ---C- ------ - - -- -









O 350 400 450 500 550 600 650 700 7'


L
i


60XO). In each case optical and fluorescence images were acquired concurrently, with


excitation provided by a 488nm argon laser for fluorescein labeled samples, and a red


laser for Texas Red labeled samples.


4.2.6.5 Fluorescence microscopy


Fluorescence microscopy was performed using a Zeiss Axioplan 2 imaging


microscope with Axiovision 3.1 software. For fluorescence images, or combined (dual)


mode images appropriate filter cubes were utilized. When fluorescein was the


fluorophore, filter set 10 was used, with the spectral characteristics shown in figure 4-9.


This filter set is expressly sold for use with fluorescein.


- citation: BP 450 -490
-beamsplitter: FT 510
emission: BP 515 -565


(typical curves)


rm [
110









50





o0
-0


51


0


Figure 4-9: Spectra for Zeiss filter set 10, used for fluorescence microscopy of
fluorescein containing particles. Sourced from Carl Zeiss, Inc. (2002),
"Upright microscopes technical data",
http://www.zeiss.de/4 12568 1F004CAO25/Contents-
Frame/286BA4D22Bl14DEE985256B4A007C3686, January 2004


Filter set 10

488010 0000









4.2.7 Preparation of Microspheres for Cell Work

All microspheres that were used for cell work were repeatedly washed and

suspended in distilled water to remove any possible impurities. No further preparation

was made in cases when the cells would be disposed of directly after testing. If the cells

were to be used further the microspheres were sterilized by exposure to UV light for

several hours under a biological hood. A sample of UV-exposed microspheres was

compared to non UV-exposed controls by fluorometric methods to determine if the

sterilization exposure generated photobleaching meaningful enough to provide significant

reduction in fluorescence intensity.

4.2.8 Preparation of Microspheres by Dispersion Polymerization

Microspheres were prepared by dispersion polymerization using only a single

general method. This method was partly developed by Leckey (1997).

4.2.8.1 Method of dispersion polymerization

The solvent phase was prepared from a mixture of absolute ethanol and water

with stabilizer mixed in. The stabilizer used was polyvinylpyrrolidone, MW 40,000

(PVP40), utilized at a concentration equal to 12 w/v % relative to the monomers mixture

amount added to the solvent phase. This mixture was stirred for at least 5 min to permit

good mixing. The amount of water in the dispersion media was a primary lever that was

utilized to control the size of the particles formed--a typical composition was 80 v/v %

ethanol with 20 v/v % water.

The monomer phase was prepared by adding together the desired combination of

monomers--in most cases approximately (measured by volume) 2 parts styrene and 1

part diethyl aminoethyl methacrylate (DEA). Divinyl benzene (DVB) was added as

crosslinker, usually around 1 v/v %. The initiator used was azo-bis isobutyronitrile










(AIBN), at a concentration equal to 3 w/v %, and stirred for at least 2 min to fully

dissolve.

The monomer mixture was combined with the dispersion phase at 5 v/v %

monomers and 95 v/v % dispersion phase, to make 5 or 10 ml samples that were sealed

into glass vials of at least twice the capacity of their filling. The combination was

vortexed (30 s) and purged with nitrogen gas before sealing. The sealed glass vials were

placed in a pre-warmed shaking water batch (Haake SWB 25), with n = 100 cycles per

minute, and temperature set between 55 and 800 C, with the most common temperature

used being 700 C. The vials were zip tied diagonal to the direction of reciprocation, fully

submerged in water. In almost all cases the polymerization time was 24 h (+ ~ 4h).

Samples were removed from the polymerization apparatus and cooled. A crust of

polymer sometimes formed under the cap, and any agglomerations or crust formations

were disposed of. The microspheres formed were washed in DI water by centrifuging

and redispersing in water at least 2 x. All microspheres were sized by Coulter light

scattering.

4.2.8.2 Incorporation of magnetic species

Several different techniques of imparting magnetic response to the microspheres

were applied. These can be divided into processes that involved addition during

polymerization, and post-polymerization modifications.

When magnetic additions were made prior to polymerization, a portion of the

dispersion medium was replaced with a suspension of magnetite in water; in some cases

the magnetite was silica coated.









Post polymerization addition of magnetite occurred by base catalyzed

precipitation of iron oxide in situ. The base catalyzed precipitation of iron chlorides to

form iron oxide particles was discussed in detail by Leamy (2003).

4.2.9 Preparation of Microspheres by Activated Swelling

Activated swelling procedures were adapted from literature sources that detailed

procedures that were separately developed by groups headed by Tuncel and Frechet,

based on original development of the process by Ugelstad (Ugelstad 1978; Ugelstad et al.

1979; Ugelstad et al. 1988; Wang et al. 1994; Tuncel et al. 2002a; Tuncel et al. 2002b).

All procedures used seed particles that could be prepared by emulsion polymerization, or

as in the present research by dispersion polymerization. In almost all cases the seeds

were low polydispersity uncrosslinked styrene microspheres, and the swelling agent used

was dibutyl phthalate (DBP). The procedure is essentially a suspension polymerization

of preformed swollen seed particles, with droplet size controlled by swelling rather than

by stirring rate. This is performed in a medium designed to preclude particle

coalescence.

The protocols from the Frechet group (Wang et al. 1992; Wang et al. 1994) used a

4 to 10-fold excess (v/w) of swelling agent over seed particles. The swelling suspension

was prepared by dispersing 60 mg seeds with at least 200 Cll DBP in 30 ml 0.25 w/v %

aqueous sodium dodecyl sulfate (SDS) solution by sonicating (using a cell disruptor

probe on 50 % intensity for 30 s). The suspension of swollen seeds was stirred for at

least one whole day. The above solution had added to it 0.03 w/v % sodium nitrate as

solution phase inhibitor and approximately 5 ml water with enough PVP40 stabilizer to

adjust the total solution to 1 w/v % stabilizer concentration, and stirred for 10 min. A









monomer mixture of 4.4 ml styrene with 2.5 w/v % benzoyl peroxide (BPO) initiator was

added, and emulsified in the solution via probe sonication for 20 s. The whole mixture

was then polymerized in a sealed glass vessel at 700 C for 24 h. The product was used in

the next step without purification. Styrene and DVB were mixed 1:1 to make 10 ml

monomers, and had 1 w/v % BPO added as initiator. The monomer mixture was

emulsified in 40 ml of 0.25 w/v % aqueous SDS solution as previously, by probe

sonication for 20 s. The emulsified monomer mixture was stirred with the polymerized

swollen seeds for 5 h at room temperature. Polymerization took place in sealed glass

vessels for 24 h at 700 C. The product was washed and purified like any other

microsphere preparation.

The Tuncel protocols that were applied to prepare batches of microspheres were

drawn from several publication by Tuncel et al. (Tuncel 1999; Tuncel et al. 1999; Tuncel

and Cicek 2000; Tuncel et al. 2000; Tuncel et al. 2002a; Tuncel et al. 2002b). The

essential protocol was similar to the above Frechet protocol, with a primary difference

being that the swelling agent (Dibutylphthalate, DBP) was used at a 1:1 ratio (v/w) with

the seed particles. The typical Tuncel-based protocol used 160 mg of uncrosslinked

styrene seeds, swollen with 160 Cll DBP that was emulsified in 20 ml of aqueous 0.25 w/v

% SDS solution, and stirred together for 24 h. A monomer emulsion was prepared that

contained equal parts of styrene, hydroxyethyl methacrylate (HEMA) and DVB to make

1.2 ml monomers, mixed with 60 mg benzoyl peroxide initiator until dissolved, and

emulsified in 20 ml aqueous 0.25 w/v % SDS solution. The monomer emulsion was

stirred together with the swollen seeds emulsion for 24 h at ~300 rpm and room

temperature. Sufficient 10 w/v % stabilizer (either PVA or PVP40) was added to bring









the concentration of stabilizer in mixture to 1 w/v %. The preparation was nitrogen

purged before polymerization, then sealed in glass vials and polymerized for 24 h at 700

C at 120 cpm in a shaking water bath, with the vials arranged diagonally to the direction

of motion. The product was then washed and purified as for other polymerizations.

4.3 Immobilization of Folic Acid onto Microspheres

4.3.1 Folic Acid Immobilization Procedure

Folic acid (figure 4-10) must be bound via the carboxylic acid groups that exist on

the glutamic acid end of the molecule rather than via the amine group on the opposite end

(at the methylpteridin). This has been proven necessary in order to preserve receptor

binding functionality, as binding to the receptor occurs across the amine group. The

procedure has been adapted from protocols by Zhang et al. (2002), who have shown

success binding folic acid onto particles with surface chemistries similar to the

microspheres being used.



HNN CHNH-NHH--O
OH
II II




Figure 4-10: Folic acid structure

4.3.1.1 Materials for folic acid immobilization

Dimethyl sulfoxide (DMSO, Acros 12779 [67-68-5]), N-hydroxy succinimide

(NHS, Aldrich 13,067-2 [6066-82-6]), 1 -ethyl-3-(3-dimethylaminopropyl) carbodiimide

(EDC, Aldrich 16,142-2, [25952-53-8]), folic acid (Fisher Bioreagents BP251925, [59-

30-8]), Ethylene triamine (triethylamine / Et3N, Aldrich 13,206-3, [121-44-2]), were all

used as delivered from the respective suppliers.









4.3.1.2 Method of folic acid immobilization onto microspheres

EDC NHS
ONH2 4 Folic acid (COOH) ONH-(CO)Folic Acid
Et3N

Figure 4-11:. Schematic of folic acid immobilization onto microspheres

Folic acid immobilization proceeded along the plan shown in figure 4-11. The

chemistry utilized is depicted schematically in figure 4-12. The carbodiimide compound

EDC is used as a zero-length crosslinking agent, mediating the formation of an amide

between the carboxyl group on folic acid (at the glutamic acid end) and the terminal

amine on the microsphere (from the aminoethyl methacrylates comonomer). Ethylene

triamine was used to activate the amine group~-rendering it more reactive. The

carbodiimide modifies the carboxyl group to form a short-lived O-isoacylurea

intermediate, which is highly reactive towards nucleophilic amine groups, to form the

strong amide linkage and an isourea by-product. NHS is added to the reactants to

increase the stability of the active intermediate (Hermanson 1996).

The microspheres were washed repeatedly prior to immobilization, and 20 mg of

microspheres were suspended in 2 ml DMSO. 20 Cll Et3N was added to 'activate' the

microsphere surface amine groups, and the suspension was mixed (sonicated and

vortexed for 10 min). A second solution was made consisting of 1 ml 10 CIM folic acid in

DMSO with 1.5 ml each of aqueous 15 mM NHS and 75 mM EDC. This second solution

was mixed (sonicated and vortexed) for 10 min. Both solutions were combined, and

mixed for 20 min. The microspheres were subsequently separated from the solution

magnetically, and washed in distilled water 8x, and finally stored in 5 ml distilled water,

away from light.









Carboxylate
compound
OH
R'
O


EDC


CH3
H--N-CH3





forms O-isoacylurea
intermediate


CH3


= O


R'- C=


H-N-C=
CH3




amide bonded
product



NH


CH3\
NNCH-CH + R--NH2
amide
compound



I isourea by product


HOCH3\
N-C-N NH+ CH3
CH3 H~


Figure 4-12: Schematic detailing the carbodiimide mediated coupling of a carboxyl
group to an amine to form an amide linkage.

4.3.2 Characterization of Folic Acid Immobilized Microspheres

The folate-immobilized microspheres were characterized to demonstrate the

attachment of folate onto the surface. The model system detailed in chapter 6 was

utilized for this process also, particularly in cases where larger and more polydisperse

microspheres generated significant light scattering to create extra noise in light

spectroscopy systems.

For spectroscopic methods, there are two independent fluorescent species

contained in folic acid: p-aminobenzoic acid (PABA, the leftmost species in the folic acid










structure shown in figure 4-8), and methylpteridin (MTE, the central portion of the folic

acid structure). PABA has an absorption band at 265 nm and emits at 336 nm (Tanojo et

al. 1997). MTE has absorption bands at 275 nm and 352 nm, and emits at 447 nm

(Espinosa-Mansilla et al. 1998). The absorption spectrum of folic acid shows maxima at

281 nm and at 355 nm.

4.3.2.1 UV-Visible spectroscopy

Absorption spectra for control and immobilized microspheres were compared to

the absorption spectrum of aqueous folic acid. Experiments were performed in spectrum

mode on dilute suspensions of microspheres in DI water.

4.3.2.2 Fluorospectrometry

Experiments were performed on a dilute suspension of microspheres in DI water,

using a Fluorolog Tau-3 (Jobin Yvon Spex Instruments, S.A. Inc) with a 450 W xenon

lamp as the excitation source.

4.3.2.3 Brookhaven zeta plus

The Brookhaven Zeta Plus zeta potential and particle size analyzer provided data

on the charges of the microspheres before and after immobilization. Phosphate buffered

saline (PBS), 7.4 pH was used as medium, and particle sizes from Coulter data were

specified as parameters. Software used was BIC Zeta Potential Plus, set to 1 run of 10

cycles. Measurements were repeated at least 3 times--more if results were inconclusive

or questionable.

4-4 Cell Testing

In order to evaluate the microspheres produced for specific binding to folate

receptors, testing on cell lines was an important model system. This required cell lines

that were known to express folate receptors (FR) as well as cell lines known to not










express folate receptors in order to serve as controls. Cells were cultured according to

standard techniques while observing any instructions specific to the individual cell lines,

and grown on circular cover slips to provide a limited and controlled testing environment.

4.4.1 Cell Lines

The cell lines used were primarily sourced from the American Type Culture

Collection (ATCC), and much of the information on the cell characteristics was directly

drawn from their database (www.atcc.org). Cells were also acquired from Clonetics

(Clonetics was acquired by Bio Whittaker, and operates under Cambrex Life Sciences).

Appendix B provides full information sheets on each cell line, as provided by the source

company.

4.4.1.1 Malignant cell line CRL-5800 / NCI-H23 human epithelial lung
adenocarcinoma

The NCI-H23 cell line is a human epithelial cell, non-small cell lung

adenocarcinoma originally derived from a 51 year old black male. Lung

adenocarcinomas are known to overexpress FR (Weitman et al. 1992; Franklin et al.

1994), making it a good first model to utilize for testing.

4.4.1.2 Secondary testing cell line CCL-163 / BALB/3T3 clone A31 mouse
fibroblasts

The BALB/3T3 cell line is a mouse fibroblast that is known to be highly

susceptible to transformation in tissue culture. It was utilized as a secondary testing cell

line.

4.4.1.3 Control cell line NHDF: normal human adult fibroblasts

Normal human adult fibroblasts from Clonetics were used as controls. There is

no indication that these cells have any overexpression of folate receptors.









4.4.1.4 Oral squamous cell carcinoma cell line SCC-9

Oral squamous cell carcinoma (OSCC) cells were utilized to provide a tissue-

specific cell type for experiments. The cells were from the ATCC SCC-9 cell line, and

were revived after extended cryo-storage (>10 years).

4.4.2 Cell Culture Procedures

The cell lines were grown in RPMI 1640 lx medium without L-glutamine

(Mediatech Inc., Herndon, VA). The medium was supplemented with 10 % fetal bovine

serum purchased as manufactured for investigational use (Mediatech, Inc., Herndon,

Virginia) and 1% Penicillin-Steptomycin 10,000 IU/ml and 10,000 Clg/ml purchased as

sterile filtered for in vitro diagnostic use (Mediatech, Inc., Herndon, Virginia). Medium

was changed every 2-3 days and cells were passage weekly using 0.05% trypsin-EDTA

(Mediatech, Inc., Herndon, Virginia). All cells were maintained in 6 multi-well plates, T-

25 flasks and/or T-75 flasks with respectively, 3ml, 5ml and 13ml of growth medium

while being kept at 370C in a humidified 95% air: 5% CO2 atmosphere.

4.4.3 Cell Culture Preparation and Testing Procedure

4.4.3.1 Cells seeded onto multi-well plates

This procedure was utilized for NCI-H23 cells. Cells were grown in 6-well

culture plates according to the procedures above and in appendix B. For testing, 200 Cll

of each suspension (particles in water, ~ 7 w/v %) was pipetted into wells (added directly

to medium), followed by 6 min gentle agitation of the plate, after which the wells were

aspirated. Each well was rinsed by addition of Hanks balanced salt solution, and 60 s

agitation to dislodge and non-bound particles. The rinse procedure was repeated 3 times.

A fourth and final rinse with cell growth medium was performed. The medium was










aspirated from the cell wells, and the plate was inverted for observation on a Zeiss

Axioplan 2 imaging microscope with Axiovision 3.1.2 image acquisition software.

4.4.3.2 Cells seeded onto coverslips

Cells were seeded onto ~ 6 mm diameter coverslip circles that were hole-punched

out of Fisher 22 mm plastic coverslips (Fisher catalog # 12-547). The coverslips were

placed at the bottom of cell wells in multi-well plates, and cell seeding took place as for

wells. Fisher specified the coverslips as being cell-adherent, but specific materials data

was unavailable.

For testing, the coverslips were picked up with tweezers and immersed into wells

of suspended microspheres and swirled for 60 s. The tweezers were cleaned with 70 %

ethanol solution before each manipulation to ensure no transfer of microspheres occurred

by tweezers, and for better hygiene. Approximately the same concentration suspension

of microspheres was used, but this factor was not strictly controlled for. The rinse

procedure involved swirling the coverslips in the same rinse solutions as above for 60 s.

Rmnsing was repeated 3 x with Hanks balanced salt solution and once with medium. In

later experiments, the rinse utilized was changed to medium.

For analysis the coverslips were placed onto microscope slides for support and

examined with the Zeiss Axioplan 2 imaging microscope.

4.4.3.3 Cell experiments for microsphere specific binding

A series of experiments with cells were conducted to determine the capacity of

the folate immobilized microspheres to specifically bind to the tumorous cell lines.

Coverslips were treated as described above with either folate-immobilized microsphere or

with controls (no folate immobilization). The coverslips were then examined under the

microscope for microspheres that were specifically bound to cells. Images were acquired









on the Zeiss Axioplan 2 microscope using a 10 x Neoplan-fluar obj ective and a collection

area of 1300 x 1030 pixels. The image captured was 1.0683 Clm per pixel (horizontal and

vertical same, calibration from microscope supplier)-hus approximately 1400 x 1100

Clm giving an area of 1.53 mm2

4.4.3.4 Fluorescently labeled microspheres

Microspheres that were fluorescently labeled possessed an additional mechanism

by which they could be distinguished from cells. The capture of this information

required additional imaging steps. Experiments involving specifically bound

microspheres on cells were imaged by transmitted light microscopy, and then by

fluorescence microscopy for the same frame and magnifieation, with appropriate

adjustments to exposure settings. In some cases it was possible to create a combined, or

dual image involving low intensity transmitted light to provide feature definition, as well

as fluorescence imaging to provide spatial information on the fluorophores.

4.4.3.5 Image treatment and analysis

Images were acquired with the goal of generating data to turn into a test statistic

to evaluate the performance of microsphere specific binding. This process required a

number of steps. First, cell culture samples were treated with microspheres either

folate-immobilized or controls. Then images were taken of representative areas from

each sample. Areas were chosen that could produce the best quality image (thus the most

reliable data). Images of these areas were acquired, and then analyzed manually. The

images were divided into smaller areas using Axiovision 3.1 software, and microsphere

contact points to cells were counted. Additional data collected included average









microsphere and cell sizes and ranges. This same general procedure was followed for all

the cell lines, as well as for the tissue samples.

4.4.3.6 Microsphere recovery experiments

Experiments involving the recovery of the test microspheres were performed in

two ways. In all cases, the samples were treated as described above, with either folate-

immobilized or control microspheres. In most instances, samples were then viewed

under the microscope to determine the extent of specific binding that had occurred. In

some cases, the samples were not imaged previous to microsphere recovery.

Recovery of the microspheres was performed using several procedures to dislodge

the microspheres from their specific binding. It was known that acidic conditions caused

endosomal release within cells (Low et al. 2001), so acidic solutions were sampled to

dislodge the microspheres. In all cases, the wash procedure was performed on the sample

using the same procedure as for rinsing (described in section 4.4.3.2). Agents used as the

rinse included 1 through 5 v/v % aqueous acetic acid (2.6 to 2.2 pH). Other agents

experimented with included trypsin and mucolytic agents, as well as acidic drink

products (CokeTM). Microspheres were collected by placing a magnet under the rinse

container for 30 s and the decanting the rinse solution. The microspheres retained could

then be counted and/or imaged.

4.5 Tissue Testing

4.5.1 Institutional Review Board (IRB) Approval

IRB approval is required for all research proj ects involving human subj ects. The

definition of human subjects includes excised tissue. IRB approval was granted on the

basis of Expedited IRB #488-2003. This approval included a waiver of the requirement

of patient consent forms, since the study posed no possible hazard to the patient. All









tissues used were excised solely on the basis of clinical diagnoses for pathological

reasons. No records exist linking the patients to the tissues studied, and the study

personnel were blinded to the records. Only information concerning the acquisition date,

the pathological diagnosis (by study pathologists and Shands pathology assistants), and

the nature of the tissue (type and source location, pathological status) were recorded.

This information was recorded on a standardized form to improve records keeping and

provide compliance with IRB office policy.

4.5.2 Tissue Preparation

The tissues used for testing were acquired exactly as outlined in the application

for Expedited IRB #488-2003. Tissue was utilized fresh, or snap frozen after surgical

excision. Several types of samples were prepared according to recommendations by

pathology personnel.

4.5.2.1 Fresh tissue samples

Fresh samples were tested in several ways. Initial testing took place using

sections cut from a larger resectioned tumour. These samples were approximately 3 mm

x 5 mm x 2-4 mm thick. Later samples were cut to the same dimension at lesser

thickness of ~ 1 mm to provide a flatter surface that would allow better focus with the

optical microscope.

4.5.2.2 Snap-frozen tissue samples

Tissue that could not be used as fresh was snap-frozen in liquid nitrogen to

preserve for later testing. The frozen samples could then be sectioned to size for analysis,

or mounted on slides. Standard histological preparation methods were followed in

mounting the slide samples. The frozen sample was attached to a backing slide and

embedded in a polymeric grout. This assembly was placed into a cryomicrotome and










sectioned to desired thickness. Samples used were sectioned to 10 Clm, 20 Clm and 30 Clm

thickness. These samples were then mounted onto microscope slides. This procedure

was performed by the histological assistants and technicians in the gross pathology

laboratory at Shands Health Care Center. Standard H & E (Hematoxylin and Eosin)

stained slides of the sample were also prepared.

4.5.3 Sample Treatment for Testing with Microspheres

The samples prepared were treated with the procedures established using the cell

lines. Variations existed to accommodate the different physical shapes of the unmounted

tissue samples and the prepared slides.

4.5.3.1 Unmounted tissue testing procedure

The unmounted samples were able to be treated in the same manner as the cover

slips that the cells had been grown on, since their size permitted the physical transfer of

the samples between cell wells in multi-well plates. The samples were dipped and

swirled in a suspension of the testing microspheres in DI water, for 60 s, then dipped and

swirled in 3 rinses of PBS 7.4 pH buffer for 60 s each.

4.5.3.2 Slide mounted tissue testing procedure

The samples that were mounted on slides could not be physically transferred

between wells as above. For these samples, the same concentration of microspheres

suspension was prepared, and the slide was placed at an angle in a petri dish, as shown in

figure 4-13. A transfer pipette was used to pour suspension over the slide. The

suspension collected in the dish and was poured over the slide again for 60 s. This same

procedure was repeated with 3 successive rinses of PBS 7.4 pH.






74


4.5.4 Microscopy of Prepared Tissue Samples

The prepared tissue samples were examined in the Zeiss Axioplan 2 imaging

microscope. Pictures were taken at 10 x magnification. The images that were recorded

included pure optical images (transmitted light, no filter), transmitted light images using

the fluorescein filter cube (bandpass filter allowing only green component of light to

pass), fluorescence mode (fluorescence excitation) and mixed/dual mode (low level

transmitted light together with fluorescence excitation). Whenever possible, each of

these four types of images would be acquired for the same area at the same focus to most

clearly delineate the structures being viewed.















Fiue -3 Tsuesie ontdo sie en rneda ar f etngpocdr









4.5.5 Magnetic Recovery of Microspheres From Tissue Samples

Following the sample preparation procedure outlined in 4.5.3, microspheres were

dislodged from receptor mediated binding through the application of an acid wash. 1 v/v

% aqueous acetic acid was utilized in the manner appropriate to the sample--applied in

the same manner as the PBS rinse. A magnet was applied to the bottom of the collection

vessel (either a dish or a cell well) into which the acetic acid rinse terminated. The acetic

acid was then decanted, and the microspheres retained magnetically. The microspheres

could then be examined in situ or transferred to permit microscopy or spectrometry.















CHAPTER 5
RESULTS AND DISCUSSION

The design requirements presented in chapter 3 called for a size range of

microspheres between approximately 2 and 100 Clm, with a primary amine group

available for ligand coupling and superparamagnetic inclusions to permit separation. The

Microspheres were prepared in this size range by a suspension copolymerization of

methyl methacrylate (MMA) and aminoethyl methacrylate (AEMH) monomers, the

former doped with coated magnetite. The design requirements also called for a method

of dye labeling, for which two approaches were demonstrated: a direct covalent

fluorophore labeling, and a swelling-based dye incorporation. Characterization for size,

surface chemistry, fluorescence, magnetic recovery and ligand adhesion was performed,

demonstrating that all the design requirements were fulfilled.

The microspheres were tested using various cell lines to demonstrate receptor

specific binding. Recovery of the bound particles was demonstrated using an acid wash

to sever the receptor binding followed by magnetic recovery of the microspheres.

Fluorescence microscopy was utilized both to analyze the microspheres themselves and

to acquire spatial information on the particle-cell interactions.

5.1 Microsphere Synthesis

Suspension polymerization was ultimately chosen as the best synthesis route to

form the desired particles. This conclusion was based on work done producing

microspheres by suspension polymerization as well as by dispersion polymerization,

including activated swelling procedures that used dispersion polymerized seed particles.









5.1.1 Dispersion Polymerized Samples

Dispersion polymerized microspheres had been produced in our group on many

occasions (Leckey 1997), in some cases with many of the required characteristics for the

present application. Styrene based microspheres polymerized with diethyl aminoethyl

methacrylate were prepared in a dispersion medium containing iron oxide to yield the

product shown in the SEM micrograph of figure 5-1. Appendix A provides detailed

polymerization records on each of the polymerized samples that are discussed in the text.

Dispersion polymerized samples offered the opportunity to investigate a number

of incorporation mechanisms for iron oxides. Several of the strategies that produced

viable magnetically responsive microspheres are detailed below, along with the

microspheres generated.




















Figure 5-1: Sample D013-4; ST-co-DEA particles dispersion polymerized in ferrofluid

5.1.1.1 D013 dispersion polymerization with ferrofluid

The dispersion polymerized batch of microspheres D013 was produced from a

feedstock of 69 vol % styrene (ST), 30 vol % diethyl aminoethyl methacrylates (DEA)










and 1 vol % crosslinkerdivinyl benzene (DVB). This monomer mixture was

polymerized in 70 % methanol and 30 % water, with 3 wt % polyvinyl pyrrolidone (PVP-

40) as stabilizer. For the particular sample shown, half of the water in the dispersion

solution had been replaced with a 3.5 wt % ferrofluid. Figure 5-1 clearly shows that the

product from batch D013 was very polydisperse with irregular shape. While the shape

and size of the product was unsatisfactory, the EDS spectrum in figure 5-2 below

demonstrates a strong response for iron in the labeled peaks. Phase separation apparently

resulted, yielding a large portion of segregated, pure white polymer and a smaller, fairly

agglomerated mass of black polymer-iron oxide mixture~oth of which can be seen in

the micrograph of figure 5-1. This system did not yield an overall satisfactory

microsphere for any application, but it did demonstrate that iron oxide could be

polymerized into particles in a one-pot polymerization procedure.

ops







8 0- Al48

60- Cg Cu~

40ur -2.ESsetao apeD1- hwn togio ek

5.1.1. D030dispesion plymerzatio with ron oide Feiiae nst





Figudrec 5-2 ESit speciptrat of ro sample 013-4 sho ingsronger ioron io) peaks









oxide was formed by the same mechanism as described in section 4. 1.2--a base

catalyzed precipitation of iron oxide in magnetite from iron chlorides. The formulation

for the microspheres utilized the same monomer feedstock as the previous batch

presented: 69/30/1 ST/DEA/DVB. For the particular sample shown-DO30-3M-the

dispersion medium was 85 vol % denatured ethanol and 15 vol % distilled water, with 12

w/v % stabilizer (PVP40). The DEA monomer (structure shown in figure 5.3) provided

sites for the iron chlorides to coordinate to at its amine junctions, as per Ugelstad patent

literature (Ugelstad et al. 1988).


O /CH,CH,

H2C= C -C OCH2 CH2 N\

CH, CH,CH,

Figure 5-3: Diethyl aminoethyl methacrylate (DEA) structure

The results with this process were microspheres of 1.88 Clm volume average

diameter, with a standard deviation of 0.29 Clm, as determined by Coulter LS 230 light

scattering particle size analysis. Figure 5-4 shows the original microspheres formed in

panels a and b. A suspension of the microspheres had iron oxide precipitated onto

(and/or into) them--these samples are shown in panels c and d and clearly show a finer

surface roughness probably due to precipitated iron oxide. The magnetized DO30_4m

samples were sized at a volume average of 4.42 Clm diameter with a standard deviation of

5.90 Clm--indicating agglomeration as shown in the particle adhesions in panels c and d

of figure 5-4.







































Figure 5-4: Samples DO30_3. Panels a and b show the dispersion polymerized
microspheres as produced. Panels c and d show the microspheres after iron
oxide precipitation.

The magnetic character of the particles was determined in several ways. First,

XRD analysis shows, as illustrated in figure 5-5 (spectrum) and figure 5-6 (spectrum data

listing), that the identity of the precipitated species was not only magnetite, but included

maghemite--an oxidation product of magnetite. This behaviour was in no way

contradictory to observations that the particles reacted appropriately for

superparamagnetic species since both magnetite and maghemite are known

superparamagnetics (Barthelmy 2004). Second, as the microspheres were magnetically

separated from suspension, and they did not agglomerate when outside a magnetic field,









the magnetic character of the material could be taken as superparamagnetic on this

empirical evidence. Third, the iron content of the sample was analyzed by ICP, and

found to amount to 4.2 wt % iron, which in turn equates to 5.8 wt % iron oxide (this

calculation is based on an assumption of 100 % magnetite. This is not strictly accurate,

but calculation with even 100 % maghemite would not alter the results beyond the

magnitude of experimental error, which is about 5 % for the instrumental measurement,

and at least another 5 % for handing and processing).


















Figure 5-5: XRD spectrum from sample DO30_4m.

The conclusions from the preparation and characterization of dispersion

polymerized samples D013 and DO30 was that small particles with superparamagnetic

character could be produced by dispersion polymerization methods. These particles were

however neither as monodisperse as non-magnetic dispersion polymerized particles

shown in figure 5-7, nor as regularly shaped. Additionally, the dispersion polymerized

particles could not be formed at an average size of greater than a few Clm.























































Figure 5-6: XRD data listing from spectrum shown in figure 5-5 for sample DO30_4m.

































Figure 5-7: D009 dispersion polymerized microspheres. 69/30/1 ST/DEA/DVB
feedstock composition.

Dispersion polymerization as a process for making the particles required for the

application detailed in chapter 3 was thus set aside. This was due partly to the above

results, and partly due to the nature of the dispersion polymerization process. The

required monomers with amine functionalities are typically more water soluble than the

main monomer used: methyl methacrylate. This mismatch in water solubility would lead

to a significant loss of the more soluble reactant to solution. Compensation can be made

in the monomer feedstock ratio, but the development time entailed in devising a

dispersion system (such systems are largely empirical in nature) was known to be

significant. It was thus judged unlikely that the time would be well spent, in view of the

above difficulties that would still apply to dispersion polymerized microspheres.









5.1.2 Activated Swelling

The activated swelling process was developed by Ugelstad, who found that the

degree of swelling that a polymer particle was able to undergo was radically increased

through 'activation' by a low molecular weight solvent (Ugelstad 1978). While non-

activated particles are able to swell by only approximately 8-fold (as measured by either

mass or volume) when swollen with monomer, activated particles were able to swell by

more than 100-fold (Ugelstad et al. 1979).

The swelling process relies on a seed particle that is amenable to

swelling--meaning that it has no crosslinker in the composition. A common particle type

used for swelling was dispersion or emulsion polymerized styrene, since monodisperse

seeds permit the formation of a monodisperse swollen particle population. Seed particles

made by dispersion polymerization of styrene are shown in 5-8. Final particle

composition can be controlled through the swelling mixture, permitting a range of

chemistries that would not otherwise be amenable to dispersion or emulsion

polymerization techniques.

The particular batch of particles pictured in figure 5-8 served as seeds for

activated swelling reactions ASO2 through ASO6, and nicely monodisperse as shown by

the size histogram in figure 5-9, where the size of ASO5 is shown alongside. The volume

average particle size was measured by Coulter LS 230 as 1.25 Clm, with a standard

deviation of 0. 16 Clm. These seeds were 'activated' by forming an emulsion of dibutyl

phthalate (DBP) in aqueous 0.25 % sodium dodecyl sulfate (SDS), and adding the seeds

as per (Tuncel 1999; Tuncel et al. 1999; Tuncel and Cicek 2000; Tuncel et al. 2002a;

Tuncel et al. 2002b). Once the seeds were activated, the emulsion of monomers and