Cancer is a very serious disease; it is the second leading cause of death worldwide. Most cancers can be effectively treated if discovered early, when the tumor is in its early stage. Current clinical cancer diagnose methods lack the necessary timelines and sensitivity, which needs 20-40 years to diagnose the cancer after first parental tumor cell genesis. However, according to the present method, it still needs more than ten years for the cancer to be diagnosed. In addition, most clinical blood biomarkers lack the requisite sensitivity and specificity to detect cancer at early stages. Therefore, there is a pressing need for discovering new biomarkers that possess higher sensitivity and specificity for diagnosing and prognosing aggressive cancers. We found that extracellular vesicle (EV) –exosomes contain many types of known biomarker proteins DNA and miRNA. The specific exosomes secreted by tumor cells normally contain the biomarkers that can be used to predict the existence of cancer. The biogenesis of exosomes is regulated by the machinery of endosomal sorting complexes required for lipid ceramide or transport (ESCRT). Since exosomes possess large enough density and size that can be isolated and purified from blood using ultracentrifugation or magnetic separation methods. After re-suspending the isolated exosomes in a pure solution, the concentration of the specific biomarkers in the exosomes can be significantly increased compared to the traditional blood based biomarker method. Moreover, the volume of re-suspension solution is controllable, which can adjust the concentration of the specific biomarkers accordingly. Therefore, exosomes based method is superior over the blood based biomarker method and other traditional cancer detection methods.

However, it is still unclear how early the exosomes based detection can diagnose cancer. In order to demonstrate the feasibility of exosomes based cancer diagnosis method, we developed a mathematical model to simulate the kinetics of cancer specific exosomes in relation to the growth of tumors that begin with a single parental cancer cell. To exemplify a more realistic scenario, we primed the model on pancreatic tumor growth using Glypican-1(GPC1) circulating exosomes (crExo). The simulation parameters are obtained from published literature.

As the result, after first parental tumor cell genesis, it requires 940 days for tumor cells to secrete 105 numbers of the GPC1 specific exosomes. Followed by isolation and purification of the secreted exosomes, the isolated and purified exosomes can be re-suspended in any desirable volume of solution, which can greatly increase the concentrations of all the biological contents including the biomarkers enclosed by the exosomes compared to the traditional blood based biomarker. Hence, the required number of secreted exosomes is significantly reduced in order to reach the same equipment detection limit. In comparision, exosome based method can diagnose the cancer in 2.58 years compared to about 10 years diagosis timeframe using traditional blood based biomarker.

My developed mathematical model is capable of predicting the cancer detection time and helping the researchers simulate the growth and size for any tumor cell type. The model presented here can be extended to virtually any solid cancer and associated biomarkers.

Living cells undergoing cellular differentiation, chronological aging, and malignant progression could exhibit dramatically different biophysical characteristics. Accurate quantification of the intrinsic mechanical properties of cells is of both practical and fundamental significance in biological research and clinical diagnostics. Here, we introduce Non-invaSIve acousTo-mechanophEnotyping (ON-SITE) for high-throughput (~ 5 × 104 cells/hr) on-the-fly quantification of single-cell mechanics in continuous flow without any physical contact and labeling. ON-SITE integrates two sequential modes of acousto-dynamic operation in a staged fashion, enabling three-dimensional focusing of cells into a tight line and subsequent touchless steering of them orthogonal to laminar flow, tracing out a parabola-like trajectory. This unique dual functionality offers real-time probing and evaluation of cell mechanical properties with excellent spatiotemporal resolution by tracking cell axial transit path lengths over its entire trajectory. We identified multiple critical biophysical factors (size, density, compressibility) that contribute to the dynamic motion of cells by analytical modeling and experimental approaches. We demonstrate highly efficient, reliable, and precise quantification of the cell elastic modulus of three distinct wild-type hematopoietic progenitors and their conditionally reprogrammed counterparts upon treatment with cytoskeleton-perturbing molecules. Remarkably, the subtle microscopic differences (~ 1 – 12 %) in cell elastic modulus are unraveled on an ultrafast timescale (~ 0.3 s) using acoustic waves that are generated at a clinically safe power (~ 2 W/cm2) and frequency (~ 13 MHz) level.

Exosomes have emerged as novel biomarkers for disease diagnostics and prognosis. Exosomes are present in bodily fluids and closely resemble the contents of their parental cells; thus, they have a huge potential to serve as a liquid biopsy tool in the diagnosis of multiple diseases. In particular, tumor exosomes have the potential as biomarkers for the early detection of cancer since their contents reflect the genomic and metabolic abnormalities in their parental cells. With the development of techniques for high-throughput purification and isolation of exosomes and exosome content analysis, exosomal proteins is rapidly becoming an important tool for the early diagnosis of cancer.

Exosomes are extracellular vesicles with a diameter of 30 - 150 nm. These nanovesicles are produced by almost all types of mammalian cells and cancer cells through fusion of an intermediate endocytic compartment, namely, multivesicular bodies, with the plasma membrane.

In the introduction part of this thesis, I introduce the concept of the exosome, its definition, biogenesis and characteristics, and composition. Then I discuss the use of exosomes for cancer prognosis, and α-synuclein circulating exosomes as an example to explain how exosomal proteins work and its high specificity to Parkinson Disease. Finally, I discuss the different methods of isolating and collecting specific types of exosomes.

One unanswered question is at what stage of disease and cancer development can the exosome-based method be useful for diagnosis. To test the practicability of exosome-based methods in cancer and diseases diagnosis, I will establish a mathematical model and analytically calculate the concentrations of cancer-specific exosomes based on tumor growth. Then I will predict how early the exosome-based method can detect cancer and other diseases considering the detection limits of current diagnostic technologies. In addition, I will introduce and discuss the parameters required to inspect and verify the feasibility of my mathematical model. This thesis is focused on two specific exosomes: α-synuclein circulating exosomes for Parkinson disease and HSP70 proteins for liver cancer.

After the exosomes are extracted and purified, they can be resuspended into a small volume solution. Therefore, exosomal samples yield higher concentrations of biomarkers in a resuspension solution than those in blood. In the following, we show that it takes about 2.05 years for a parental cell to expand to a cell population that can secrete a baseline value of the α-synuclein circulating exosomes in Parkinson disease (PD) patients, and 0.73 years for a parental tumor cell to expand to a cell population that can secrete the baseline number of HSP70 proteins circulating exosomes in liver cancer patients. In conclusion, the mathematical model I established can help us predict how exosomal protein can be used to detect cancer and PD.

We are in a thrilling period defined by the rapid integration of many research disciplines and groundbreaking technical advancements fundamentally transforming our diagnostic and clinical procedures. The fusion of micro-nanotechnology and biotechnology is driving our technical capabilities to unprecedented levels. Despite notable advancements in sensitivity and multiplexing capability over the past few decades, there remains ample opportunity for innovation to enhance performance as a practical instrument for real-world applications. In order to accomplish this objective, numerous complex issues pertaining to the engineering, plasmonics, electrochemistry, and biomedical applications of bead-based assays must be addressed. To achieve this, I present a novel method in assay development, plasmonic and electrochemical sensing techniques, and automation in this dissertation. The first chapter introduces reagents used in the diagnostic assays, and current diagnostic methods such as ELISA, PCR, and bead-based assays. We then discuss three examples of widely used detection techniques. These techniques are plasmonic sensing, electrochemical sensing, and fluorescence sensing.The second chapter introduces a novel sandwich bioassay called Biomarker-to-surrogate conversion (B2S). We develop this assay to overcome the one challenge that has been neglected over the past decades: the need for solutions in the current diagnostic world when it comes to amplifying signals for protein detection. We convert target proteins (antigens) into surrogate beads that are much heavier and larger than antigens. That enables us to detect them much more reliably at attomolar-level concentrations. This remarkable noble assay enables us to have background free surrogate particles as final product of the assay. Hence, it opens us many doors to develop and execute different detection setups. The third chapter aims to develop a novel signal collection method to generate signal from surrogate products of B2S described in Chapter 2. In this chapter, we overcome the mass transport problem that has been prevented most research from going below pg/mL detection ranges, by introducing a flow-through setup for processing the final buffer obtained through our B2S assay. We introduce plasmonic nanohole arrays (NHA) and their implementation in surrogate particle sensing. We employ Extraordinary Optical Transmission (EOT) with the manufactured NHA chips, combining nanofluidics and nanoplasmonics into a single unit. We measure the peak shift in the spectrum of the transmitted light due to the resonance wavelength, and hence detect the target molecule in our sample. This setup also enables us power-free and naked-eye detection of the samples, by utilizing merely sun light, at femtomolar levels. The fourth chapter is about automating the novel assay, B2S, described in Chapter 2, with the novel electrochemical sensing technique we have developed and explained in this chapter. By using a low-cost and open-source liquid handling platform called OT-2, we are able to automate the B2S assay procedure which significantly reduces the workload on the scientist. Also, we designed a printed circuit board (PCB) to control multiple SPEs, allowing us to do electrophoretic deposition simultaneously at eight different chips. Then, by combining this PCB with a commercial potentiostat, we can extract the readout data of that automated and high-throughput diagnostic setup. The finalized setup is automated from start to finish and can process 8 samples at a time, with the possibility of increasing the number.

Abstract

A Surrogate Based Characterization Technique

For Antibody Coupling Efficiency

by

Mustafa Mutlu

The immunoprecipitation protocol is of key importance to antigen purification, to downstream protein analysis, and to clinical research. The success of an immunoprecipitation experiment depends on many variables of which antibody immobilization is the most important. Antibodies must be coupled with specific solid surfaces, such as agarose beads or superparamagnetic beads, during immunoprecipitation. Although antibody coupling is crucial for the protocol, there is not an easy-to-apply, low cost, or fast method to maximize and characterize the immuno-binding capacity. In this work we propose a characterization method which gives a clear result of the coupling efficiency, which does not require complicated experimental steps or expensive laboratory equipment. By forming a two-particle complex that sandwiches an antibody, the total amount of antibody coupled to the given number of magnetic beads can be readily translated into the number of surrogate beads that are attached to the magnetic beads through the antibodies. This method doesn’t require additional biological reagents. The surrogate bead alone suffices to quantify immobilization efficiency. This method requires only fluorescence imaging, which can be done with any portable fluorescence platform. It has multiplexing capability, provides high-throughput and high-yield, is cost-effective and time-efficient. Researchers can easily track the shelf-life, quality, degradation, and deterioration of the immobilized beads. This research has characterized a couple of variables, as well as tested automating the protocol. Our design also allows other researchers to explore more variables and automate their protocols.

We are living in an exciting era of a fast-paced confluence of wide-ranging research fields and technological breakthroughs that are profoundly altering our biomedical practices. In particular, the marriage of micro-nanotechnology and biotechnology is propelling our technical capabilities to new horizons. Despite tremendous progress made over the past 20 years, many challenges still exist for a broad range of biomedical applications spanning from few nanometers to tens of microns.

In this thesis, I aim to address three important biomedical problems at three different length scales. I first present an innovative and low-cost acousto-microfluidic approach that overcomes the diffusion limitations imposed by laminar flow profile in microfluidic channels at a challenging flow regime corresponding to extremely low Reynolds (Re~10-4) and high Péclet (Pe~106) numbers. Using this novel approach, we achieve immunoaffinity based isolation of cells using a simple rectangular channel without resorting to chaotic mixing processes or magnetic forces. Our acousto-microfluidic method uses a combination of pure and quasi-standing surface acoustic waves that are generated at a safe power (~ 1-10 W/cm2) and frequency (~10 MHz) regime comparable to those used in ultrasound imaging (e.g. fetal imaging). By exploiting the local shear forces in the vicinity of the bottom channel boundary (defined capture surface), we achieve selective separation of desired particles/cells using pre-functionalized cell specific biomolecules. We demonstrate selective enrichment of low-abundance T lymphocyte cells with a separation efficiency of about 80% at a high flow rate of 1.2 mL/hr, a separation efficiency that is comparable to most advanced rare cell isolation technologies. In the second part of my thesis (Chapter 2), I introduce a novel nanophotonic approach for enhanced inactivation of multidrug-resistant bacterial pathogens using visible light at 405 nm. To achieve this, I introduce precisely engineered aluminum nanoantenna arrays using a space mapping algorithm enabling inverse design of nanostructures from the desired spectral response. The nanoantenna arrays are optimized using radiative decay engineering principles to maximize the near-field enhancements and photothermal heating at 405 nm, a wavelength which bacteria is extremely sensitive to light and used as an emerging light-based disinfection technology alternative to ultraviolent (UV) light. In our experiments with low illumination intensities (2.5 mW/mm2), we demonstrate that our aluminum nanoantenna array (ANA) enables rapid eradication (~20 min) of multidrug-resistant Vibrio Cholerae biofilms with more than 500-fold inactivation efficiency (~99.995%) with respect to the 405 nm light alone. In the final part of my thesis (Chapter 3), I introduce a subwavelength thick (<200 nm) Optofluidic PlasmonIC (OPtIC) microlens that effortlessly achieves objective-free focusing and self-alignment of opposing optical scattering and fluidic drag forces for selective separation of exosome size bioparticles at low flow rates (< 5 m/s). Our optofluidic microlens provides a self-collimating mechanism for particle trajectories with a spatial dispersion that is inherently minimized by the optical gradient and radial fluidic drag forces working together to align the particles along the optical axis. I demonstrate that this facile platform facilitates complete separation of small size bioparticles (i.e., exosomes) from a heterogenous mixture through negative depletion and provides a robust selective separation capability for same size nanoparticles based on their differences in chemical composition. Unlike existing optical chromatography techniques that require complicated instrumentation (lasers, objectives and precise alignment stages), our OPtIC microlenses with a foot-print of 4 μm × 4 μm open up the possibility of multiplexed sorting of nanoparticles on a chip using low-cost broadband light sources (LEDs).

Vibrio Cholerae is considered an important human pathogen causing worldwide health problem and commonly exists in aquatic reservoirs. Photoinactivation of 405 nm light and photothermal disinfection by near infrared lasers have been investigated on various types of bacterial species with a notably high killing efficien. This work combines the photoinactivation of 405 nm light with plasmonic enhanced photothermal effect on the surface of an Al nanoantenna disk array. A significant cell reduction of Vibrio Cholerae was found after 20 min illumination with almost 99.99% of bacteria inactivated which suggests potential use for clinical and medical applications.

Cancer is the second leading cause of death worldwide, with as many as 13 million deaths estimated by 2030. [31]. Of these 13 million, approximately 3.9 million of these deaths would be preventable if patients are diagnosed and treated before the occurrence of cancer metastases, which are believed to cause 90% of cancer-related deaths [20]. Cancer metastases arise after the dissemination of circulating tumor cells (CTCs) into the peripheral bloodstream from primary or secondary tumor sites [28, 62]. CTCs belong to a class of cells known as ultra- rare, with an occurrence frequency of around 1 in 107 leukocytes in the peripheral blood of cancer patients [52]. In the past 10 years, the use of CTCs as a real-time liquid biopsy has received much attention, and further analysis of these cells may greatly advance our understanding of the metastatic cascade, tumor evolution, heterogeneity and resistance to therapy. However, because CTCs are ultra rare, they must be isolated from the peripheral streams in order to be effectively studied. Hence, there is great incentive for a robust cell separation technique that allows fast and efficient isolation of CTCs for downstream analysis.

Over the past decade, the introduction of surface acoustic waves (SAW) has revolution- ized the field of microfluidics, opening up a sub-field commonly known as acoustofluidics, which has become indispensable in manipulation of microparticles in a simple, compact, non-invasive and biocompatible manner. Acoustofluidic devices have prevalently been used in separating [23], patterning [66], and transporting microparticles [7], to name only a few of their many ap- plications. Many state-of-the-art acoustofluidic devices utilize active devices called interdigital

transducers (IDTs) to generate and pattern the required acoustic fields. Typically, these IDTs are arranged in matched pairs to form a standing SAW (SSAW). Recently, many groups have demonstrated that SSAWs can be used to isolate CTCs from biological samples [16, 32, 68, 3].

In this work, we present a summary on the design, fabrication, and optimization of SSAW based microfluidic devices for high throughput, label-free rare cell separation. Using existing designs in the literature as a reference, we have fabricated a novel design separation of sub-micrometer sized microparticles. This design has successfully facilitated separation of 4.0 μm diameter from those with 2.0 μm diameter. In addition to the development of fabrication and design processes, we developed both two-dimensional and three-dimensional finite-element method simulations of IDT-based SSAW generation on lithium niobate substrates. A particle tracing code was also developed in MATLAB to simulate trajectories of microparticles under the influence of specified SSAW field and drag forces due to fluid flow. We hope that the simulation techniques and the devices presented here can have a significant impact on the field of rare-cell isolation and detection.