Search

As early diagnosis is essential for improving patient outcomes when treating both communicable and noncommunicable diseases, it is necessary to develop diagnostic tests that are accessible at the point of care. Thus, there is a need for rapid, robust, and inexpensive diagnostic tests, especially in resource-limited areas where access to healthcare is reduced. One such test is the lateral-flow immunoassay (LFA), which has seen a drastic rise in use during the COVID-19 pandemic. However, the conventional LFA suffers from a couple limitations, which this thesis aims to address. First, we address the limited sensitivity of the LFA by incorporating nanozyme signal enhancement to improve detection of the SARS-CoV-2 nucleocapsid protein in serum. By developing dehydrated signal enhancement reagents and an innovative 3D printed casing to store and deliver reagents at the push of a button, we were able to improve the limit of detection of the LFA by at least 10-fold without introducing additional complexity for the user.Next, we shifted our focus to introducing semi-quantitative results to the LFA through the multicolor etching of gold nanorods (GNRs). We demonstrated the first integration of the LFA with GNRs to produce a multicolor readout for the quantification of digoxin levels in serum. We then expanded upon this work by developing a fully paper-based assay to incorporate both the LFA detection and GNR etching steps together, which involved the design of novel color-changing GNR pads. The color hues generated by GNR etching are easily distinguishable by the naked eye, thus allowing for quantification of target biomarker concentrations at the point of care.

The clinical diagnosis of urinary tract infections (UTIs) faces challenges characterized by prolonged test results, contributing to over-prescription of antibiotics, and missed diagnoses. Point-of-care (POC) tests such as the lateral-flow immunoassay (LFA) attempt to alleviate this issue by providing patients with rapid results; however, their effectiveness is limited by low sensitivity. To improve the performance of the LFA, this thesis focuses on the development of an automated device that enables the addition of a sample preconcentration step at the POC using an aqueous two-phase system (ATPS). The device utilizes a photoresistor-based interface detection algorithm to selectively extract the bottom phase of the ATPS, containing concentrated biomarker, thereby avoiding the dispensing of the undesired top phase onto an LFA. The device incorporates a custom syringe pump module to apply the sample fluid to the LFA, and a gravity-based buffer dispensing module to facilitate the application of running buffer following sample addition. The affordability of the device is ensured through the use of reusable electrical components and cost- effective consumables, making it accessible to physician offices and mobile clinics at the POC. Experimental results demonstrate the successful automation of ATPS application onto the LFA, effectively extracting the bottom phase without any of the top phase. This advancement brings sample preconcentration using an ATPS one step closer to clinical adoption, offering a sensitive, rapid, and user-friendly UTI diagnostic solution.

The development of patterning high-resolution microfluidic circuits onto cellulose paper in 2007 initiated widespread research into the use of the paper as a low-cost, easy-to-use alternative substrate over the glass and plastics of traditional microfluidics. Paper, as a porous hydrophilic material, naturally wicks fluid through itself, without the need to external pumps or power sources. The patterning of paper into hydrophobic and hydrophilic regions, now achievable with consumer-grade office printers, allowed the design of new 2D devices, capable of multi-analyte detection. 3D devices, made from multiple stacked layers of paper, offer even more possibilities for complex, multi-fluid routing in smaller overall device footprints. The use of patterned aerosol adhesives are investigated as an improved method of attaching multiple paper layers together rapidly and with minimal interference of interlayer fluid transport. Patterned aerosol adhesives also enable the development of nonplanar 3D devices, which represent a novel platform upon which to develop new microfluidic devices, which would otherwise be impossible to construct or function in a planar device.Much of paper microfluidics research is focused on developing more sophisticated detection methods that provide quantitative data, instead of simple colorimetric qualitative yes/no answers. Frequently quantification is obtained by scanning the device and performing a color intensity analysis to relate a color change to concentrations of a target analyte. This technique suffers due to variations in the quality of imaging equipment and the ambient lighting conditions during image acquisition. To address this, some have proposed a distance-based lateral flow device, where the distance traveled by a colored substance is proportional to the target analyte concentration. The use of a microsphere aggregation-based sandwich assay was investigated for semi-quantitatively determining the concentration of a target ssDNA strand.

Advances in diagnostic technology and microscale analysis are transforming the field of biomedical research, enabling more precise and efficient detection and analysis of diseases and biological processes. This work explores innovative approaches in point-of-care diagnostics, microfluidic platforms, and high-throughput screening methods to address challenges in disease diagnosis and cellular analysis.In Chapter 1, we present a point-of-care (POC) serological test for Lyme disease (LD), using synthetic peptides specific to LD antibodies on a paper-based platform for rapid and cost-effective diagnosis. By targeting antigenic epitopes conserved across Borrelia burgdorferi genospecies, we develop a multiplexed panel to detect LD-specific IgG and IgM antibodies from patient sera. The combination of multiple peptide epitopes and a machine learning-based diagnostic model achieves high sensitivity without sacrificing specificity. Blinded validation with samples demonstrates 95.5% sensitivity and 100% specificity, accurately matching lab-based two-tier testing and differentiating LD from similar diseases. This diagnostic test has the potential to replace the cumbersome two-tier testing, improving early diagnosis and treatment while enabling better immune monitoring and surveillance.In Chapter 2, we discuss how Lab on a Particle (LoP) platforms leverage microparticles for microscale reactions and molecular and cellular analysis. We describe the evolution of hydrogels for single-cell encapsulation and the emergence of hollow core-shell microparticles that are highly parallelizable and analyzable with standard lab instruments like flow cytometers and microscopes. These particles can be used to design custom screening workflows for cell and molecular screening pipelines, enabling high-throughput, massively parallelized analysis and advanced assay capabilities.In Chapter 3, we present a high-throughput function-first screening platform using core-shell microparticles to compartmentalize individual variants into suspendable colonies for large-scale analysis. By integrating this approach with flow cytometry and deep mutational sequencing, we screen extensive libraries of GCaMPs, genetically encoded calcium indicators, achieving unprecedented throughput by assessing millions of variants within hours. This method allows for the selection based on multi-functional criteria and identified GCaMP variants with over a 10-fold improvement in dynamic response. This innovative platform holds potential to revolutionize fluorescent biosensor engineering and can be broadly applied to screen cells and other proteins, facilitating the discovery of rare, high-performing variants and measuring complex functions that are challenging with traditional methods.Overall, this thesis presents significant advancements in point-of-care diagnostics, microfluidic platforms, and high-throughput screening technologies that address current limitations in disease detection and protein engineering. By introducing novel diagnostic tools for Lyme disease, expanding the capabilities of Lab on a Particle platforms, and developing innovative approaches for screening biosensors, this work lays the foundation for more efficient, scalable, and precise biomedical applications. The technologies developed have the potential to not only improve clinical diagnostics but also accelerate research in discovery, cellular analysis, and synthetic biology. The commercialization and widespread adoption of these tools could bridge critical gaps in healthcare and biotechnology, ultimately improving patient outcomes and driving progress in biomedical research.

In order to more effectively diagnose and treat both communicable and non-communicable diseases, it is crucial to develop diagnostic tests that are accessible at the point of care. This is especially pertinent in resource-limited settings areas, where there is a need for affordable, rapid, and robust diagnostic tests due to limited access to healthcare services and laboratory-based tests. One diagnostic modality that is well-suited for these settings is the lateral-flow immunoassay (LFA), which saw widespread use during the COVID-19 pandemic. However, the applicability of the conventional LFA to other diseases and conditions is limited by its inability to provide a quantitative readout, which this thesis aims to address.First, we sought to improve the quantitative capabilities of the LFA by utilizing the multicolor etching of gold nanorods (GNRs). As a proof-of-concept, our group had previously combined the LFA with GNR etching in suspension to produce a multicolor readout that quantifies the concentration of digoxin in human serum. To make the assay more point-of-care friendly, we expanded upon this work in this thesis by adapting the workflow to be fully paper-based, which involved designing an all-in-one 3D-printed casing that combined both LFA detection and GNR etching via novel color-changing GNR pads. The resulting color hues from the GNR pads are easily distinguishable by the naked eye, thereby enabling a workflow that easily and quickly quantifies target biomarker concentrations in point-of-care settings.Next, we turned to utilizing the chemiluminescence reaction of luminol as another method to improve the quantitative capabilities of the LFA. After detecting the target biomarker using the LFA, the light output generated from the chemiluminescence reaction can be quantified using a smartphone-based reader to determine the concentration of the target biomarker in the patient sample. To assist with the development of this assay, a mathematical model of the chemiluminescence reaction in the presence of phenolic enhancers—reagents that can significantly increase the light output of the reaction—was developed in MATLAB to optimize the reaction parameters and conditions.

Aqueous two-phase systems (ATPSs), traditionally utilized in industrial bioseparations, are showing increasing potential as an approach for concentrating components in paper-based point-of-care bioassays. Specifically, our lab was the first to demonstrate that ATPSs can increase the sensitivity of the lateral-flow immunoassay (LFA) by concentrating biomarkers into one of the two phases, after which, the phase containing the concentrated biomarkers was extracted prior to their detection. While this method demonstrated consistent improvements to the LFA, its applicability in point-of-care settings was restricted by two main factors. First, the time to separate into two distinct phases varied among different classes of ATPSs, but usually required hours to achieve effective biomolecule concentration. Second, the method required several user steps in the form of sample mixing with the ATPS components, and the subsequent extraction and application of the phase containing the concentrated biomolecule to the LFA after phase separation occurred. More recently, our lab demonstrated that when the mixed, homogenous ATPS was applied to a paper membrane, phase separation was observed within the paper membrane itself as the solution wicked across the paper. This largely unexplored phenomenon reduced the phase separation time of a polyethylene glycol (PEG)-salt ATPS from hours in a tube to minutes on paper. This thesis focuses on advancing the phenomenon of paper-based phase separation as a means of enhancing the phase separation rate of ATPSs and making them more suitable for point-of-care applications. First, we extended the paper-based phase separation phenomenon to a naturally slow phase separating system, the Triton X-114 ATPS. Next, we investigated the dehehydration of the phase forming components directly into the paper matrix, and the subsequent phase separation upon resolubilization of the components by a liquid sample. Within these two investigations, the ATPS was then integrated with the LFA to improve the detection of biomarkers of infectious diseases such as malaria and chlamydia. Furthermore, we investigated the use of the Washburn equation as a mathematical framework to describe the flow behavior of ATPS phases within porous media in order to better predict phase separation behavior within paper. The Triton X-114 ATPS is a micellar ATPS that is comprised of the Triton X-114 nonionic surfactant. This particular ATPS, previously used by our lab to concentrate biomarkers for the LFA, is one of the slowest separating systems, partially due to the small interfacial tension and density difference between its two phases (the micelle-poor and micelle-rich phases). We applied the Triton X-114 ATPS to paper membranes and demonstrated, using a distinctively different design in which the solution flows vertically up a multilayered paper wick, that paper-based phase separation can also be achieved with the Triton X-114 system. In this case, we found that gravitational effects had no influence on the flow of the dense gold nanoparticles as the less dense micelle-poor phase containing the gold wicked ahead of the micelle-rich phase. This was the first time that a micellar ATPS was applied directly to a fiberglass paper membrane to significantly speed up its macroscopic separation from at least 8 hrs in a test tube to approximately 3 min on paper. The paper-based Triton X-114 ATPS was then integrated with the LFA to simultaneously concentrate a malaria protein biomarker into the leading micelle-poor phase, and then detect it without the need of a user-dependent phase extraction step. The single-step integration improved the LFA detection limit for the protein by 10-fold in buffered saline and complex serum media. This was also the first time within our lab that we concentrated an infectious disease biomarker in complex biological fluids. The design used for the abovementioned Triton X-114 study is applicable when dealing with oral and vaginal swab samples as the swab would need to be mixed with a buffer solution to solubilize the target. However, in urine, saliva, or blood applications, one would prefer to just add the biological fluid to the device. The second focus of the thesis was therefore to further improve the user-friendliness of the ATPS-LFA integration by removing initial sample preparation steps. To achieve this, we investigated the novel concept of sequential rehydration of the two-phase components that were initially dehydrated into the paper matrix as a way to achieve paper-based ATPS phase separation. We used two different polymer-salt ATPSs: the PEG-salt system and the UCON-50-HB-5100 (UCON)-salt system, optimizing the component concentrations and rehydration order to yield the appropriate phase separation conditions for each system. Upon rehydration of the components, phase separation successfully occurred within the paper, leading to the formation of a leading polymer-poor phase and a lagging polymer-rich phase. As mentioned above, the benefit of this method is that a biological sample no longer needs to be manually mixed with the components and instead can be directly added to the device. These dehydrated systems were then integrated with the LFA to produce paper-based assays in which all the necessary components for target concentration and detection were stored within the paper matrix. The dehydrated PEG-salt ATPS and UCON-salt ATPS were integrated with the LFA to simultaneously concentrate and detect large Chlamydia trachomatis whole bacteria and smaller human IgM antibodies, respectively. Ultimately, our designs demonstrated 10-fold improvements to the detection limits for both Chlamydia trachomatis and IgM, and could therefore improve LFA sensitivity without adding steps to the user. These exciting developments solve the initial limitations of the ATPS-LFA integration, making their use as point-of-care diagnostic devices for infectious diseases one step closer to reality. Although paper-based phase separation of several different aqueous two-phase systems have been experimentally demonstrated, there is a need for a mathematical model that can accurately predict two-phase system wicking behavior in paper-based devices, as it would benefit the process of device design. We decided to evaluate the Washburn model as a framework for fluid flow of the PEG-salt ATPS and Triton X-114 ATPS in porous media. Using a combination of imbibition studies and characterization of the Washburn fluid parameters for each individual phase, we determined that the viscosity difference between the two phases is a dominant factor in the ability of a given ATPS to phase separate in paper. More specifically, the Washburn model correctly predicts that the less viscous phase will constitute the leading phase, independent of flow direction in a horizontal or vertical orientation. In this validation of the model, we then applied it to predict the phase separation capabilities of two polymer-polymer systems, the PEG-Dextran system and the PEG-polyacrylic acid (PEG-PAA) system. We predicted and successfully demonstrated that the PEG-PAA system could phase separate due to its large enough difference in phase viscosities, while the PEG-Dextran system could not phase separate due to its small viscosity difference. Furthermore, this theoretical framework was extended to predict the phase separation of ATPSs in polyethylene glycol dimethacrylate-based microporous hydrogels. We were able to predict and show for the first time that phase separation of various ATPSs could be achieved within hydrogels, demonstrating that the phase separation enhancement phenomenon can occur in various types of porous media.

Every year infectious diseases are among the top major causes of death worldwide. Infectious diseases disproportionately impact resource-poor areas. While most developed countries have the ability to screen and quickly detect infectious diseases in humans, animals, and plants, developing countries lack this ability. In such areas, access to medical treatment and testing facilities is severely limited. Additionally, developing countries are more densely populated, which makes infectious disease surveillance exceedingly more difficult. As a result of this, infectious diseases tend to spread more quickly and outbreaks tend to occur more often in developing countries. Additionally, the increasing rate of antimicrobial resistant infections and emergence of novel infectious diseases increase the likelihood of global pandemic occurrence. It is evident that developing novel devices for detection of infectious diseases would be greatly beneficial to global health. It is important to note that developed countries would benefit equally from such a device. Pandemics tend to burden the economy and put strain on resources even in developed countries. The impact of such a device is determined by a set of criteria put forth by the World Health Organization (WHO). Based on the ASSURED criteria, an ideal device for point-of-care (POC) diagnostics must be affordable, sensitive, specific; and yet it must also be user-friendly, rapid and robust, equipment free, and deliverable to end-users. Additionally, this device must be easily adaptable for detection of novel pathogens. The nucleic acid amplification test (NAAT) meets many of these criteria. However, NAAT’s current workflow involves many user liquid handling steps as well as a thermocycler. This thesis aims to address these issues by first automating isothermal nucleic acid amplification, followed by coupling the amplification step with the lateral-flow immunoassay (LFA), which is a paper-based diagnostic device. Thus, taking advantage of the affordability and POC features of the LFA. The aim of these efforts is to work toward the ultimate goal of developing a device that receives the patient sample, then carries out all the steps in NAAT in an automated manner and produces results with just a push of a single button.

Every year infectious diseases are among the top major causes of death worldwide. Infectious diseases disproportionately impact resource-poor areas. While most developed countries have the ability to screen and quickly detect infectious diseases in humans, animals, and plants, developing countries lack this ability. In such areas, access to medical treatment and testing facilities is severely limited. Additionally, developing countries are more densely populated, which makes infectious disease surveillance exceedingly more difficult. As a result of this, infectious diseases tend to spread more quickly and outbreaks tend to occur more often in developing countries. Additionally, the increasing rate of antimicrobial resistant infections and emergence of novel infectious diseases increase the likelihood of global pandemic occurrence. It is evident that developing novel devices for detection of infectious diseases would be greatly beneficial to global health. It is important to note that developed countries would benefit equally from such a device. Pandemics tend to burden the economy and put strain on resources even in developed countries. The impact of such a device is determined by a set of criteria put forth by the World Health Organization (WHO). Based on the ASSURED criteria, an ideal device for point-of-care (POC) diagnostics must be affordable, sensitive, specific; and yet it must also be user-friendly, rapid and robust, equipment free, and deliverable to end-users. Additionally, this device must be easily adaptable for detection of novel pathogens. The nucleic acid amplification test (NAAT) meets many of these criteria. However, NAAT’s current workflow involves many user liquid handling steps as well as a thermocycler. This thesis aims to address these issues by first automating isothermal nucleic acid amplification, followed by coupling the amplification step with the lateral-flow immunoassay (LFA), which is a paper-based diagnostic device. Thus, taking advantage of the affordability and POC features of the LFA. The aim of these efforts is to work toward the ultimate goal of developing a device that receives the patient sample, then carries out all the steps in NAAT in an automated manner and produces results with just a push of a single button.

Over the past decades the dramatic increase in computational resources coupled with the advent of machine learning and artificial intelligence has profoundly impacted sensor technology. Together, these trends present a new opportunity for data-driven computational sensor design, where acquisition hardware is fundamentally changed to “lock-in” to the optimal sensing data with respect to a user defined cost function. This dissertation addresses this emerging opportunity, referred to herein as computational sensing, with a focus on applications in biophotonics and diagnostics where there is currently high-demand for cost-effective and compact sensing systems for democratizing global health. Therefore, several examples of computational sensing systems will be presented, all of which leverage machine learning and data-driven design principles to enable low-cost, compact, and sparse implementations well-suited for the point-of-care among other settings. Firstly, I will discuss a multiplexed paper-based vertical flow assay and reader for biomarker quantification at the point-of-care. I will show how deep learning and statistical analysis of the multiplexed sensor response can be leveraged to improve accuracy, dynamic range, and limit-of-detection, greatly enhancing the sensing capabilities beyond that of the traditionally employed lateral flow assay. Additionally, I will demonstrate feature selection techniques which can serve as a toolbox for iterative assay development, sensor design, and cost-performance optimization. Two applications in point-of-care diagnostics will be presented with this platform, one being a clinical study for early-stage Lyme Disease diagnoses, and the other, a clinical study for cardio-vascular risk assessment by the high-sensitivity C-Reactive Protein test. Secondly, Localized plasmon resonance (LSPR) sensors and their corresponding readers will be discussed as computational sensing systems for label-free biomolecular sensing. Through machine learning of spectral characteristics, I will present how the corresponding LSPR sensor reader can be jointly designed with the LSPR sensor to optimize read-out accuracy while minimizing cost by computationally selecting from cost effective illumination optics. Lastly, a fundamentally different hardware design for spectroscopy, enabled by deep learning, will be demonstrated and discussed for application-specific performance advantages in terms of acquisition rate, form factor, and cost.

Chemoselective reactions are powerful tools for the progression of basic science and biochemical technologies. Bond-forming reactions are routinely used for monitoring the status of cellular activity and in the production of precision therapeutics and diagnostics. However, the development of such reactions is a challenging feat, as one must account for the broad swathes of interfering chemical entities present within the biological milieu. One way to address this complicating factor is through the use of “bioorthogonal” chemistry. These reactions involve two reactive groups with xenochemical properties, i.e., reactivity preferences that are inert to biological moieties yet exquisitely reactive with one another. By necessity, bioorthogonal reactions are forged from deep chemical insight. A complementary strategy to achieve reaction selectivity in biological settings seeks to use proximity enhancement. Positioning two groups in close proximity to one another raises the higher effective local concentration of the desired partners, increasing the rate by which they react to afford highly selective conjugation, even amidst complex systems. In this dissertation, I explore several applications of this concept to develop a variety of tools for basic bioscience and clinical diagnostics. In Chapter 1, I provide a brief overview of several landmark papers that employ proximity-enhanced reactions in chemical biotechnology.In Chapter 2, I describe the use of proximity-enhanced reactivity to direct ligation and detection of protein-specific glycosylation on live cells. In Chapter 3, I describe the application of proximity enhancement to selectively accelerate the ligation of DNA strands that have been chemically attached to different antibodies. In this particular project, we used paired antibodies that bind either to biotin or a protein of interest, driving the ligation and formation of a DNA amplicon in the presence of a biotinylated protein. In combination with chemistries that selectively attach biotin to glycans, this technology allows the use of qPCR to detect very low levels of glycosylated proteins. In Chapter 4, I describe the development of a new technology, termed Antibody Detection by Agglutination-PCR (ADAP), which uses PCR to detect antibodies with a high degree of multiplexing and sensitivity.In Chapter 5, I describe a method for re-purposing cyclooctynes to selectively ligate to thiols present on serum albumin as a strategy to improve the serum-half lives of therapeutic molecules.In Chapter 6, I describe the computational modeling and synthesis of dendrimer-based glycan mimetics.