<p> Chromatin conformation and dynamics remains unsolved despite the critical role of the chromatin in fundamental genetic functions such as transcription, replication, and repair. At the molecular level, chromatin can be viewed as a linear array of nucleosomes, each consisting of 147 base pairs (bp) of double-stranded DNA (dsDNA) wrapped around a protein core and connected by 10 to 90 bp of linker dsDNA. </p><p> Using small-angle X-ray scattering (SAXS), we investigated how the conformations of model nucleosome arrays in solution are modulated by ionic condition as well as the effect of linker histone proteins. To facilitate ensemble modeling of these SAXS measurements, we developed a simulation method that treats coarse-grained DNA as a Markov chain, then explores possible DNA conformations using Metropolis Monte Carlo (MC) sampling. This algorithm extends the functionality of SASSIE, a program used to model intrinsically disordered biological molecules, adding to the previous methods for simulating protein, carbohydrates, and single-stranded DNA. Our SAXS measurements of various nucleosome arrays together with the MC generated models provide valuable solution structure information identifying specific differences from the structure of crystallized arrays.</p>

<p> Past studies have shown that tumor growth generally follows an exponential growth function or, with a limiting growth constraint, the sigmoid Gompertzian function, where a terminal tumor size is reached at late times. The classical Gompertzian description of tumor growth applies in the case of two-dimensional (2D) <i>in vitro</i> cell studies due to the effect of physical limitations on possible growth area. This project asked whether Gompertzian form applies to the <i>in vitro</i> growth of multifocal 3D tumor nodules, whose size is determined by aggregation events as well as cell proliferation. Previous reports have indicated that these three-dimensional (3D) spheroids appear to reach a terminal size, even though the full available 3D volume is not occupied. In this scenario it is not immediately obvious if individual nodules are growth-constrained by nutrient or oxygen diffusion, or rather if the ensemble of all nodules exhibits Gompertzian form. 3D <i>in vitro </i> ovarian cancer cells were chosen as the population to be studied. The ovarian cancer cells were grown in overlay on a laminin-rich extracellular matrix (ECM). This model system is a common and widely used cell culture platform in cancer cell research. Using this system, division of the ovarian cancer cells into heterogeneous clusters that aggregate into larger clusters, and then reach a steady bimodal distribution of small and large aggregates, was observed. The average volume as well as the total volume of these two cell aggregate groups were measured over time to determine the nodules&rsquo; growth behavior would plateau without a growth area limitation. Biological processes may limit the size and behavior of cells within sphere-like multicellular nodules differently than a simple layer of cells on a petri dish. The standard deviation of the rapidly growing nodule volume population within a 3D <i> in vitro</i> ovarian cancer sample was shown to grow according to a quadratic function, while the population of small nodules stays constant over time. The overall growth behavior of the total volume of the rapidly growing nodules was Gompertzian. The spread between the increasing average size of the large and growing nodule population and the constant average size of the population of small nodules increased exponentially. A particle velocity tracking program was used to search for a relationship between the lateral velocity of the nodules within the field of view and the average size of the rapidly growing nodules. The average lateral velocity of all nodules was shown to weakly decrease over time. This indicates that the behavior of 3D grown ovarian cancer cells follow a dissemination pattern in which small cells or nodules of ovarian cancer cells demonstrate higher dissemination than large nodules. The motion of smaller cell nodules or single cells may be advantageous in the <i> in vivo,</i> as well as the in vitro settings. This advantage may produce the bimodal distribution of mobile small aggregates and large slow-moving and growing aggregates, and in turn, this behavior may demonstrate that dissemination of small aggregates of ovarian cancer cells occurs in a 3D environment.</p>

<p>Photosynthesis&mdash;the conversion of sunlight to chemical energy&mdash;is fundamental for supporting life on our planet. Despite its importance, the physical principles that underpin the primary steps of photosynthesis, from photon absorption to electronic charge separation, remain to be understood in full. Electronic coherence within tightly-packed light-harvesting (LH) units or within individual reaction centers (RCs) has been recognized as an important ingredient for a complete understanding of the excitation energy transfer (EET) dynamics. However, the electronic coherence across units&mdash;RC <i> and</i> LH or LH <i>and</i> LH&mdash;has been consistently neglected as it does not play a significant role during these relatively slow transfer processes. Here, we turn our attention to the absorption process, which, as we will show, has a much shorter built-in timescale. We demonstrate that the&mdash;often overlooked&mdash;spatially extended but short-lived excitonic delocalization plays a relevant role in general photosynthetic systems. Most strikingly, we find that absorption intensity is, quite generally, redistributed from LH units to the RC, increasing the number of excitations which can effect charge separation without further transfer steps. A biomemetic nano-system is proposed which is predicted to funnel excitation to the RC-analogue, and hence is the first step towards exploiting these new design principles for efficient artificial light-harvesting. </p>

<p> Taking inspiration from nature has promoted the production of man made pores, analogous to protein channels, providing a platform in unraveling the mechanism for nanoscale transportation. This gives insight to applications such as drug delivery, cancer detection and bio-electronics. Therefore, to gain more insight, hydrogels of different structural stiffnesses were translocated through pores with 800 <i>nm</i> and 500 <i>nm</i> diameters. The stiffness of the hydrogels were classified into 3 categories of 2%, 5% and 10% crosslink percentage, 2% being the least rigid and 10% the most. The characterization of each set of particles were classified by their resistive pulse, charge intake, average translocation time and effective particle size. It was determined that the characterization of each resistive pulse is defined by the inner structure of each pore&rsquo;s imperfections. Now by focusing on each individual pore, we could see slight variations between each set of hydrogels as they translocate the unique pore. What we obtained was the translocation times for more compressible hydrogels being faster than for more rigid ones. The charge intake gave insight into the particle size, with larger charge intake into the pore referring to larger particles and smaller intake for smaller one. This made clear the notion that bigger diameter hydrogels with same crosslink percentages will translocate at a slower rate than smaller hydrogels. The other piece of data obtained was the average effective diameter for all 3 hydrogel sets at given voltages. Both 500 <i>nm</i> and 800 <i>nm</i> pores coincided in the order of effective particle size translocated, from smallest to greatest starting at 10%, 2%, and 5%, respectively. This indicates that even though the 2% crosslink is the most compressible hydrogel, it did not translocate the biggest particles. Instead the 5% crosslink hydrogels translocated the bigger sizes, meaning that the 2 extremes are not ideal for bigger particle translocation. The 10% is obviously too rigid to allow big hydrogels than the other 2 sets, but the 2% indicates that some phenomena exists that tells us a compressibility to an extreme won&rsquo;t produce bigger sized translocations. The latter phenomena is still not fully understood, but because both sized pores gave the same effect and follow the same trend, it indicates the probability of this being true pretty high and with good certainty. We&rsquo;ve now characterized these translocations and understand that by classifying the effects of our parameters, the properties of any given translocated object can be deciphered. This leads to the ultimate goal of using these techniques to develop applications for bio technology.</p>

<p> Single molecule experiments are a set of experiments designed specifically to study the properties of individual molecules. It has only been in the last three decades where single molecule experiments have been applied to the life sciences; where they have been successfully implemented in systems biology for probing the behaviors of sub-cellular mechanisms. The advent and growth of super-resolution techniques in single molecule experiments has made the fundamental behaviors of light and the associated nano-probes a necessary concern amongst life scientists wishing to advance the state of human knowledge in biology. This dissertation disseminates some of the practices learned in experimental live cell microscopy. The topic of single particle tracking is addressed here in a format that is designed for the physicist who embarks upon single molecule studies. Specifically, the focus is on the necessary procedures to generate single particle tracking analysis techniques that can be implemented to answer biological questions. These analysis techniques range from designing and testing a particle tracking algorithm to inferring model parameters once an image has been processed. The intellectual contributions of the author include the techniques in diffusion estimation, localization filtering, and trajectory associations for tracking which will all be discussed in detail in later chapters. The author of this thesis has also contributed to the software development of automated gain calibration, live cell particle simulations, and various single particle tracking packages. Future work includes further evaluation of this laboratory's single particle tracking software, entropy based approaches towards hypothesis validations, and the uncertainty quantification of gain calibration.</p><p>

Over recent decades, the structure and movement of a linear, flexible polyelectrolyte is a phenomenon of fundamental interest, both in experiment and theory. The investigation of the phenomenon guides approaches to fundamental biological sciences and polymer sciences. Computer simulations offer us the opportunity to treat macromolecules as entity and build structural models. This dissertation effectively modeled the polyelectrolyte with a coarse-grained model. We applied Langevin dynamics to the modeled system with a flexible polyelectrolyte chain and counterions. The globule-coil transition of the polyelectrolyte chain was induced by changing the solvent quality or by tensions on the ends of the polyelectrolyte. With Langevin dynamics simulations, we demonstrated that the counterion condensation was changed corresponding to changes of solvent quality or stretching forces. In accordance with the theory of polymer translocation, simulations in this dissertation uncovered the nonuniversal features of polymer translocation through a short solid state nanopore. Based on these studies, we presented the dependence of the polymer translocation events and the polymer translocation time on the applied voltage and the sequence of the polymer chain. We also modeled a system of template DNA, MspA pore and phi29 DNA polymerase with coarse-grained models and applied Langevin dynamics to the simulation. This simulation investigated and uncovered how the phi29 DNA polymerase controlled the DNA sequencing under the applied external voltage. Polymer dynamics of the template DNA during the replicative activity of phi29 DNA polymerase were also investigated.

This thesis explores the impact of surface heterogeneities on colloidal interactions and translates concepts to biointerfacial systems, for instance, microfluidic and biomedical devices. The thesis advances a model system, originally put forth by Kozlova: Tunable electrostatic surface heterogeneity is produced by adsorbing small amounts of cationic polyelectrolyte on a silica flat. The resulting positive electrostatic patches possess a density that is tuned from a saturated carpet down to average spacings on the order of a few hundred nanometers. At these length-scales, multiple adhesive elements (from tens to thousands) are present in the area of contact between a particle and a surface, a distinguishing feature of the thesis. Much of the literature addressing surface “heterogeneity” engineers surfaces with micron-scale features, almost always larger than the contact area between a particle and a second surface. With a nanoscale heterogeneity model, this thesis reports and quantitatively explains particle interaction behavior not typical of homogeneous interfaces. This includes (1) an adhesion threshold, a minimum average surface density of cationic patches needed for particle capture, (previously observed by Kozlova); (2) a crossover, from salt-destabilized to salt-stabilized interactions between heterogeneous surfaces with net-negative charge; (3) a shift of the adhesion threshold with shear, reducing adhesion; (4) a crossover from shear-enhanced to shear-hindered particle adhesion; (5) a range of surface compositions and processing parameters that sustain particle rolling; and (6) conditions where particles arrest immediately on contact. Through variations in ionic strength and particle size, the particle-surface contact area is systematically varied relative to the heterogeneity lengthscale. This provides a semi-quantitative explanation for the shifting of the adhesion threshold, in terms of the statistical probability of a particle being able to find a surface region sufficiently attractive for capture. Though neglecting hydrodynamics, the resulting (κ-1a)1/2 power law scaling for the density of patches at the adhesion threshold roughly captures the general shape of the data. The study also reveals that at high ionic strength, particle-surface interactions are most influenced by the patchy surface heterogeneity; however, at low ionic strengths, the system becomes most sensitive to the average system properties. Thus for heterogeneous interfaces, the extent to which heterogeneity is influential depends on other factors (particle size, ionic strength). While this comprises a crossover from heterogeneity-dominated to mean field behavior, it is worth noting that even in the mean field regime, the spacing between patches always exceeds the Debye length, making the regions of different surface charge always distinct. Comparison with the simulations of Duffadar and Davis reveals that the criterion for particle capture is a nearly constant number of cationic patches per unit area of contact between a particle and a heterogeneous collector. The heterogeneous surface model displays a shear crossover seen with bacteria and other complex systems: At low shear, particle capture is enhanced, while at higher shears it is reduced. This behavior, sometimes rationalized in terms of the complex energy landscapes of biological bonds, is clearly explained in the heterogeneity model. For weakly adhesive systems engaging only a few adhesive elements or receptors, shear compromises the ability of a few bonds to capture particles. For more strongly adhesive systems, shear increases particle transport. The convolution of this competition leads to the non-monotonic effect of shear seen in biology. The complex variety of particle behaviors combined with the large number of independently variable parameters, each with different scaling of interfacial forces, necessitates a state-space approach to mapping regimes interactions and motion signatures. Following the approach taken by biophysicists for describing the interactions of leukocytes with the endothelial vasculature near an injury, the state spaces in this thesis map regimes of free particle motion, immediate firm arrest, and persistent rolling against macroscopic average patch density, Debye length, particle size, and shear rate. Surprisingly, the electrostatic heterogeneity state space resembles that for selectin-mediated leukocyte motion, and reasons are put forth. This finding is important because it demonstrates how synthetic nanoscale constructs can be exploited to achieve the selective cell capture mechanism previously attributed only to specialized cell adhesion molecules. This thesis initiates studies that extend these fundamental principles, developed for a tunable and well-characterized synthetic model to biological systems. For instance, it is demonstrated that general behaviors seen with the electrostatic model are observed when fibrinogen proteins are substituted for the electrostatic patches. This shows that the nature of the attractions is immaterial to adhesion, and that the effect of added salt primarily alters the range of the electrostatic repulsion and, correspondingly, the contact area. Also, studies with Staphylococcus aureus run parallel to those employing 1 μm silica spheres, further translating the concepts. Inaugural studies with mammalian cells, in the future work section, indicate that application of the surface heterogeneity approach to cell manipulation holds much future promise.

Model membranes allow for the exploration of complex biological phenomena with simple, controllable components. In this thesis we employ model membranes to determine the effect of vesicle properties such as line tension, lateral tension, and shape on membrane-binding proteins. We find that line tension at the boundary between domains in a phase separated vesicle can accumulate model membrane-binding proteins (green fluorescent protein with a histidine tag), and that those proteins can, in turn, alter vesicle shape. These results suggest that domains in biological membranes may enhance the local concentration of membrane-bound proteins and thus alter protein function. We also explore how membrane mechanical and chemical properties alter the function of the N-BAR domain of amphiphysin, a membrane-binding protein implicated in endocytosis. We find that negatively charged lipids are necessary for N-BAR binding to membranes at detectable levels, and that, at least for some lipid species, binding may be cooperative. Measurements of N-BAR binding as a function of vesicle tension reveal that modest membrane tension of around 2 mN/m, corresponding to a strain of around 1%, strongly increases N-BAR binding. We attribute this increase in binding with tension to the insertion of N-BAR's N-terminal amphipathic helix into the membrane which increases the membrane area. We propose that N-BAR, which was previously described as being able to sense membrane curvature, may be sensing strain instead. Measurements of membrane deformation by N-BAR as a function of membrane tension reveal that tension can hinder membrane deformation. Thus, tension may favor N-BAR binding yet suppress membrane deformation/tubulation, which requires work against tension. These results suggest that membrane tension, a parameter that is often not controlled in model membranes but is tightly controlled in biological cells, may be important in regulating protein binding and assembly and, hence, protein function.

Thesis (Ph. D.)--Syracuse University, 2009. / "Publication number: AAT 3385846 ."

<p> We show that a combination of different theoretical methods is a viable approach to calculate and explain the relative binding affinities of inhibitors of the human butyrylcholinesterase enzyme. We probe structural properties of the enzyme-inhibitor complex in the presence of dialkyl phenyl phosphates and derivatives that include changes to the aromatic group and alkane-to-cholinyl substitutions that help these inhibitors mimic physiological substrates. Monte Carlo docking allowed for the identification of three regions within the active site of the enzyme where substituents of the phosphate group could be structurally stabilized. Computational clustering was used to identify distinct binding modes and their relative stabilities. Molecular dynamics suggest an essential asparagine residue not previously characterized as strongly influencing inhibitor strength which may serve as a crucial component in catalytic and inhibitory activity. This study provides a framework for suggesting future inhibitors that we expect will be effective at sub-micromolar concentrations. </p><p>