BIOCOMPATIBLE NANOPARTICLES AND BIOPOLYELECTROLYTES
by
OLENA ZRIBI
B.Sc., Kyiv National University named after Taras Shevchenko, 1999 M.Sc., Kyiv National University named after Taras Shevchenko, 2000
M.Sc., State University of New York at Binghamton, 2001
A dissertation submitted to the Graduate Faculty of the University of Colorado at Colorado Springs
in partial fulfillment of the requirements
for the degree of Doctor of Philosophy in Applied Science - Physics Department of Physics and Energy Science
This dissertation for the Doctor of Philosophy degree by
Olena Zribi
has been approved for the
Department of Physics and Energy Science
by
___________________________ Anatoliy Glushchenko, Chair
___________________________ Robert Camley ___________________________ Zbigniew Celinski ___________________________ Karen Livesey ___________________________ Anatoliy Pinchuk ___________________________ Kathrin Spendier __________________ Date
iii
DEDICATION
To my husband Anis and my daughters Nadia and Sonia – my enormous gratitude for your everlasting patience that made this dissertation possible.
To the three generations of scientists in my family who inspired the passion for physics research in me – I am honored to continue the family tradition. To my great-grandfather professor of mathematics Dr. Ivan Shymansky, grandfather professor of physics Dr. Yuriy Shymansky, grandmother professor of physics Dr. Olena Shymanska, father Dr. Volodymyr Rudko, mother professor of physics Dr. Galyna Rudko and sister Dr. Valentyna Nosenko.
BIOFRONTIERS CENTER RESEARCH TEAM
Adviser: Dr. Anatoliy Glushchenko
v ABSTRACT
The research presented in this manuscript encompasses a merger of two research directions: a study of aqueous nanoparticle colloids and a study of biological polyelectrolytes.
The majority of biomedical applications of nanoparticles require stable aqueous colloids of nanoparticles as a starting point. A new one-step method of preparation of aqueous solutions of ultra-fine ferroelectric barium titanate nanoparticles was developed and generalized to the preparation of stable aqueous colloids of semiconductor nanoparticles. This high-energy ball milling technique is low cost, environmentally friendly, and allows for control of nanoparticle size by changing milling time. Aqueous colloids of BaTiO3 nanoparticles are stable over time, maintain ferroelectricity and can be used as second harmonic generating nanoprobes for biomedical imaging.
Biopolyelectrolytes exhibit a variety of novel liquid-crystalline phases in aqueous solutions where their electrolytic nature is a driving force behind phase formation. We study medically relevant mixtures of F-actin, DNA and oppositely charged ions (such as multivalent salts and antibiotic drugs) and map out phase diagrams and laws that govern phase transitions.
We combine these research directions in studies of the condensation behavior in aqueous solutions of biocompatible nanoparticles and biopolyelectrolytes.
AKNOWLEDGMENTS
The projects in this manuscript were supported by the following funding agencies: UCCS BioFrontiers Center; NSF grant #1102332 “Liquid Crystal Signal Processing Devices for Microwave and Millimeter Wave Operation”; NSF Nanoscience and Engineering Initiative (UIUC), NSF Grants No. DMR-0409769 and No. DMR-0409369 (UIUC), NIH Grant No. 1R21DK68431-01, DOE Division of Materials Sciences under Award No. DEFG02-91ER45439 through the Frederick Seitz Materials Research Laboratory (UIUC), STC WaterCAMPWS of the NSF under agreement No. CTS-0120978, Cystic Fibrosis Foundation.
Portions of this research were carried out at BioFrontiers Center (UCCS), the Fredrick Seitz Materials Research Laboratory (FS-MRL; UIUC, Urbana, IL), the Stanford Synchrotron Radiation Laboratory (SSRL), and the Advanced Photon Source (APS). I am indebted to my advisers Prof. Anatoliy Glushchenko and Prof. Gerard Wong for their great efforts in teaching me all I know about nanoparticle colloids and biopolyelectrolytes. I am equally grateful to Dr. Yuriy Garbovskiy for insightful scientific discussions and brainstorming innovative ideas during my research.
Also, I would like to thank all my colleagues at UCCS and UIUC for our friendship and on-going collaboration, which helped me complete the research contained in this manuscript. I especially value the climate of leadership, boundary-less innovation and university support and would like to thank the BioFrontiers team that made biophysics research possible at UCCS: Professors Robert Camley, Zbigniew Celinski, Anatoliy Glushchenko, Karen Livesey, Anatoliy Pinchuk and Kathrin Spendier.
vii
TABLE OF CONTENTS
Chapter 1: Introduction to the field of colloids and biopolyelectrolytes ……...………...1
Chapter 2: Materials and methods………28
Chapter 3: Biocompatible nanoparticle production using high energy ball milling……..67
Chapter 4: Self-organization and condensation in aqueous solutions of
biopolyelectrolytes………...………..86
Chapter 5: Nanoparticles and biopolyelectrolytes: current and future research
directions………..123
References………138
Appendix 1: Direct particle size measurements using atomic force microscopy ...…....148
LIST OF TABLES
Table 1. Solid/surfactant combinations tested………...69
Table 2. Raw materials concentration, milling time, and average particles size evaluated through dynamic light scattering………...70
ix
LIST OF FIGURES
Figure 1. Atomic origin of ferroelectricity: BaTiO3 crystal structure and polarization…...3
Figure 2. Nanoparticle colloids: from production to biological applications, specific and nonspecific nanoparticle-mediated cell imaging routes………...5
Figure 3. Actin structure………...31
Figure 4. Relative sizes of F-actin, DNA and aminoglycoside cations……….35
Figure 5. Schematic of cell washing step during RBL cell fixation process……….40
Figure 6. Factors that influence the outcome of high-energy ball milling process………41
Figure 7. Retsch PM200 planetary ball mill………..…42
Figure 8. Sawyer-Tower electric loop and applied triangular signal schematic…………44
Figure 9. Schematic of two-photon second harmonic signal generation………..45
Figure 10. Second harmonic measurement setup with Zeiss LSM 780 Meta Microscope and Chameleon Ultra II laser………46
Figure 11. Incident laser power dependence on attenuated excitation laser power generated by Chameleon Ultra II laser ....……….………47
Figure 12. Small angle X-ray diffraction experimental geometry ..………..60
Figure 13. Schematics and relative sizes of surfactants used for ferroelectric and semiconductor nanoparticle experiments...………68
Figure 14. Optical transmission spectra of the aqueous colloid of BaTiO3 nanoparticles for different milling times………..73
Figure 15. Current through the cell filled with aqueous colloid of BaTiO3 nanoparticles prepared through high energy ball milling……….76
Figure 16. Current through the cell filled with BaTiO3 nanoparticles milled in heptane..77
Figure 17. Current through the cell filled with BaTiO3 nanoparticles prepared through high energy ball milling in aqueous fluid carrier, dried, and re-suspended in heptane….78
Figure 18. Second harmonic generation of barium titanate nanoparticles and aggregates dried on the slide………....80
Figure 19. Semiconductor nanoparticle solutions produced through high-energy ball milling ………...………....82
Figure 20. ZnO quantum dots schematic: visible emission colors and corresponding particle sizes ...………...83
Figure 21. Microscopy images of salt-free mixtures of F-actin and DNA………89
Figure 22. SAXS data of salt-free DNA/actin mixtures and inter-actin nematic spacing dependence on DNA concentration ……….………….90
Figure 23. Confocal microscopy micrographs of solutions containing F-actin and MgCl2………...93
Figure 24. Phase diagram of condensation behavior in systems containing F-actin, divalent cations and monovalent anions………94
xi
Figure 25. Phase diagram of condensation behavior in systems containing F-actin, divalent cations and divalent anions………..94
Figure 26. Effect of salt anion valence on condensed F-actin bundles: monovalent anion, divalent cation…..………..95
Figure 27. Effect of salt anion valence on condensed F-actin bundles: divalent anion, divalent cation………96
Figure 28. SAXS of multicomponent Actin + DNA + monovalent salt mixtures……….98
Figure 29. SAXS of multicomponent Actin + DNA + divalent salt mixtures ..………..100
Figure 30. Schematic of condensed phases observed in actin + DNA + multivalent salt system …………...………..101
Figure 31. Confocal images of F-Actin , DNA and Spermidine mixture ..……….103
Figure 32. Confocal image of F-Actin + DNA + Tobramycin mixture ..………104
Figure 33. SAXS data for actin/Spermidine mixture for actin of various lengths…..….105
Figure 34. SAXS plot of F-Actin, DNA and Spermidine3+ mixture………106
Figure 35. X-ray diffraction and phase diagram of F-Actin, DNA and Spermidine3+…107
Figure 36. pH-dependence of DNA condensation behavior in mixtures of DNA and Tobramycin………..108
Figure 37. SAXS of binary mixture of DNA + Tobramycin…………..………...109
Figure 39. SAXS of multicomponent system containing F-Actin, DNA and
Tobramycin………..111 Figure 40. Phase diagram of multicomponent mixtures comprised of F-actin, DNA and Tobramycin………..112
Figure 41: Schematic ternary phase diagram section for the rod/polymer/solvent system
for the relevant parameters………...117
Figure 42. Cell viability diagram for RBL cells in suspension incubated for 20 hours with and without added nanoparticles……….…….124
Figure 43. Confocal microscopy micrograph of non-specific nanoparticle-cell interaction: RBL cells and BaTiO3 nanoparticles coated with Tween20………...126
Figure 44. Second harmonic generation by RBL cells incubated with barium titanate nanoparticles………128
Figure 45. Second harmonic generation by reference RBL cells incubated without the nanoparticles ..………129
Figure 46. Possible condensed phase schematic for system containing F-actin and surfactant-coated nanoparticles………131
Figure 47. Schematic of potential nanoparticles-mediated drug delivery method……..132
Figure 48. Multiphoton microscopy of semidilute F-actin solution with barium titanate nanoparticles………133
xiii
Figure 49. Confocal microscopy revealing self-assembled bundles of F-actin in solution with 80 mM CaCl2, 0.1 mg/ml BaTiO3 nanoparticles……….………134
Figure 50. AFM image and schematic of “coffee ring” formation by nanoparticle solution air-dried on mica surface ………149
Figure 51. AFM image and size distribution of barium titanate nanoparticles spin-coated on mica surface ..……….150
CHAPTER 1
Introduction to the field of colloids and biopolyelectrolytes
1.1. Outline………...…..2 1.2. Overview of prior scientific discoveries in the field of nanoparticle
colloids………....…5 1.2.1. Ferroelectric nanoparticle colloids……….………..………..5 1.2.2. Colloids of semiconductor nanoparticles……....………..……….9 1.2.3. Our research contribution to aqueous nanoparticle colloid production……...10 1.3. Aqueous solutions of biopolyelectrolytes……….12
1.3.1. Theories of polyelectrolyte condensation………...…………..15
1.3.2. Biomedical relevance of polyelectrolytes……….…23
1.3.3. Our research contribution to solutions of the biopolyelectrolyte self-organization problems………..…25
2 1.1. Outline
Microscopic interactions between the components of colloids including ions, molecules and suspended particles can be very complex to describe and often result in a wide variety of micro and macroscopic behaviors. In this manuscript, we study aqueous colloids of nanoparticles with focus on ferroelectric nanoparticles and solutions of biological polyelectrolytes.
Ferroelectricity has been discovered well over a hundred years ago but due to the poor quality of the early-discovered ferroelectric materials, it remained an academic curiosity with no practical applications until a ferroelectric phase was found for barium titanate during World War II [1,2]. This most famous ferroelectric material - barium titanate - has been first synthesized about ninety years ago [3] and has been used in an increasing number of applications ever since. In the first three decades, the phase diagram of bulk barium titanate (BaTiO3) was mapped out and the boundaries of its ferroelectric phase were outlined. The stability and easy synthesis of barium titanate lead to widespread applications of ferroelectrics in electronics [4]. With the advent of electronic miniaturization, the focus shifted from bulk ferroelectrics to thin film and nanostructured ferroelectrics in various geometries: nanowires, nanopowders and nanocolloids. The discovery of novel effects in nanoscale ferroelectrics provided exciting opportunities to implement these materials in new nanoscale device applications (reviewed in Ref. [5]). More recently, ferroelectric research expanded to new potential fields of applications including nanomaterials technologies, medicine, and biology [6].
ferroelectric phase is suppressed and the paraelectric phase dominates. The crystalline structure of barium titanate (BaTiO3)has been known for bulk materials since the 1950s [7]. In bulk form, the cubic phase of BaTiO3 is stable at temperatures in excess of 120C while the tetragonal phase (ferroelectric phase) is stable between -5C and 120C (see Fig. 1). However, at the nanoscale BaTiO3 particles can have either cubic or tetragonal crystalline structures depending on the method of synthesis.
a) b) c)
Figure 1. Atomic origin of ferroelectricity: BaTiO3 crystal structure and polarization.
Non-ferroelectric cubic structure (a), Non-ferroelectric tetragonal structure with polarization directed up (b) and down (c).
In recent years, there has been a significant effort aimed at producing ferroelectric nanoparticles with specified properties. In the last decade, a few methods were successful at producing nanoparticles that exhibit ferroelectricity [8-10]. Chapter 1.2 provides an overview of production methods and the types of nanoparticles these methods can deliver. We developed a new aqueous ferroelectric nanoparticle colloid production method based on high-energy ball milling, which is described in Chapters 2.2 and 3. Optical and electrical techniques were devised for aqueous ferroelectric colloid characterization. Second harmonic generation capabilities were demonstrated and potential imaging applications delineated for aqueous ferroelectric nanoparticle colloids.
4
The newly-developed high-energy ball milling method was generalized to generate stable aqueous semiconductor nanoparticle colloids (Chapter 3).
Another area of soft condensed matter physics that has been developing gradually over the past 100 years is biological polyelectrolytes. The most studied and understood of bio polyelectrolytes is DNA but other molecules such as actin have garnered increasing interest more recently. Many processes in biology are unraveled by studying the physics of long charged flexible biological molecules with examples including the wrapping of DNA around histones, DNA tight packing within viruses and three-dimensional networks of actin filaments in cytoskeleton [11]. Many theories describing polyelectrolyte behaviors fail to delve into multi-body effects observed in aqueous solutions of biological polyelectrolytes (such as network phase existence or finite bundle size in systems comprised of F-actin [12-15]). However, these experimentally observed effects are crucial for proper biological functioning of the polyelectrolytes in question. We review major classes of existing theories and prior experimental data that provide a foundation for our research in Chapter 1.3. In chapter 4, we build upon existing expertise in the field to probe the behavior in multicomponent biopolyelectrolyte solutions of F-actin and DNA.
We investigated the laws that govern phase behavior in medically relevant solutions containing F-actin and/or DNA and various small charged molecules including antibiotic drugs such as Tobramycin (dug of particular importance in cystic fibrosis therapy). In model cystic fibrosis mucus comprised of actin, DNA and small multivalent cations we observed competitive condensation of actin and DNA depending on multivalent cation concentration. We found that the existing condensed phases and phase boundaries could
be explained using simple criteria based on ab initio DNA charge neutralization. The obtained results formed the basis for a theory derived to describe the behavior of actin and DNA in solutions.
We further expand this research in Chapter 5 to include mixtures of polyelectrolytes (F-actin, DNA) and nanoparticles. We outline current and future research directions aimed at potential applications of aqueous nanoparticle colloids toward cell imaging (Fig. 2), tuning biopolyelectrolyte condensation behavior and potential therapeutic strategies based on nanoparticle colloids prepared by the high-energy ball milling method described in Chapter 3.
Figure 2. Nanoparticle colloids: from production to biological applications, specific (1-2-3) and nonspecific (1- 2*) nanoparticle-mediated cell imaging routes. We describe step 1 of this process in Chapters 2-3; Chapter 3 focuses on 1-2* route; Chapter 5 outlines possible 1-2-3 targeted routes.
1.2. Overview of prior scientific discoveries in the field of nanoparticle colloids 1.2.1. Ferroelectric nanoparticle colloids
Ferroelectric nanoparticles have a wide range of applications ranging from the ceramic industry and electronics [6], liquid crystal display manufacturing and photonics [6,16,17], to bio-medicine and medical engineering [6,18-22]. Possibilities of practical applications rely heavily on the existence of controllable production methods of stable ferroelectric nanoparticle colloids [6,16].
Antigen NP Nanoparticle Ligand NP Cell NP Receptor Antibody Cell NP or
6
The conventional methods used to produce such stable colloids usually include at least two basic steps: the production of ferroelectric nanoparticles and the dispersion of these nanoparticles in a fluid carrier using surfactants [23,24]. The existing methods of ferroelectric nanoparticles production can be classified as physical, chemical or biological [6] and in most cases, a combination of at least two methods is required to produce particles of good quality (narrow size distribution, tetragonal phase).
Chemical synthesis methods of ferroelectric nanoparticles include, for example, solid-state reactions, sol-gel techniques, solvothermal methods, hydrothermal methods and molten salt methods [6]. Dry and wet mechanical grinding techniques are the most cost effective mechanical techniques used to prepare nanoparticles. Wet grinding allows better control of the ferroelectric nanoparticle size distributions and during the last few years, substantial progress was made in this field [17,25]. Biological methods have also been successfully used for preparation of BaTiO3 nanoparticles: the first group of methods is based on peptides (this group of methods was reviewed in detail by Chen [26]); another group of methods relies on nanoparticle synthesis by living organisms such as fungus or bacteria [27,28]. However, many biological methods produce nanoparticles with cubic structure (i.e. paraelectric), and in a number of cases, the claims of ferroelectricity have not been directly experimentally confirmed.
The aforementioned two-step process of colloids production starts with a nano-powder (produced by one or combination of the methods described above) that is subsequently dispersed in a carrier fluid. Frequently, these two steps are not only done at separate times, but often by separate groups of people. This staging of the production process steps does not allow for much particle property tunability and flexibility.
Nanoparticle production is typically performed by the nanoparticle vendor and subsequent dispersion in aqueous media using surfactant is performed by researchers who pick dispersant and dispersion method based on the intended application. In addition, sizes of the dispersed nanoparticles produced using such methods are quite large (~40-100 nm) and are not tunable during the dispersing step [18,20-23,29]. Such techniques are usually particle- and dispersant- specific and cannot be implemented broadly.
Low energy aqueous ball milling has been used sporadically as an auxiliary method for dispersing the particles and getting their size smaller, e.g. surfactant-free and surfactant-assisted mixing of graphite nanoparticles [30], industrial titania nanoparticles [31], metal alloys such as Co70.4Fe4.6Si15B10 [32], metal oxides such as Fe3O4 [33] and BaFe12O19 ferrite nanoparticles [34]. In most cases when ball milling or ball mixing in aqueous media has been used in the past, it was not a primary method of nanoparticle production, but rather a supplementary technique, such as part of mechano-chemical methods [24,30,35].
In contrast, we will focus in Chapter 3 on high-energy ball milling, a technique that falls into the wet grinding methods category. High-energy ball milling combines nanoparticle production and dispersion in a single step and enables tuning of nanoparticle sizes by changing the milling time. Furthermore, the proposed ball milling method can be universally applied to a wide variety of nanoparticles, dispersants and fluid carriers. Only within the last year, simultaneously with our research, a few studies emerged focused solely on production of nanoparticle colloids through high energy ball milling in aqueous media (such as aluminum, multi-wall carbon nanotubes, niobium monoxide and lithium
8
niobate [36-40]). These studies emphasize the importance of surfactants in stabilization and dispersion of aqueous colloids.
The wet grinding method usually requires three key ingredients: a powder of the ferroelectric material (e.g. BaTiO3), a surfactant and a fluid carrier. The surfactant is required to prevent aggregation and overheating during grinding, and to stabilize the suspension via electrostatic and/or steric mechanisms [6,23]. The fluid carrier, nanoparticle size and dispersant selection are driven by the target application. Industrial applications (such as electronics and display manufacturing) are based on non-aqueous colloids whereas the majority of biological and medical applications require aqueous colloids.
A considerable body of research (driven by intended industrial applications) has been developed on the preparation of non-aqueous stable colloids of ferroelectric nanoparticles during the last few decades [23,25,41]. However, these methods utilize organic fluid carriers (toluene, hexane, heptane, isopropanol, acetone, benzyl alcohol, benzaldehyde etc.) that tend to be toxic and therefore biologically incompatible, hazardous and generally more expensive than simple aqueous solutions. The use of aqueous media instead of organic fluid carriers enables numerous biomedical applications of ferroelectric colloids such as biological imaging probes [18,42] and intracellular nanovectors [20-22,29]. Economic and environmental considerations are driving a general trend of replacing non-aqueous colloids with aqueous ones [23].
Our aim is to create stable colloids for biomedical applications. In most cases, “biomedical applications” implies that the items could be suspended or dissolved in
aqueous solutions. There are few exceptions to this rule, when the target biological sites are actually hydrophobic. For example, hydrophobic BaTiO3 colloids were recently used as the matrix for phospholipids and as extracting probes for hydrophobic proteins [43]. However, such hydrophobic colloids in non-aqueous carriers have very narrow and specific range of applications and they cannot be used in living systems.
Several recent studies [18,20,21,29,42,44-46] suggest the use of ferroelectric nanoparticles (e.g. aqueous barium titanate nanoparticles) produced by the previously discussed two-step method for imaging and sensing applications. This group of studies was performed using barium titanate nanoparticles with overall sizes ~100-400 nm. Yust et al., for example, focused on studying fairly large (~200nm) surface treated BaTiO3 nanoparticles of various shapes, their properties and interaction with endothelial cells and potential applications as biomarkers or biosensors [44]. Unmodified nanoparticles have been shown [44] to bind non-specifically to endothelial cell membranes with no cytotoxic effects. These findings as well as research by Hsieh et al. [18,42] indicate that barium titanate nanoparticles are compatible with the type of cells under investigation, thus making barium titanate a promising candidate for further cell-related research. Semenov et al. suggested using ferroelectric nanoparticles for contrast enhancement of microwave tomography [47], while Hsieh et al. and Ciofani et al. relied on second harmonic generation by ferroelectric materials to facilitate nanoparticle-mediated imaging [18,21,29,42,45]. Multiphoton imaging of cells and tissues in vitro and in vivo has been gaining widespread use in recent years[48]. For example, probe-free multiphoton imaging of ex vivo cancer biopsy of tissues rich in connective fibers performed by Wang et al. suggests imminent use of multiphoton microscopy for rapid ex vivo tissue diagnosis or
10
implementing multiphoton endoscopes for in vivo cancer diagnosis and surgery [49]. It has been noted that nanoparticles can enhance second and third harmonic generation, which has been used to improve the performance of microscopes [50]. The indication is that producing tunable aqueous colloids of ferroelectric nanoparticles, such as barium titanate proposed in this manuscript, holds great promise for biological and medical applications, especially in imaging.
1.2.2. Colloids of semiconductor nanoparticles
Semiconductor nanoparticles have been the subject of numerous research studies during the past three decades due to the discovery of practical applications for the quantum confinement effect. More specifically, quantum dots have been first discovered in pure semiconductor nanoparticles and tremendous effort went into improving their quality (e.g. increasing brightness, stability and biocompatibility). As a result new research directions emerged in the areas of doped semiconductor materials and core-shell structures. The method we propose in this manuscript does not allow for core-shell or surface-doped nanoparticle production, thus we will focus on materials that are known to yield quantum dots with just decreasing the particle sizes into single-digit nanometer scale. Such materials include, but are not limited to, zinc oxide, zinc sulfide, cadmium selenide, cadmium telluride, etc. Quantum dots made of these types of materials have been used for many years in a variety of imaging applications. However, semiconductor nanoparticle uses are not limited to just imaging. For example, in biomedicine ZnO nanoparticle-based systems are known for their antibacterial activity and have been used as prophylactic agents against bacterial infections ([26,51] and references therein). ZnO UV-blocking properties are used to protect the skin in cosmetics applications, and there
are recent reports of preferred killing of cancer cells and activated human T-cells using ZnO nanoparticles [51].
Certain methods previously introduced for the production of barium titanate particles also work for semiconductor nanoparticle production, e.g. peptides were also used for preparation of ZnS and CdS nanoparticles [26]. Quantum dots can be produced by variety of physical and chemical methods, such as molecular beam epitaxy, ion implantation, e-beam lithography, and X-ray lithography, wet-chemical and vapor-phase methods [52]. Just as in the case of ferroelectric nanoparticle powder production techniques, these methods usually involve an additional surface-functionalization step to promote solubility and reduce cytotoxicity. We provide an alternative one-step mechanical method that follows into the ferroelectric nanoparticle production footsteps.
1.2.3. Our research contribution to aqueous nanoparticle colloid production
We developed a one-step high-energy ball milling method of preparation of aqueous solutions of ultra-fine ferroelectric barium titanate nanoparticles (detailed in Chapters 2 and 3). This technique is low cost, environmentally friendly, and allows for colloidal nanoparticle size control by changing milling time. Investigation of the optical and ferroelectric physical properties of these aqueous colloids revealed remarkable differences from the properties of colloids in non-aqueous fluids. We developed a technique that allows us to conclude that barium titanate nanoparticles maintain ferroelectricity in aqueous colloid form.
In addition, the high-energy ball milling method was applied to generate of stable aqueous colloids of barium titanate coated with variety of surfactants and the designed technique was generalized to the preparation of stable aqueous colloids of semiconductor
12
nanoparticles. The stability of aqueous colloids of BaTiO3 and semiconductor nanoparticles was examined over time. Second harmonic generation by barium titanate nanoparticles milled in aqueous media was demonstrated.
We concluded that BaTiO3 nanoparticles are good candidates for second harmonic generating nanoprobes for biomedical imaging. This potential biological application was explored in Chapter 5, where we performed the experiments on mammalian cells incubated with aqueous colloids of nanoparticles produced through high energy ball milling. Additionally, surfactant-coated aqueous colloids provide interesting research possibilities in the field of biopolyelectrolyte studies.
In the last decade, there has been increasing interest in the integration of nanoparticles into systems made of biopolyelectrolytes and/or lipids, such as quantum dots + DNA and quantum dots + Actin + lipids systems [53-55]. Many of the structures formed by the nanoparticles self-assembled with polyelectrolytes are reminiscent of the complex multicomponent condensates of anionic biopolyelectrolytes with cations that have been rigorously studied for the past several decades [56-64]. In the past 20 years, multiple new self-assembled soft condensed matter structures were discovered, many of them in the systems comprised of highly charged biological macromolecules, such as DNA, lipids, actin, viruses [56-64]. The basics of the polyelectrolyte behavior in solution necessary to understand such complex systems are described in Chapter 1.3.
1.3. Aqueous solutions of biopolyelectrolytes
A review of polyelectrolytes research within the last 50 years reveals a large body of theoretical and experimental work motivated by numerous potential applications in industry, biology and medicine. Industrial applications, for example, range from paint,
paper making, oil recovery and water treatment to cosmetics, food, protein purification, antibacterial reagents and medical diagnosis, implant coatings, and controlled drug release. Polyelectrolytes could be both synthetic and naturally occurring. Examples of synthetic polyelectrolytes include polystyrene sulfonates, polyacrylic acid, polyvinylpyrrolidone and carboxymethyl cellulose [65]. More familiar polyelectrolytes are biological in nature – and these include nucleic acids (DNA, RNA) and proteins, such as F-actin, microtubules, etc. Polyelectrolytes have big practical value because they are soluble in water, in comparison to most neutral hydrocarbon polymers, that are only soluble in organic solvents. Tunable properties of synthetic polyelectrolytes, such as hydrophobicity, charge density and chain stiffness [65] may help in understanding biological systems.
The widespread use of polyelectrolytes is often motivated by their attractive rheological, chemical and electrostatic properties. Polyelectrolytes are polymers whose monomers bear an electrolyte group. Upon interactions with polar solvents (such as water) they dissociate and produce a charged polyion (all charged monomers of which are of the same sign) and multiple small ions of opposite charge, known as counterions. Strictly speaking, protein molecules are polyampholytes, i.e. polyion chain carries charges of both signs. However, charges of one sign often dominate and thus many proteins, such as F-actin, are often considered as polyelectrolytes.
Biological polyelectrolytes can be found in all sorts of biological systems such as viruses, bacteria, cells of animal, plant or human origin. They carry out diverse roles, e.g. DNA-carries genetic information, F-actin forms cell skeleton and is involved in cell motility, endocytosis, etc. Despite the multitude of functions, the behavior of biological
14
polyelectrolyte solutions is in large determined by their electrostatic properties and thus is generic. Biopolyelectrolytes can self-organize in the presence of oppositely charged ions and this self-organization is finding new applications in many areas, such as medicine, biology, nanoscience and water purification.
Studying aqueous solutions of biopolyelectrolytes is essential to be able to predict their interactions with each other and with various agents (such as drugs or pathogens), and to be able to alter these interactions. Biopolyelectrolytes stand apart from other polyelectrolytes and polymers because they are very sensitive to ambient conditions and are meant to carry out specific biological functions. Unlike neutral polymers, biopolyelectrolytes can self-organize in the presence of oppositely-charged ions.
Condensation in aqueous biological media is associated with self-assembly of ions in aqueous biological media (like charged and/or oppositely charged ions). In air or vacuum, like charges repel each other while opposite charges attract each other. However, in aqueous biological media like charged macroions can attract each other in the presence of oppositely charged ions that act as a “glue”. There are a number of poorly understood phenomena exhibited by aqueous biopolyelectrolytes, such as ionic valence and charge threshold for this attraction to occur and spatial distribution of macroions and “glue” ions upon condensation, as well as the effect of macro ion structure on condensation.
Because of polyelectrolytes’ dual nature as polymers and electrolytes they exhibit properties of both types of moieties. Their polymeric nature is responsible for their commonly observed liquid crystalline phases induced by high molecular aspect ratio, viscous properties and their low mobility in solutions, while the presence of electrolyte
groups results in their conductive behavior and is believed to be the reason behind most aggregate formations [65,66] in the presence of chemical entities of high opposite charge (e.g., DNA-lipid complexes). These attributes of polyelectrolytes make for a complex study subject due to a combination of several phenomena including long-range coulomb interactions, electrostatic screening, counterions condensation, steric effects, etc. These phenomena often induce structural and morphological changes, such as self-organization, in systems containing polyelectrolytes.
There are many types of interactions that can cause self-organization of polymers in solutions. Examples include surface tension, steric effects, entropy driven processes and electrostatically driven processes [67]. The most notable driving force for self organization in polyelectrolyte systems are electrostatic interactions [68], but these are not necessarily the only possible causes of self-organization in polyelectrolytes. For example, in a system of elongated molecules entropy maximization considerations can lead to liquid crystalline phase formation [69].
The behavior of weakly charged polyelectrolytes in solvents is governed by a combination of long-range Coulomb and short-range Van der Waals interactions. On the other hand, the behavior of strongly charged polyelectrolytes is mostly governed by Coulomb interactions. This manuscript will focus on studying the role of electrostatic interactions/forces on self-organization and phase formation in solutions comprised of highly charged biopolyelectrolytes and salts.
In biology, one often encounters systems with strong electrostatic interactions (such as high surface charge densities, and/or multivalent ions). Electrostatics in complex
16
biological fluids often produces counterintuitive effects [68,70-73]. For example the coexistence of polyelectrolytes in solution with multivalent counterions can lead polyions to overcome mutual repulsion and to attract each other if the counterions valence and concentrations are above certain threshold value [74,75]. This could lead to system collapse and precipitation of polyions, in other words self-organization or more specifically polyelectrolyte condensation (also referred to as aggregation).
1.3.1 Theories of polyelectrolyte condensation
Basics of mean field Poisson-Boltzmann and Debye-Hückel approaches
Poisson-Boltzmann theory and its linearized version known as Debye-Hückel theory are the usual starting point for considering any electrolyte system in mean-field approximation [68]. This theory introduces a parameter known as Debye length, which describes the rate of electrostatic interactions decay in solution (as opposed to vacuum). In water, Debye length is on the order of 7Å. Based on the value of this parameter relative to the persistence length of charged polymer, it may or may not be possible to approximate a polyelectrolyte as a cylindrical rod for purposes of theoretical modeling electrostatic interaction. Since 7Å is much smaller than the persistence length of the biopolyelectrolytes used in our work, this narrows down the field of applicable theories to the interactions of rod-like polyelectrolytes with ions and each other. Mean field approach always leads to net polyelectrolyte repulsion [76].
Polyelectrolyte interaction with counterions. Manning condensation.
Many theoretical approaches originate from the concepts first introduced in a theoretical description of condensation by Manning [77,78]. The original Manning theory
provides a criterion for condensation of monovalent counterions onto polyion but it does not provide description of multivalent counterion condensation onto polyelectrolyte, overcharging and subsequent polyion attraction and aggregation (also known as condensation). The limitations of the theory become clear in light of a large body of experimental work that sheds light on like-charge attraction in solutions of F-actin filaments, filamentous viruses, DNA [74,75,79-86].
Condensation in polyelectrolyte systems is the result of a powerful attraction between either polyion and counterions or between multiple polyions (these terms were coined in the early works by Manning and Oosawa [77,78,87-89]). The complexity of these phenomena is such that no general theories, to our knowledge, exist that describe systems with multiple types of like-charged macroions. The focus of this introductory section will be on systems with single types of macro-ions. Theoretical descriptions of polyelectrolyte solutions have several levels of complexity: one class of theories is concerned with the interactions between polyions and counterions while the other class of theories is concerned with interaction between the multiple polyions mediated by counterions.
Early mean field theories, such as those based on Poisson-Boltzmann formalism (see review by Levin [68]), fail to describe the physics of polyelectrolyte solutions adequately (Oosawa [87-89], Manning [77,78], DLVO[90]). Pioneering work of Oosawa and Manning demonstrated that the Debye-Hückel formalism (linearized version of mean- field Poisson-Boltzmann theory), designed to describe ion-counterion interaction for simple ionic solutions, would not work well for polyelectrolyte solutions. The mean field theories fail due to the fact that they cannot describe attractive interactions between polyelectrolytes observed experimentally.
18
In principle, Debye-Hückel theory could be applied to polyelectrolyte solutions as well, but the polymer concentrations for which it is valid are extremely low. Manning created one of the first theoretical descriptions of interaction between polyion and monovalent counterions [77,78], incorporating important inherent properties of polyelectrolytes, such as charge distribution along the polymer (i.e. charge density). He introduced a dimensionless parameter to quantify counterion condensation onto polyion – the Manning parameter =lB/b , where b is unit charge separation along the polyelectrolyte and lB is Bjerrum length, defined as the length at which energy of interaction of two unit charges in dielectric media equals kT: lB e2
4
0kT
. Manning concluded that counterions would condense onto the polyion until the value of is reduced to one. Manning’s theory, however, does not describe polyion-polyion attraction origin in solutions of multivalent salts. It was experimentally observed (see next section below) that such like-charge attraction could occur in a variety of polyelectrolyte systems. In an attempt to explain these findings, several new classes of theories were developed proving that different forms of correlations between counterions can produce new interactions [91-95].Salt-free polyelectrolyte mixtures
There are a few cases where mean-filed theories can be sufficient for description of polyelectrolyte behavior, such as multicomponent polyelectrolyte system without addition of multivalent cations described in Chapter 4.1. Observed phase separation in polyelectrolyte mixtures can be induced by a number of different effects. One of them can be induced by electrostatic instability - strong positional correlations between the oppositely charged ions. Another effect is instability caused by surface charge density
fluctuations on the polyion. Yet another possibility is instability caused by combination of abovementioned electrostatic effects and molecular crowding. In large part of our work we are mainly considering electrostatics-driven phase separation, but other mechanisms characteristic of neutral polymeric systems also apply. In chapter 4.1 we use Flory-Huggins type of theory to describe phase separation in actin/DNA mixtures [96]. Mean-field description for polymers in solvents, known as Flory-Huggins model, is often a basis for explanation of the majority of the effects in more complex polymer systems, such as phase separation in polymer blends [97].
Mutual polyelectrolyte attractions
Experimental basis for mutual attractions between polyelectrolyte molecules
Contrary to intuitive expectations and conclusions of mean field theories, the presence of multivalent counterions can induce an attraction between like-charged polyelectrolytes. This has been experimentally observed for many different polyelectrolytes, including DNA [81,98], F-actin [74,80,99,100], microtubules [100,101], and multiple filamentous viruses (fd, M13 and TMV) [100,102-104].
The most well-studied condensed phase of polyelectrolytes is bundled phase, where polyions arrange parallel to each other and in the plane perpendicular to polyelectrolyte axis they are nearly hexagonally close-packed [85,100,101,105-107]. This phase is observed for all abovementioned filamentous biopolyelectrolytes. In case of more flexible polyelectrolyte DNA, in addition to this phase there is also a toroidal phase [98,108-110], where DNA chain (or chains) is wound in hexagonally packed (in cross-section) toroids.
20
Experiments conducted using a variety of divalent salts helped us better understand the phase diagram of dense solutions of F-actin filaments with different lengths [63]. A number of interesting new polyelectrolyte-specific effects have also been found. For example, in the presence of divalent ions, F-actin gradually condenses into close-packed bundles via an intermediate state comprised of liquid crystalline F-actin networks [63,111,112].
Numerous experiments proved that ions of different minimum valences are generally required to condense different polyelectrolytes: F-actin [15,74,100] and many filamentous viruses [113-115] require counterion valence of 2 or higher. Most of the DNA molecules require valence of 3 or higher in general [77,81,82,108,109,116-125]. In contrast to the above behavior for trivalent and divalent ions, monovalent ions do not condense any of these polyelectrolytes [126]. The dependence of DNA and F-actin condensation on the ion valence, size and structure has been systematically studied experimentally [15,116], and recent experiments conducted by Butler et al on filamentous viruses and 'dumbbell' divalent ions provide us with an experimentally motivated criterion that explains this polyelectrolyte-specific valence threshold [75].
Classes of theories for the origin of attraction between polyelectrolyte molecules
Several classes of new theoretical concepts were introduced to explain the experimentally observed attractions between cylindrical polyelectrolytes [83,84,86,102,106,127-159]. They can be divided into two major classes of theories both of which rely on correlations between counterions to induce attraction. First class of theoretical mechanisms was pioneered by works of Kirkwood [160,161] and Oosawa
[87,88] and is based on correlated dynamic long-wavelength density fluctuations (fluctuating density of condensed counterion cloud). Multiple developments and upgrades of this theory emerged in subsequent years [83,89,102,133,134,136,162,163].
Oosawa’ work showed that uniform counterion distribution always leads to repulsion.
On the other hand, correlated fluctuations in density of counterion cloud result in large attractive force. Only polyvalent ions generate attraction. Oosawa considers counterion fluctuations and shows that the resulting force balance between two polyelectrolytes could go from repulsion to attraction [88]. He theoretically approximated this force polyion-polyion interaction (valid at large inter-rod separations R) and proved that long-wavelength charge-density fluctuations can produce attractions when R<lB. This counterion fluctuation-induced mechanism is only one of many possible ways to generate attraction in the system.
Strong positional correlations between counterions can also induce polyelectrolyte attraction and consequently condensed phase formation or phase separation. In the last decade, this second mechanism based on static positional correlations of counterions (along polyelectrolyte axis) was introduced [86,106,132,146]. The most well-known example from this class was work done by Grønbech-Jensen et al. [132]. They showed that theoretically the crossover from repulsion to attraction could be caused by onset of positional correlations between the counterions condensed on the two rods. They introduced another fundamental length scale into the problem – the ion-rod separation r0 at minimum of ion-rod interaction potential. In contrast to Oosawa’s theory, which did not account for this parameter at all, Grønbech -Jensen et al. show that their result is very sensitive to r0 [132]. To visualize the positional correlations between the condensed
22
counterions, they suggest that at low temperatures counterions form a Wigner lattice (concept originally introduced for ionic crystals) exhibiting long-range positional order, and at higher temperatures the remains of melted Wigner crystal still retain short-range positional order. These short-range positional correlations existing at small inter-rod separations R give rise to attraction at finite temperatures.
As it was noted above, stiff rod-like polyelectrolytes can form bundles of parallelly aligned rods and multitude of other phases where conformation of polymer chain stays extended. On the other hand, flexible polyelectrolytes can collapse on poor solvents and self condense. The most common example in this class is DNA toroid formation [81,164]. Flexible polyelectrolyte collapse was also extensively explored theoretically [112,137,162,165-170].
In subsequent years, there were many attempts to elaborate on both fluctuations-mediated attractions and positional correlation-induced attractions. For example, many-body effects were included to refine Oosawa’s formalism [83,134], the formation of
Wigner crystal of counterions condensed in the bundle of polyelectrolytes was examined [106,132,151]. A number of excellent general reviews have also recently been published [68,71,72,171,172].
Experiments that probe counterion behavior in condensed phases
Numerous studies focused on like-charge polyelectrolyte attraction, although until recently, none have directly observed counterion correlations. Angelini et al. carried out the first direct measurements of counterion correlations in bundled phase of F-actin – it indicates that counterion organization in the form of a charge density wave occurs within
a tightly packed F-actin bundle and that formation of a bundle is coupled to overtwisting of individual filaments [99]. Further experiments, such as neutron spin-echo spectroscopy experiments [173] and inelastic X-ray scattering [80] were aimed at discovering coupled nature between counterion and polyelectrolyte dynamics. SAXS and anomalous SAXS [174] on DNA solutions were recently performed to visualize counterion condensation.
In this manuscript we do not focus on counterion dynamics. However, since counterion dynamics induces increase in polyelectrolyte helical twist, it potentially has an effect on bundle formation mechanisms that we examine in Chapter 4.2. Further study of this connection is not within the scope of this work but could be interesting topic for future research.
In Chapter 4, we study model anionic polyelectrolyte systems containing F-actin and DNA to improve our understanding of their interactions with a variety of cationic biological/chemical agents such as pathogens and antibiotics. One of the potential applications of the conclusions derived from interactions in such model system is a problem of antibiotic treatment of lung infections associated with cystic fibrosis.
1.3.2. Biomedical relevance of polyelectrolytes Underlying biomedical problem: Cystic Fibrosis
Cystic fibrosis (CF) is one of the most common fatal genetic diseases in the United States (in USA, about 30,000 people are affected by CF) [175-188]. Life expectancy for CF patients is short (~38 years) and more than 90% of these patients die of lung disease [177,178,180,189-191] and respiratory failure [176,178]. CF is caused by mutation of a
24
single gene [183-185], which codes for the protein called cystic fibrosis transmembrane conductance regulator (CFTR)[186] that controls ion transport through cell membranes. CF patients’ lungs are infected with bacteria, most common of which is Pseudomonas aeruginosa. Pseudomonas aeruginosa form colonies (biofilms) in CF patients' lungs making the infection almost impossible to eradicate [188,192,193]. The inflammatory response results in abnormally high concentrations of actin and DNA in the airways. CF patients’ mucus (and respiratory sputum in particular) becomes unusually thick and viscous [194], clogs patients' lungs, causing respiratory distress, persistent infections, and ultimately death. Viscoelastic properties of CF mucus are in part determined by unusually high content of DNA, F-actin and salts (these are three major components of CF mucus) [195-197].
Polycationic small molecule drugs are commonly used for treatment of lung infections associated with CF. Antibiotics of the aminoglycoside family (e.g. Tobramycin, Gentamicin, Amikacin) are the most effective in battling Pseudomonas aeruginosa bacteria. One of the most widely used aminoglycosides for treatment of infections associated with CF, in clinical studies, is Tobramycin [192,198-202]. Aminoglycosides are positively charged and can get sequestered and inactivated by polyanionic (F-actin, DNA) mucus components [179,188,191,192,195,196,199,203-205]. Aminoglycoside drugs are ototoxic (can cause hearing loss) and nephrotoxic (toxic to kidneys), thus administering high doses is ill-advised. At the same time, these drugs get easily sequestered in mucus of CF patients and not all the drugs that are administered are involved in antimicrobial activity [206].
It has been previously shown that cationic antibiotics could bind to anionic macromolecules (such as DNA and actin) present in abundance in CF mucus [188,192,203,207]. Binding of antibiotic by mucus prevents it from reaching the target bacteria (in case of Tobramycin antibiotic, only 1/20 of the overall Tobramycin concentration is not bound to mucus and is free to perform its function against Pseudomonas aeruginosa [206]) and thus antibiotics activity is diminished by CF mucus [188,192,193]. The implications of aminoglycoside binding by mucus extend even further than its inactivation. At sufficiently high concentrations of multivalent cations, such as aminoglycosides and lysozyme, actin and DNA (which comprise majority of anionic sputum components) can form dense condensate complexes. Presence of such condensates significantly increases mucus viscosity [195-197,208,209], thus making it's removal more difficult.Common strategies for decreasing mucus viscoelastic modulus include DNase and gelsolin treatments [195,197,209-211]. These agents cut long polyelectrolyte components of mucus (DNA and F-actin, respectively), preventing polyelectrolyte condensate formation. Another new technique for decreasing mucus viscosity is based upon the introduction of other anionic polymers into mucus, thus shifting the balance between anionic and cationic mucus components into the phase region where bundling does not occur [196].
1.3.3. Our research contribution to solutions of the biopolyelectrolyte self- organization problems
We examined the behavior of binary polyelectrolyte mixtures on the model solutions comprised of F-actin rods and DNA coils. We studid phase separation in such model system and structural properties of the resulting phases in Chapter 4. As a starting point
26
in Chapter 4.1, we considered a simplified system composed only of F-actin and DNA with no added salt. We discovered the rules that govern phase domain formation in such system: DNA compresses F-actin into elongated ribbon-like nematic domains, and inter-actin spacing is dependent as a power –½ law on DNA concentration. This rule was proven to be independent of F-actin and DNA molecule length experimentally, and described with the aid of simple theoretical model.
In Chapter 4.2 we explored the phase formation in solutions of F-actin and divalent salt cations depending on salt anions valence. We mapped out the phase diagrams in such systems and studied internal structure of condensed phases using X-ray diffraction. We confirmed salt ion charge density wave formation in condensed bundled phase of F-actin.
Given that our binary anionic polyelctrolyte mixture consists of individual components with different contour length, charge density, and flexibility, the addition of cations of different valences was explored. Ion valence was varied from 1 to 4, such that it can effectively condense either none, one or both of the polyelectrolytes in isolation (divalent and higher valences are required to condense F-actin, and trivalents and higher are required to condense the DNA). The order in the binary polyelectrolyte mixture can in principle have a wide range of structural possibilities, from ordered or disordered composite states to complete phase separation.
We probed the binary polyelectrolyte mixtures structure using synchrotron small angle X-ray scattering (SAXS) and laser-scanning confocal microscopy. As expected, monovalent ions do not condense either component, and divalent ions selectively condense F-actin rods into bundles out of the DNA / F-actin polyelectrolyte mixture. The
situation is remarkably different with addition of multivalent cations capable of condensing both polyelectrolyte species in isolation (Spermidine, Tobramycin). We chose Tobramycin for its common antibiotic use in treatment of lung infections associated with cystic fibrosis [192,199-201,203,212,213].
We found that the existing condensed phases and phase boundaries could be explained using simple criteria based on ab initio DNA charge neutralization. We consider a simple phenomenological Flory-Huggins theory for a mixture of polymers, rods, and the solvent to explain observed condensation behavior in solutions with multivalent ions, and propose a model based on polyelectrolyte competition for ions. Observed trends describing condensation behavior in multicomponent biopolyelectrolyte systems helped us understand interactions between different components in CF mucus and layed foundation for future steps toward making drug delivery through the CF mucus more efficient.
In Chapter 5, we performed initial investigations of the compatibility of aueous nanoparticle colloids produced in Chapter 3 with mammalian cells and biopolyelectrolytes studied in Chapter 4. We layed the basis for several future research directions including aqueous nanoparticle and biopolyelectrolyte self-organization studies and development of therapeutic strategies based on aqueous colloids described in Chapter 3 and biopolyelectrolyte condensation trends explained in Chapter 4.
CHAPTER 2
Materials and methods
2.1. Materials………29
2.1.1. Nanoparticle production: powders, fluid carriers and surfactants……….29
2.1.2. Motivation in choosing F-actin and DNA as model polyelectrolytes, and in selecting particular salts and drugs………..30
2.1.3. F–actin: properties, preparation……….33
2.1.4. DNA: properties, preparation………34
2.1.5. Salt solutions and drug solutions………...35
2.1.6. Rat basophilic leukemia cell culture……….…….36
2.2. Methods……….……..40
2.2.1. High energy ball milling ...………..…40 2.2.2. Dynamic Light scattering……….…42 2.2.3. UV-Vis Spectroscopy……….….43 2.2.4. Sawyer-Tower method……….……43 2.2.5. Two-photon confocal microscopy and second harmonic generation ...……..44 2.2.6. Polarizing microscopy……….49 2.2.7. Confocal fluorescent microscopy ……….……..49 2.2.8. Small angle X-ray scattering……….……..58 2.2.9. Cell viability test ……….………65
2.1. Materials
2.1.1. Nanoparticle production: powders, fluid carriers and surfactants
Barium Titanate powder with average particle size 3 m, 99% pure (BaTiO3) and Cadmium Telluride (CdTe) powder with average particle size less than 250 m, 99.99% pure were obtained from Sigma-Aldrich. Zinc Sulfide (ZnS) crystals with sizes less than 6 mm, 99.995% pure and Zinc Oxide (ZnO) powder, 99.99% pure were purchased from Alfa Aesar.
Double deionized water (average resistivity of 18.2 MΩ) was used as the carrier fluid,
to which a surfactant was added to prevent agglomeration of the ferroelectric nanoparticles and achieve suspensions with small particle sizes through milling. Double de-ionized water (adjustable pH ~68) and media (PBS, isotonic solutions, cell media) was used to replace the carrier liquid in colloidal solutions obtained by the high-energy ball milling method, in certain described below cases.
As dispersants we chose the following materials: aminomethyl phosphonic acid (AMPA, Acros Organics, >95% pure), polyacrylic acid (PAA, MW=2000, Acros Organics, 63 wt % solution in water), 3-aminopropyl triethoxysilane (APES, Acros Organics, 99% pure) and polysorbate 20 (also known as Tween20) (Alfa Aesar). N-(6-aminohexyl) aminopropyl trimethoxysilane (N6APMS), 95% pure was purchased from Gelest, Inc. These dispersants are commonly used for stabilization of colloids in aqueous and physiological media [23,24,214-218]. At the same time, such surfactant coating is frequently aimed at non-biological applications, e.g. silane-coupled barium titanate particles were used in variety of solvents (water, ethanol, xylene [219,220]), just as
30
phosphonic acid conjugate barium titanate nanoparticles were used for nanocomposites production [221,222]. This prior body of research makes such surfactants good candidates to test the efficiency of the experimental method we propose in this manuscript. The choice of surfactants was dictated by several factors: their ability to prevent the leakage of Ba ions from the surface for the nanoparticles made from barium titanate, good physical/chemical adhesion with particles surface and reaction capability for bio-conjugation. Several other dispersants were attempted but did not produce stable aqueous colloids, namely glycine, oleic acid and asolectin lipids.
The weight ratio of the dispersant to BaTiO3 was varied from 10:1 to 1:10. Typical concentrations of nanoparticle solids in successful colloidal samples were on the order of 1mg/ml.
2.1.2. Motivation in choosing F-actin and DNA as model polyelectrolytes, and in selecting particular salts and drugs
From a physics point of view, F-actin and DNA are highly negatively charged polyelectrolytes that could be used as models for charged rod [223] and coil [78], respectively.
Actin in monomer form (G-actin) is a globular protein with average molecular weight of 43,000. When G-actin is polymerized it forms long stiff filaments, otherwise known as F-actin (Fig. 3). The molecular structure of F-actin filaments was determined by X-ray fiber diffraction experiments, first presented by Holmes at al. [224]. They concluded that G-actin units pack into 13/6 helical structure to form F-actin helical rods [224]. The average length of F-actin filaments can be controlled and the charge density is known.
Persistence length for F-actin (average of 10m) is comparable to the filament length and thus F-actin can be described as a rod. Rabbit skeletal muscle tissue F-actin is a polyampholyte with 54 positive and 43 negative charges per monomer [225], but it is traditionally modeled as polyelectrolyte with a net negative charge of 11 per monomer at circumneutral pH.
(a) (b) (c)
Figure 3. Actin structure. (a) Actin monomer (G-actin) in ribbon representation with 4 distinct domains labeled by roman numerals (image from classic biology textbook by Lodish [11]; (b) sphere model of G-actin as suggested by Al-Khayat [226]; (c) actin polymer (F-actin) in 4-sphere formalism – in isolation F-actin possesses 13/6 helical symmetry. (Schematics (b) and (c) courtesy T.E. Angelini).
-DNA, on the other hand, has persistence length (~50 nm) that is much smaller than
molecular length (16 m). This means that it would assume coil conformations in solution. Nevertheless, Debye length (length over which electrostatic interactions are screened) in water is approximately 8Å. Consequently both actin and DNA can be viewed as rods from electrostatics point of view. Both F-actin and DNA are overall negatively charged, DNA being one of the most highly charged polyelectrolytes known in nature.
Monovalent salts are known to screen repulsion in binary polyelectrolyte + salt systems, divalent salts have been shown to induce condensation in actin + salt system,
32
and DNA requires salt valence to be +3 or more in most cases to induce condensation F-actin [15,74,77,81,82,100,108,109,116-125]. Compared to the large volume of research on binary polyelectrolyte and salt mixtures, there are only few studies on multicomponent systems (multiple types of polyelectrolytes + salt) that provide closer approximation of certain biological environments [158,195,196,208,227]. We are aiming to close the existing knowledge gap and to derive rules that govern behavior in multicomponent polyelectrolyte systems.
Polyamines are polyvalent cationic molecules ubiquitous in most eukaryotes and prokaryotes. The simplest examples in this class are Putrescine+2, Spermidine+3 and Spermine+4. Polyamines are essential for cell growth and proliferation, and their raised levels in cancer cells lead to development of certain cancer therapies based on interference with polyamine metabolism [228]. Condensation of DNA that occurs in the presence of polyamines with charges +3 and +4 has been intensely studied during the last fifty years since this phenomenon is believed to be the driving force behind compaction of DNA in various phages and plays important role in nucleosome formation. Condensation and aggregate formation in the (Spermidine, DNA) system has been previously noted by many other research groups [109,117,164,229-232] and the structure of such aggregates was studied [119,231]. In contrast to such multitude of studies, multicomponent systems involving Spermidine, DNA and other polyelectrolytes received much less attention. Such systems have tremendous practical applications in medicine and biology, which motivate our interests.
2.1.3. F–actin: properties, preparation
Lyophilized powder of rabbit skeletal muscle actin (G-actin, molecular weight = 43,000) was purchased from Cytoskeleton (Denver). G-actin was suspended at 2 mg/ml in non-polymerizing solution containing a 5 mM Tris buffer at pH 8.0, with 0.2 mM CaCl2, 0.5 mM ATP, 0.2 mM DTT, and 0.01 NaN3. Monovalent salt (KCl, final concentration 100 mM) was added to polymerize G-actin forming F-actin. F-actin (filamentous-actin) is a polyampholyte (diameter Dactin=7.5 nm, length lactin=0.3–10 m, persistence length actin= 10 m, at pH ~7-8, 43 positive charges and 54 negative charges per monomer, monomer spacing ~27.5 Å in bare actin or ~28.7 Å in bundled actin). Bare F-actin has overall linear density Actin~e-/0.25 nm. Samples were allowed to polymerize for 1 hour at room temperature. The average filament length lactin=0.3–10 m was controlled by adding human plasma gelsolin (MW 87 000), an actin severing and binding protein, purchased from Cytoskeleton (Denver). F-actin average length dependence on gelsolin concentration has been previously calibrated in Ref. [15]. To prevent F-actin de-polymerization, actin was stabilized by adding phalloidin (molecular weight = 789.2; purchased from Sigma-Aldrich) at 1:1 molar ratio of phalloidin:G-actin. F-actin solutions were ultracentrifuged at 100,000 g for 1 hour to pellet the filaments. The supernatant buffer solution was then removed and F-actin was resuspended in Millipore H2O (18.2 M) to the desired concentration (usually ~10-20 mg/ml before mixing with salts and/or DNA).
For the experiments in Chapter 4, DNA and F-actin solutions were mixed at different ratios, with the global actin concentration kept constant at 5 mg/ml.
