Chimeric antigen receptor (CAR) T-cell therapies present transformative potentials for progressive and refractory cancer treatment. However, therapy-associated neuronal toxicities, cytokine release syndromes, relapse rates, and the complex responses of patients and medical management have increased the cost of patient care. Prompt point-of-care (POC) quantification of circulating CAR T-cells and associated cytokines could enhance safety, simplify patients' management, and decrease patient care costs. While effective, existing standard detection methods, such as Enzyme-Linked Immunosorbent Assay (ELISA), quantitative Polymerase Chain Reaction(qPCR), and Flow cytometry, are not conducive to quick POC testing due to their complexity and expense. This research introduces a centrifuge-free Rapid Optical Imaging (ROI)-based platform to quantify CAR T-cells and therapy-related cytokine (Interleukin-6) from a single drop of whole blood. Through machine learning, label-free ROI-based CAR T-cell detection has been improved for accuracy compared with fluorescent staining results, and the morphological characteristics of CAR-T cells have been applied to attribute for differentiation and reduce false positives. This multi-layered microfluidic chip integrates cell and cytokines separation, collection, and detection steps, reducing the need for centrifugation or staining procedures. The microfluidic channel system separates white blood cells from whole blood after red blood cell agglutination and membrane filtration. The non-agglutinated samples are then extracted into a subchannel with a functionalized sensor surface for CAR-T-specific detection. Calibration curves were established using blood samples spiked with varying CAR-T cell concentrations. Another subchannel, featuring dual-layer membrane filtration, has been designed for cytokine detection using gold nanoparticle-labeled detection antibodies. Cytokine concentrations are digitally measured by tracking the number of gold nanoparticles in designated zones. This platform aims to offer a rapid and cost-efficient prognostic tool for timely assessment of key molecular and cellular biomarkers of CAR-T therapy patients, facilitating timely and evidence-based treatment adjustments.

Over the past four decades, DNA nanotechnology has grown exponentially from a field focused on simple structures to one capable of synthesizing complex nano-machines capable of drug delivery, nano-robotics, digital data storage, logic gated circuitry, nano-photonics, and other applications. The construction of these nanostructures is possible because of the predictable and programmable Watson-Crick base pairing of DNA. However, there is an increasing need for the incorporation of chemical diversity and functionality into these nanostructures. To overcome this challenge, this work explored creating hybrid DNA nanostructures by making self-assembling small molecule/protein-DNA conjugates.In one direction, well studied host-guest interactions (i.e. cucurbituril[7]-adamantane) were used as the choice of self-assembling species. Binding studies using these small molecule-DNA conjugates were performed and thereafter they were used to assemble larger DNA origami nanostructures. Finally, a stimulus responsive DNA nano-box that opens and closes based on these interactions was also demonstrated. In another direction, a trimeric KDPG aldolase protein-DNA conjugate was probed as a structural building block by assembling it into a DNA origami tetrahedron with four cavities. This hybrid building block was thereafter characterized by single particle cryo-EM and the resulting electron density map was best fit by simulating origami cages with varying number of proteins (ranging from 0 to 4). Next, to increase access and for larger democratization of the field, an automation designer software tool capable of making DNA nanostructures was made. In this work, the focus was on making curved 3D DNA nanostructures. The last direction probed in this work was to make optical metamaterials based on complex 3D DNA architectures. Realization of a self-assembled 3D tetrastack geometry is still an unachieved dream in the field of DNA self-assembly. Thus, this direction was probed using DNA origami icosahedrons. Finally, the work covered in my thesis probes multiple directions for advancing DNA nanotechnology, both fundamentally and for potential applications.