In this work a system-level approach was taken to the single-molecule fluorescence microscopy of living cells. This primarily involved the unification of relevant information within appropriately structured artefacts that were used to inform and enhance experimentation. Initially the diversity of emerging single-molecule techniques was reviewed and presented with a novel article structure to suit the purpose of designing an experiment (Harriman and Leake 2011). Techniques were grouped by the type of information they could access, rather than the standard organisation centred on the techniques themselves. A bespoke microscope was conceived and built with reference to knowledge and tools from the fields of Architecture and Systems-Engineering. The microscope layout would enable multiple experiment types through independent control of multiple illumination beams. A technique was developed enabling the prescription of evanescent field penetration depth for each incident beam. The various empirical and theoretical results that are used to understand and modify a microscopy experiment were integrated into an internally consistent simulation model (Harriman and Leake. 2013). This was used to inform the selection of experimental components and parameters and ultimately acquire higher data quality as measured by functions such as signal-to-noise ratio (SNR). The combined experimental system of microscope and simulation model was applied in two live-cell investigations. In Escherichia coli, the spatial distribution of membrane bound proteins was investigated and a novel technique was applied to the analysis of colocalisation. Results indicate that NADH dehydrogenase and ATP synthase follow uncorrelated trajectories. This supports the hypothesis of spatial decoupling of molecules that energise the membrane and molecules that use membrane energy. In human carcinoma cells, the mechanism of ligand-receptor binding was investigated. Data was collected prior to and periodically after the addition of ligands, and fluorescence images were acquired of both ligands and receptors. Analyses based on single particle tracking are currently being carried out by a collaborator to extract information on stoichiometry and dynamics at the single-molecule level.

The self-assembly of amyloidogenic proteins to form cytotoxic species that give rise to brain deterioration underlies numerous neurodegenerative disorders such as Alzheimer’s and Parkinson’s diseases. Increasing evidence indicates that it is the rare, low-molecular-weight species (oligomers) rather than the more abundant high-molecular-weight fibrils of certain proteins that are the most cytotoxic in several neurodegenerative diseases. However, these species have proven difficult to study using traditional methods due to their transient nature and the heterogeneity of aggregation mixtures. In this thesis, I describe my work to develop advanced methods where I combine single-molecule and ensemble fluorescence techniques with microfluidic strategies to enable the study of protein aggregation, spanning small, transient oligomers to large, insoluble aggregates. In Chapter 1 I give an overview of the biological context and relevance of this work, including the background of neurodegenerative disease, amyloidogenic aggregation and key proteins involved. I then briefly review fluorescence microscopy techniques and the field of microfluidics. In Chapter 2 I describe how complex microfluidics can be integrated with single-molecule confocal techniques to provide a highly sensitive method to continuously probe protein aggregation in vitro. I show, for the first time, that the dilution of aggregating mixtures may be automated, by up to five orders of magnitude, down to the picomolar concentrations suitable for single-molecule measurements. By incorporating this microfluidic dilution device I greatly improve the temporal resolution of the technique and facilitate the observation of more transient species through the ability to rapidly dilute and take fluorescence measurements of samples. In Chapter 3 I overcome the need for in situ labels to monitor amyloidogenic aggregation using single-molecule confocal microscopy. I describe my work to adapt the single-molecule confocal technique to achieve the ultrasensitive detection of individual aggregate species under flow without covalently-attached labels. I have demonstrated the ability of this new method to monitor the aggregation of label-free amyloidogenic proteins using extrinsic labels ex-aggregation, opening the way for biological samples to be probed in a high-throughput manner. In Chapter 4 I describe my work to combine the high precision of confocal microscopy with a microfluidic device developed to directly characterise the sizes and interactions of biomolecules in the continuous phase. By monitoring the spatial and temporal mass transport on the micron scale, the diffusion coefficient, and thus hydrodynamic radius, of species may be determined. The technique delivers much greater sensitivity for size quantification, allowing scarce and other challenging samples to be characterised, and provides significant steps towards accurate sizing for single-molecule aggregation experiments under flow. In Chapter 5 I describe my work to determine the microscopic driving force for the spatial propagation of amyloid-beta. The epifluorescence instrument I built has enabled the proliferation of aggregate species to be monitored over a macro distance on a timescale of minutes. This has greatly improved the scope of the experimental data attained, which will be used in conjunction with Monte Carlo simulations to deliver a model for the propagation of amyloid-beta in vitro. Together this thesis represents my work developing the above novel fluorescence techniques to improve their temporal and size resolution, sensitivity and adaptability to study highly complex and fundamental protein aggregation linked to neurodegenerative disease.

The development of single-molecule and super-resolution fluorescence techniques has revolutionised biological imaging. Nano-scale cellular structures and heterogeneous dynamic processes are now able to be visualised with unprecedented resolution in both time and space. The achievable localisation precision and therefore the resolution is fundamentally limited by the number of photons a single-fluorophore can emit. The ideal super-resolution dye would emit a large number of photons over a short period of time. On the contrary, an optimal single-molecule tracking probe would be highly photostable and undergo no transient dark-state transitions. Single-molecule instrument development is beginning to reach technological saturation and as the frontiers of bioimaging expand, exorbitant demands are placed on the gamut of available probes that often cannot be met. Thus, the next key challenge in the field is the development of the better fluorophores that underlie these techniques; this includes both the synthesis of new chemical derivatives and alternative novel strategies to augment existing technologies. The results of this thesis are divided into two distinct parts; Project One details the development of new synthetic fluorescent probes for the study of amyloid protein aggregates implicated in neurodegenerative diseases. This includes a study of the photophysical and binding properties of a novel fluorophore library based on the amyloid dye Thioflavin-T. Following on from this, is the presentation of novel bifunctional dyes capable of simultaneously identifying hydrogen peroxide and amyloid aggregates by combining existing tools for the independent detection of these species. The sensing capabilities of these dyes are explored at the bulk and single-molecule levels. Project Two describes a new photo-modulatable fluorescent-protein fusion construct that can undergo Förster resonance energy transfer (FRET) to an organic dye molecule. This FRET cassette is comprised of a photoconvertible fluorescent protein donor, mEos3.2 and acceptor fluorophore, JF646. This strategy imparts a strong photostabilising effect on the fluorescent protein and a resistance to photobleaching. The functionality of this approach is demonstrated with in vitro single-molecule fluorescence studies and its biological applicability shown by tracking single proteins in the nuclei of live embryonic stem cells. Furthermore, initial characterisations of the excited state dynamics in effect are presented through the systematic modification of parameters.

This thesis is dedicated to the study of fluorescent conjugated polymers made of fluorene labelled with rylene moieties. Those polymers are important candidates for use in Organic Light Emitting Devices (OLEDs). The dyes present in the polymers were studied at the single-molecule level. The first part of the work is devoted to the construction and validation of an epi-fluorescent confocal/widefield/Total Internal Reflection microscope. The ensemble properties of the samples are first measured in solution. The combination of steady-state and time-resolved spectroscopies allows us to unravel the photophysics of the conjugated polyfluorene polymer containing perylenediimides in its backbone. Energy transfer is found to occur between the polyfluorene and the perylenediimide units. Beside energy transfer, a photoinduced electron transfer is also supposed to take place. Widefield microscopy is used to measure the end-to-end distance in single polymer chains. From those measurements the polymer is shown to present a quasi linear shape inside its host matrix. From the simulation of the end-to-end distance distribution, a conjugation length of 4-6 fluorene units is found. The introduction of a new subtraction method associated with defocused imaging allows us to study a more complicated polymer containing more perylenediimide units. The location and the 3D orientation of the incorporated dyes were measured at the same time by this new technique named SPIDER. Finally, the sequential two-color measurements allow us to get useful informations concerning the energy transfer occurring between polyfluorene backbone and perylenediimide units at the single molecule level.

This thesis will present the synthesis, characterization and magnetic measurements of lanthanide complexes with varying nuclearities (Ln, Ln2, Ln3 and Ln4). EuIII, GdIII, TbIII, DyIII, HoIII and YbIII have been selected as the metal centers. Eight polydentate Schiff-base ligands have been synthesized with N- and mostly O-based coordination environments which chelate 7-, 8- or 9-coordinate lanthanide ions. The molecular structures were characterized by single crystal X-ray crystallography and the magnetic properties were measured using a SQUID magnetometer. Each chapter consists of crystal structures and magnetic measurements for complexes with the same nuclearity. There are eight DyIII SMMs in this thesis which are discrete molecules that act as magnets below a certain temperature called their blocking temperature. This phenomenon results from an appreciable spin ground state (S) as well as negative uni-axial anisotropy (D), both present in lanthanide ions owing to their f electron shell, generating an effective energy barrier for the reversal of the magnetization (Ueff). The ab initio calculations are also included for the SMMs with high anisotropic energy barriers to understand the mechanisms of slow magnetic relaxation in these systems.

Photoactivated fluorescence from individual silver nanoclusters ranging in size from 2 8 atoms has been demonstrated at room temperature. The optical properties of such clusters are far superior to those of fluorescence dyes with absorption cross sections ~50 times stronger than those of even the best organic dyes. The strong oscillator strengths produced from such nanoclusters has been shown to yield comparable enhancement factors in the surface-enhanced Raman spectroscopy (SERS) process to those observed in the presence of a plasmon- supporting nanoparticle. Raman transitions are in fact so strong that antistokes scattering is also observable on a single molecule (SM) level marking the first true demonstration of SM-SERS to date. Capable of generating true scaffold specific Raman scattering on the single molecule level, the combination of fluorescence from the small nanoclusters and strong observed Raman signals in the absence of a nanoparticle strongly indicate a chemical or charge transfer SERS enhancement mechanism.

The objective of this work is to develop a micro/nanofluidic-based single molecule detection (SMD) scheme, which would allow us to inspect individual protein or protein complex study protein-protein interactions and their dynamics. This is a collaboration work with MD Anderson Cancer Center and we applied this scheme to study functions of various proteins related to cancer progression in hope to shed new light on cancer research. State-of-the-art micro/nano-fabrication technology is used to provide fused silica micro/nano-fluidic channel devices as our detection platform. Standard contact photolithography, projection photolithography and advanced electron-beam lithography are used to fabricate micro/nano-fluidic channel with width ranging from 100nm to 2?m. The dimensions of these miniaturized biochips are designed to ensure single molecule resolution during detection and shrinking the detection volume leads to increase in signal-to-noise ratio, which is very critical for SMD. To minimize surface adsorption of protein, a fused silica channel surface coating procedure is also developed and significantly improved the detection efficiency. A fluorescent-labeled protein sample solution is filled in the fluidic channel by capillary force, and proteins are electro-kinetically driven through the fluidic channel with external voltage source. Commercial functionalized Quantum Dots (Qdots) are used as fluorescent labels due to its various advantages over conventional organic dyes for single molecule multi-color detection application. A fluorescence correlation spectrometer system, equipped with a 375nm diode laser, 60x water immersion objective with N.A. of 1.2 and two avalanche photodiodes (APD) is implemented to excite single molecules as well as collect emitted fluorescence signals. A two-dimensional photon burst analysis technique (photon counts vs. burst width) is developed to analyze individual single molecule events. We are able to identify target protein or protein complex directly from cell lysate based on fluorescence photon counts, as well as study the dynamics of protein-protein interactions. More importantly, with this technique we are also able to assess interactions between three proteins, which cannot be done with current ensemble measurement techniques. In summary, the technique described in this work has the advantages of high sensitivity, short processing time (2-3 minutes), very small sample consumption and high resolution quantitative analysis. It could potentially revolutionize the area of protein interaction research and provides us with more clues for the future of cancer diagnostics and treatments.

This thesis aims to study external field induced alignment of semiconductor quantum dot by utilizing single molecule spectroscopy. Wurtzite structure semiconductors, such as CdSe, exhibit strong electric dipole moment along its c-axis. It is proposed that quantum dot can be aligned along the applied field with sufficient strength. Experiments with two kind of matrix: PMMA mixing with wax, and liquid crystal thin film, were performed for that quantum dots are able to rotate freely in the matrix. Experiments with PMMA matrix were also performed as its rigid matrix for comparison. Interdigitated structure electrodes was deposited on the cover glass for the electric field experiments. The topical transition (absorption and emission) of CdSe quantum dots has a bright plane perpendicular to its c-axis, and a dark axis along the c-axis. It thus used for characterizing the field alignment. For each observing quantum dot, we record the fluorescence intensity, anti-bunching, polarization anisotropy, and fluorescence lifetime information. In addition, we also analyze the fluorescence correlation spectroscopy to probe the small modulation signal from the fluctuating fluorescence intensity. However, the results indicate that we didn¡¦t observe the field induced change with the field up to 1E7(v/m).

I present two applications of fluorescence spectroscopy in biophysics. The first is an instrument, the anti-Brownian electrokinetic (ABEL) trap, which is capable of trapping individual small molecules in aqueous solution at room temperature. The second is an investigation of the bending mechanics of double-stranded DNA using a novel DNA structure called a "molecular vise". Both projects take advantage of the sensitivity and specificity of fluorescence spectroscopy, and both benefit from the interplay of experimental work with theoretical and computational modeling. The ABEL trap uses fluorescence microscopy to track a freely diffusing particle, and applies real-time electrokinetic feedback forces to oppose observed motion. Small molecules are difficult to trap because they diffuse quickly and because their fluorescence emission is typically weak. I describe the experimental and algorithmic approaches that enabled small-molecule fluorophores to be trapped at room temperature. I additionally derive and discuss the theory of the molecules' behavior in the trap; this mathematical work informed the design of the trapping algorithm and additionally enabled trapped molecules to be distinguished on the basis of their diffusion coefficient and electrokinetic mobility. Molecular vises are DNA hairpins that use the free energy of hybridization to exert a compressive force on a sub-persistence length segment of double-stranded DNA. In response to the applied force, this "target strand" may either remain straight or bend, depending on its flexibility and length. Experimentally, the conformation can be monitored via Förster resonance energy transfer (FRET) between appended fluorophores. The experimental results quantitatively matched the predictions of the classic wormlike chain (WLC) model of DNA elasticity at low-to-moderate salt concentrations. Higher ionic strength induced an apparent softening of the DNA which was best accounted for by a high-curvature "kinked" state. The molecular vise is exquisitely sensitive to the sequence-dependent linear and nonlinear elastic properties of dsDNA and provides a platform for studying the effects of chemical modifications and small-molecule or protein binding on these properties.

The localization, trafficking and regulation of messenger ribonucleic acids (RNA) and viral RNA play crucial roles in cellular homeostasis and disease pathogenesis. In recent years biochemical and molecular biology methods used to study RNA function have made several important advances in the areas of RNA interference, expression of transgenes, and the sequencing of transcriptomes. In contrast, current technologies for imaging RNA in live cells remain in limited use. Previous studies of RNA localization and dynamics have relied primarily on the expression of a reporter RNA and a fluorescent protein fusion that binds to aptamer sequences in the expressed RNA. While these plasmid based systems offer methodological flexibility, there remains a need to develop methods to image native, non-engineered RNA as plasmid derived RNAs may not have the same regulatory elements (3'UTR and introns) or copy number as the native RNA. Additionally, viral pathogenesis is often sensitive to the size and sequence of their genomic RNA and may not be suitable for study using engineered systems. We sought to develop and validate a new method for imaging native, non-engineered RNA with single molecule-sensitivity. These probes have four important properties. They are modular, compatible with fixation and immunostaining, bind quickly and specifically to targets, and do not interfere with RNA function. We built upon the technique of delivering exogenous, linear probes that bind to their target by Watson-Crick base pairing. The probes are multiply labeled and tetramerized to increase their brightness. To validate the probes, targeting and utility was demonstrated in two model systems: beta-actin mRNA to show targeting of an endogenous target and the genomic RNA of human respiratory syncytial virus to show targeting of a viral RNA target. All video files are in QuickTime format.