The analyses of patient tumor biopsy samples and biopsy-based diagnostics have emerged as important tools for cancer diagnostics. However, the techniques employed currently are restricted to centralized laboratories as they are time-consuming, manual labor intensive and vary considerably in their effectiveness amongst institutions and countries. Point of Care testing (POCT) for cancer with the capacity for multiplexed detection of numerous biomarkers in biopsy samples in a rapid, precise and portable manner is emerging as an area with enormous potential to disseminate universal diagnostics to cancer patients. Additionally, POCT can be used as a screening tool, to discern malignant from benign tumors at the physician's office, and lead to reduction in the need for expensive and time-consuming laboratory tests, hence minimizing the cost and anxiety, for patients with benign tumors. A POCT platform for multiple biomarker analysis can not only improve the operational characteristics of assays but can also help ascertain drug efficacy, ushering in personalized medicine for the patients. The reduced volumes and diffusion distances, which enable multiplexed, portable and quick assays, in microfluidic devices makes such devices a promising platform to realize POCT systems. But current microfluidic devices for cancer diagnostics suffer from the lack of a generalized on-chip sample preparation module and a simplified fluid actuation technique.

The overall goal of the reported dissertation research is to develop microfluidic modules that will enable the development of integrated microfluidic diagnostic platforms for the multiplexed detection of cancer biomarkers in tumor biopsy samples. The main focus of the thesis is on the development of novel microfluidic sample preparation modules. The purpose of the sample preparation component is to pre-concentrate cancerous cells, remove background proteins in the sample and to subsequently lyse the cells to release the proteins of interest. The pre-concentration of the adherent cells, including the cancerous cells, in the sample is reported by trapping them using a novel hydrodynamic cell trap. The sample washing methods, to remove extracellular proteins that could interfere with downstream assays, is also optimized. An electrochemical lysis technique is then integrated to the cell pre-concentration module, to effectively lyse the cells without having to add external reagents. Microfluidic modules for the separation of bacterial and mammalian cells from mixed samples are also reported. The immortalized cancer cell lines used in this research include the human breast cancer cell lines BT-474, known to over express the Her-2 protein, and T47D along with cervical cancer cell line HeLa.

The development of a novel fluid actuation technique, termed Proximal Degas-driven Flow (PDF), is also reported in this thesis. PDF takes advantage of the high porosity and air solubility of PDMS to reduce the pressure inside the fluidic channels leading to fluid flow in the channel. This actuation technique enables bubble-free fluid flow, can be used to fill up dead-end chambers in contrast to traditional pressure (positive or negative) driven flows and does not require the priming of the channels. Unlike degas-driven flow, PDF alleviates the need for pre-degassed and sealed devices, enabling consistent and longer-lasting fluid flow. This portable technique also requires very simple and cheap hardware like a vacuum bulb or membrane pump (thumb pump).

In conclusion, several microfluidic modules to enable Point of Care biopsy-based cancer diagnostics are introduced. The research presented in this dissertation is an effort to transform point-of-care cancer testing and provide universal diagnostics and personalized medicine to cancer patients.

Today, the development of eco-friendly and renewable energy sources is imperative because we are facing a severe global warming and energy depletion. In this regard, the microalgae that absorb carbon dioxide from the atmosphere and convert it into a high-energy carrier which can be used as biofuel are in the spotlight. Microalgae have been considered as a next-generation energy source in that it is not only extremely renewable and reducing carbon dioxide but also highly productive even with non-arable lands, seawater or wastewater. However, its applicability to next-generation energy sources still faces practical limitations, such as the higher unit cost of producing microalgae-based biofuel compared to fossil fuels. The reason is that technologies essential to efficient and economically competitive production of microalgae-based biofuel have not yet been fully developed. The representative bottlenecks include the lack of technologies as follows; 1) methods for screening the most productive microalgae strain from the hundreds of thousands of natural and genetically engineered strains rapidly and efficiently. 2) on-site, real-time, and non-invasive methods for monitoring of mass production of microalgae over the extensive area. In this respect, this thesis discusses the novel methods for superior microalgal species selection and real-time microalgal cultivation monitoring utilizing various technologies of BioMEMS, optics, and electronics.

First, we demonstrate a buoyancy-based cell separation method that can easily sort out the best microalgal strain accumulating the largest amount of lipids using BioMEMS technology. Our simple microfluidic channel structure inducing hydraulic jumps of incoming cells allows the selection of only the lightest species (i.e., superior one which accumulates the largest amount of lipid) among numerous microalgae cells having various densities depending on their lipid amount. This method has the advantage of being able to select superior species easily and rapidly without additional labeling process and observation of the entire microalgal growth process.

Second, we demonstrate an integrated microalgae analysis photobioreactor for rapid strain selection using optics. Our photobioreactor, hemispherical cavity embedded with metal nanostructures, scatters specific wavelengths favored by microalgae, which will increase the photosynthetic efficiency of microalgae, resulting in shorter screening times through rapid growth. Besides, the additional optical property (i.e., light focusing) provided by the geometry of a hemispherical cavity helps rapid screening even with naked-eye by facilitating rapid growth through cell gathering and cell-cell interactions.

Lastly, we demonstrate a novel electrical sensor that can effectively monitor microalgae in the real field. Our approach utilizes electric signal, photocurrent, derived from the channel-rhodopsin, a unique feature of microalgae. This new method overcomes the limitations in terms of the field application that other existing approaches have not solved in that it can monitor microalgae in real-time and in-situ even with low cost without the need for additional sample processing steps.

We believe this series of new approaches can elevate the productivity in real fields of microalgae-based biofuel production by contributing to the superior strain selection and production management and are expected to provide an opportunity to step closer to the realization of microalgae as a next-generation energy source.

Infectious and non-communicable diseases impose a global burden of health on both developing and developed countries despite technological advancements in medicine and healthcare. In this perspective, point-of-care testing (POCT) has been paid attention as an area with enormous potential to solve this problem. Point-of-care molecular diagnostic enables to provide diagnoses from clinical samples to clinicians without wasting time such as sample transporting or sample preprocessing. However, most of nucleic acid diagnostics still are performed in central health facilities because they need top-notch equipment and well-trained personnel. To overcome these difficulties, the emergence of microfluidics has been attracted because of its inherent features including consuming small volume of sample and reagent, precise controllability of laminar flow, short processing time for analysis, and miniaturization of devices. This dissertation is a part of effort to develop an integrated molecular diagnostic platform for improving global health care and the quality of life. Firstly, we developed a microfluidic device for sample preparation. The purpose of sample preparation is to enrich target cells, remove unnecessary cells, and lyse target cells for nucleic acid detection. The preconcentration and selective capture of cells are demonstrated in the microfluidic device by using porous materials and antibody without any external bulky equipment such as a centrifuge. Then, photonic PCR is presented as a new technology of thermal cycling. When light comes to gold (Au), Au can absorb light efficiently due to its plasmonic-assisted high optical absorption (~ 65% at 450 nm) and quickly release the light energy as a thermal energy. Using this simple principle, we demonstrate the target DNA amplification within 5 minutes on 30 cycles. In addition, a comprehensive heat transfer simulation of the photonic PCR is carried out to improve the performance of the photonic PCR. Simulation of heat transfer model demonstrates the importance of optical absorption of Au and thermal diffusivity of materials on heating and cooling ramping rates. Finally, the development of a new integrated molecular diagnostics is demonstrated for bacterial detection in urine samples.

Recent advances in micro/nano- technologies have shown high potentials in the field of quantitative biology, biomedical science, and analytical chemistry. However, micro/nano fluidics still requires multi-layered structures, complex plumbing/tubing, and external equipments for large-scale applications and nanotechnology-based sensors demand high cost. Interestingly, nature has much simpler and more effective solutions. The goal of this dissertation is to develop novel microfluidic platforms and nanobiosensors inspired by biological systems.

In this dissertation, I report the development of a biologically inspired bidirectional fluidic diode motivated by the xylem pores, which allows designing a functional large-scale microfluidic circuit and autonomous fluidic controls without any delegate efforts on fluid regulation. The biologically inspired bidirectional fluidic diode requires only a single-layered structure and a single pressure source to regulate flow in both directions through the entire platform. The operational conditions are precisely estimated based on the fully developed analytical model, which considers the hysteresis of contact angles and effects of fabrication limitations. To demonstrate its many possible applications, I show large-scale, spontaneous droplet-patterning and colonized cell-patterning programmed with the uni- and bi-directional fluidic diodes in the microfludic platform. In addition, inspired by target recognition in nature, an aptamer-based nanoplasmonic sensor, `aptasensor' is presented by detecting a coagulation protein, human α-thrombin. Also, I present an aptasensor targeting vascular endothelial growth factor-165 (VEGF165), a predominant and effective cancer biomarker for the diagnostics of various solid cancers. The integration of an aptasensor into a microfluidic platform is demonstrated by applying the VEGF165 aptasensor to detect secreted VEGF165 from breast cancer cells cultured in a microfluidic platform.

I envision that elucidation of mechanisms underlying in biological systems will inspire to create new technologies for the application to precision biology, biotechnology, and medicine. Furthermore, the developed biologically inspired engineering will provide physical insights, analytical models, and useful tools for the better understanding of biological systems.

Biophotonics denotes the utilization of light to study biology and has provided many new tools and insights for life science. For example, molecular imaging by light microscopes with fluorescent probes has gained considerable attention owing to its high sensitivity and multiplexing capabilities. Optical tweezers employing focused laser beams can individually trap micrometer-sized objects and measure biomolecular interactions. Optogenetics has demonstrated photonic control of genetically modified cells and opened up new avenues for manipulating biological events in live cells and animals. Although conventional biophotonic techniques often satisfy the requirements of cell biologists, the diffraction of light inevitably restricts their applicability to single biomolecules (e.g. proteins, DNA, and RNA) with characteristic sizes on the nanometer scale. Based on the localized surface plasmon resonance, plasmonic nanoantennae-mediated molecular sensing, manipulation and fabrication offer a promising way to overcome the diffraction limit of light by leveraging highly confined and strong electromagnetic hot-spots. Creatively utilizing interactions between hot-spots, biomolecules, cells and materials generates novel biological applications both in vitro and in vivo.

Here, I first present a method, named nanoplasmonic optoporation, for precise gene regulation by the creation of transient nanopores in cell membranes. Induced by the irradiation of near-infrared light, the enhanced surface plasmon resonance of gold particles allows us to photothermally create ~40 nm nanopores with a ~30 min lifetime. Sequential molecular delivery into somatic cells is demonstrated with single-cell precision, as well as at the larger centimeter-scale of conventional cell culture. Nanoplasmonic-mediated gene regulation in stem cells is demonstrated with higher efficiency than conventional transfection methods.

Second, graphene nanopores are created and integrated with an optical antenna with a single fabrication step, light-induced gold nanorods melting. Nanometer-sized heating source created by photothermal effect of gold nanorods resting on a graphene membrane created nanoscale pores with hemispherically shaped nanoantennas. The approach yields the significant advantage of parallel nanopore fabrication compared to the conventional sequential process using an electron beam. The atomically-thin nature of graphene, together with nanoplasmonic properties of integrated antennas, makes this unique nanopore a potential platform for high resolution and high throughput nucleic acid analysis in single-molecular level. The optical function of an integrated plasmonic nanoantenna is demonstrated by multifold fluorescent signal enhancement during the translocation of single DNA molecules through a graphene nanopore. Graphene nanopores integrated with optical antennae could offer a new avenue for simultaneous electrical and optical enhanced DNA sequencing in the future.

Third, a simple, high-sensitive, and robust method for the detection of microRNA cancer markers has been demonstrated by oligonucleotide-conjugated plasmonic nanoparticles. Due to surface plasmon resonance effects, single metallic nanoparticles provide an extremely strong light scattering signal that lowered the detection threshold and improved sensitivity for the detection of single microRNAs. In addition, dimer-type probes were designed to induce red-shift in their scattering when hybridized with target microRNAs. It is possible to achieve single-molecular detection by simply counting red-shifted nanoparticles. To further apply this method for point-of-care diagnosis, a portable total internal reflection illumination system was designed and integrated with nanoplasmonic-based sensors.

In the end of this thesis, I conclude with the outlook of nanoplasmonic integrated nanofluidics for medical diagnostics and in vivo nanoplasmonic manipulation of genetic circuits. Nanoplasmonic integrated nanopores are promising for the precise control of nano-scaled environments for DNA or RNA analysis. On the other hand, another source of excitement comes from applications of multiple plasmonic nanomaterial-controlled gene circuit engineering. Nanoplasmonic gene regulation mediated by photothermal effects might provide multiple manipulations with high spatial and temporal resolution. Studying nanoplasmonic-mediated sensing and delivery could provide new avenues for understanding how biological systems function at the molecular level.

The global population faces significant challenges in not only improving health care, but even just in maintaining it. Environmental and economic disparities are increasingly creating a massive population that lacks adequate medical attention. Traditional tools employed to detect, monitor, and treat sickness are woefully inadequate in the face of this grand challenge in medicine. However, a new generation of technologies is under development that promises to increase both the access and the quality of care available on a global scale. These tools provide simple but powerful means for both medical workers and patients to take a predictive and preventative approach to medicine – lowering the overall burden on the system and leading a new state of wellness.

There is no shortage of problems to solve in this effort to vastly energize the medical system. However, at the core, molecular diagnostic tools hold particular promise for revolutionary gains. As the core machinery of life, proteins and nucleic acids tell the full story of health and illness through their rise and fall. The ability to capture these changes, and to draw meaning from them, will be a cornerstone of modern proactive medicine. Early detection of illness often leads to a far reduced burden on the medical system, and significantly improved outcomes for the patient. In this work, I focus on the development of several technologies aimed at bringing health care a step closer to the world that needs it most.

In the first chapter, I provide a brief overview of the acute clinical needs in remote and low-resource settings, as well as an historical perspective on the periodic evolution of medicine over the centuries that has led us to the cusp of a molecular diagnostic world. I also review some of the enabling technologies that have been developed to facilitate point of care medicine.

Chapter 2 focuses on new assays and methods for nucleic acid testing in low-resource settings. I particularly focus on the development of robust isothermal amplification to glean information about both the presence of infection and the proper course of treatment. Through genotypic analysis of the invading pathogen, drug susceptibility can be determined. Furthermore, these assays are incorporated into several microfluidic and electronic systems that enable inexpensive and rapid diagnosis.

Chapter 3 provides an insight into the world of protein analysis in point of care settings. In this work, I’ve focused on robust, DNA-based sensors to detect proteins and initiate a downstream signal that can be quantified through fluorescent or nanophotonic mechanisms. In both cases, there is a focus on maintaining a rapid and robust signal that can be delivered at a low cost.

Chapter 4 deviates slightly, as it explores a means to revolutionize the traditional polymerase chain reaction by leveraging nanoplasmonic phenomena. I provide a description and initial empirical insights on a means to perform multiplexed nucleic acid quantification with little more than an LED driving both thermocycling and specific detection. While this technology certainly has applications in the developed world as well, the primary focus is on facilitating the use of gold standard assays with simple and affordable tools.

The final chapter briefly reviews my work in this doctorate degree program, while providing some insights into next steps to move these tools even closer to applications in the developing world.

Ideal point-of-care medical diagnostic devices are low cost assays capable of performing quantitative on-site rapid testing with high sensitivity and minimal manual steps.

Current mainstream assays have several key limitations. Take, for instance, the common lateral flow assay—e.g. the pregnancy dipstick test. Such assays produce rapid results at low cost; however, they are mostly qualitative tests yielding only positive/negative results rather than quantitative figures. Other standard immunosorbant assays such as ELISA yield quantitative results but require several hours and extensive manual operation. At the other end of the spectrum, nucleic acid amplification techniques such as quantitative real-time PCR can deliver much higher sensitivity and selectivity. Unfortunately, these require costly equipment and several sample preparation steps.

In this thesis, an integrated low-cost microfluidic chip and peripheral technologies for quantitative molecular diagnostics is described. These technical advances are designed to address the prevailing dilemmas described above.

Researchers have developed and integrated several key components with microfluidic lab-on-chip miniaturization technology. In line with cutting-edge technology, a novel reagent patterning method, termed “digital micro-patterning”, was developed. A very simple method, it can be adopted at low-resource laboratory settings with mainstream equipment. Digital micro-patterning is unique in the sense that it can digitally pattern and concentrate reagents into highly defined micro-patterns. As a proof of concept, it was possible to pattern isothermal amplification reagents in hundreds of microwells and run amplification reactions in these wells.

Next, a next-generation passive microfluidic pumping technology, termed the “vacuum battery system”, has been developed. This system allows for precise passive microfluidic pumping without external pumps, controls, or power sources for up to several hours. It does not require opaque fibers as in capillary systems (e.g. lateral flow assays), thus rendering this pumping method very attractive for optical detection platforms. The vacuum battery system is also significantly more robust compared to previous degas pumping techniques. Due to its portability, excellent optical properties, low cost, and the ability for complete integration with microfluidics, this platform technology opens exciting new opportunities to create a nouveau generation of standalone microfluidic chips readily operable in field settings.

Additionally, a microfluidic sample preparation technology termed “digital plasma separation” has been developed. This technology uses parallel micro-cliff-like structures and gravity sedimentation to simultaneously separate plasma and compartmentalize samples into hundreds of micro-wells within minutes. Such sample preparation method enables isothermal digital nucleic acid amplification in one step.

As a proof of concept, these technologies were integrated into a single microfluidic chip, termed the Integrated Molecular Diagnostics Chip (iMDx). This chip is capable of performing one-step quantitative nucleic acid detection directly from human whole blood samples (10~10^5 copies/ μl in 30 minutes). One low-cost disposable chip (~ $10) is designed to integrate and automate sample preparation, quantitative isothermal digital nucleic acid detection, and next generation autonomous microfluidic pumping. This portable integrated chip can acquire template concentration data similar to bench top real-time PCR machines. As the latter can cost three or more orders ($30~80k), this chip opens exciting opportunities for rapid point-of-care diagnostics in resource-low settings.

Finally, in summarizing these cutting-edge methods, a blueprint for next-generation technical development plans is laid out. The key areas of focus are downstream microfluidic integration for advanced functionality such as protein and nucleic acid multiplexed detection on a single chip, telemedicine, mass production, and clinical studies in field settings.

Ultimately, the implication of the research in this dissertation is that these platform technologies can be adopted into future medical diagnostic devices to enable rapid on-site quantitative molecular level detection at significantly lower costs. Both the system-level design rationale and component technologies developed herein provide promising building blocks for future point-of-care diagnostic assays.

Regenerative medicine is an exciting, rapidly moving field that promises to launch an entirely new industry with an entirely new product: human cells. But biology, as it is currently practiced, is not easily adapted to an industrial scale, where cost-effectiveness, throughput, and quality control are the metrics of success. For regenerative medicine to progress, new technologies must be developed to address some key questions: How can we cost-effectively expand stem cells to industrial scale quantities (on the order of 108 - 1010 per therapeutic dose) while maintaining a homogeneous pluripotent cell phenotype? How can we uniformly differentiate these cells into therapeutic phenotypes and mitigate the risk of teratoma formation? By what metric do we assess the functionality of these cells? How do we assemble these cells into functional tissues for therapy? And how can we automate these processes so that they cost-effectively produce repeatable results. What must ultimately arise is a "cell therapy pipeline" populated with a range of new technologies for stem cell expansion, differentiation, purification, and functional tissue assembly. My graduate work has focused on developing new technologies which address two major challenges currently facing the stem cell field: scalability and robustness.

In this dissertation, I introduce electrophysiology-activated cell sorting (EPACS), a label-free, non-genetic method of purifying cells based on a functional response to stimulus (namely, electrophysiology). As many of the cell types relevant for regenerative medicine are electrically-excitable (e.g. cardiomyocytes, neurons, smooth muscle cells), EPACS is well-suited for purifying cells from heterogeneous stem cell progeny in clinical applications without the risk and expense associated with labeling molecules. This concept represents an entirely new approach to cell sorting, in which a cell's functionality is assessed rather than its expression profile or physical characteristics. I present the theoretical underpinnings of EPACS in chapter 2. In chapter 3 I present a system which sorts stem cell derived clusters based on their the evoked extracellular field potential signals. In chapter 4, I present an alternative technique for EPACS which achieves single cell resolution using a non-toxic fluorescent calcium indicator dye. I demonstrate that EPACS is capable of distinguishing undifferentiated human induced pluripotent stem cell clusters (iPSC) from iPSC-derived cardiomyocyte (iPSC-CM) clusters, and that individual HL-1 cardiomyocytes can be distinguished from 3T3 fibroblasts. For these cell sorters, I have developed microfluidic devices with integrated electrodes for electrical stimulation and recording of extracellular field potential signals from cells in flow along with the associated electronic instrumentation and automation software.

In chapter 5, I present a simple method for controlling stem cell colony geometry using a silicone stencil which leads to an improvement in stem cell pluripotency maintenance and a 3-fold increase in cardiomyocyte differentiation yield. Cell colony geometry plays a significant role in cell fate determination, and factors such as colony shape, spacing, and cell density are impossible to control with conventional cell culture techniques. My technique is easily scalable and provides repeatable results across many colonies, making it well-suited for applications in high-throughput screening and industrial cell production.

Finally, I provide detailed protocols for cell culture, EPACS, stencil patterning, and microsystem fabrication in the appendices, and I briefly present two side projects that I have worked on in my graduate career: a point-of-care pathogen genotyping assay for the developing world and a microfabricated cell motility assay which enables controlled studies of 3D cell migration. In the online supplementary accompanying this dissertation, I have uploaded my MATLAB and Labview source code for cell sorting and epifluorescence image analysis.

In this dissertation, nanoplasmonic optical antennae are utilized as "nanoplasmonic gene switches" for on-demand and systematic gene regulation in living systems. The plasmon resonance of nanoplasmonic gene switches is specifically tuned to the near-infrared spectral region where cells and tissues are essentially transparent. Due to their extraordinarily large surface-to-volume ratio, nanoplasmonic gene switches are ideal carriers of interfering oligonucleotides, such as antisense DNA and short interfering RNA oligonucleotides (siRNA). Interfering oligonucleotides enable direct, sequence-specific silencing of genes, but alone, lack the temporal control necessary for precise spatiotemporal manipulation. While interfering oligonucleotides are attached to nanoplasmonic gene switches, oligonucleotide functionality is inactivated. Using light illumination as a remote trigger to release free oligonucleotides and "activate" their functionality, genes can be silenced on-demand. In addition to inhibitory effects, genes are also expressed on-demand. A transcriptional pulse of target gene expression is generated using nanoplasmonic gene switches of different aspect ratios to selectively and temporally manipulate the activities of repressors and activators upstream from the target gene. In this way, the magnitude and timing of genetic activities can be systematically varied on-demand. Equipped with new nanoplasmonic tools to directly probe the intracellular space, quantitative approaches should capture many dynamic activities within the living cell that were otherwise previously impossible to detect using conventional methods.

Current cellular assays are limited to multi-well based cell culture samples, and the existing cellular protein detection methods are restricted to labeling the targeting molecules with fluorescent dyes or other reporters. Microfluidic cell culture technology can provide precise and physiologically relevant microenvironment control to improve the quality of cell based assays. Nanoplasmonic optical probes enable the label-free detection of cellular protein with high temporal and spatial resolution. The goal of this dissertation is to develop and integrate real-time label-free nanoplasmonic detection with high-throughput microfluidic cell culture platforms.

In this dissertation, several novel nanoplasmonic geometries, such as crescent-shaped nanoholes and nanocorals are developed to provide sensitive, robust and low-cost solutions for biomolecular detection. The sharp tips and large curvature features on the nanostructures are designed to maximize the sensitivity of both localized surface plasmon resonance (LSPR) peak shift and surface enhanced Raman spectroscopy (SERS) sensing applications. Two robust cell culture platforms - long-term single cell culture arrays and self-assembled tumor spheroid arrays - that facilitate monitoring arrayed cells with precise micro-environment control are also demonstrated.

The integration of nanoplasmonic biosensors and microfluidic cell culture platforms are demonstrated on both chip-size and at the whole 4" wafer scale. Such integrated systems provide high resolution dynamic information on large array of precisely controlled cell populations. We expect this new tool will have significant impact on both fundamental cellular research and new drug evaluation.