Date of Award
Doctor of Philosophy (PhD)
The expansion of our understanding of human health and disease depends on our ability to interrogate the deep complexity of biological tissues with realistic experiments that can be deployed at large scale. To this end, we present an integrative, bottom-up approach towards emulating the human bone marrow in vitro. Our method harnesses the regenerative capacity of adult stem cells to self-assemble a complex, specialized microenvironment of human hematopoietic stem cells (HSCs) in a vascularized three-dimensional microphysiological system. We show that the microengineered niche can reconstitute HSC self-renewal, multilineage hematopoiesis, and complex ligand-receptor signaling pathways of the native human marrow. To demonstrate the advanced application of the bone marrow-on-a-chip, we present i) a specialized model of bone marrow ablation by proton beam radiotherapy and ii) a multiorgan model of innate immune response against bacterial lung infection. This work advances our ability to reconstruct, probe, and deconvolve the complexity of the bone marrow niche, and may enable new capabilities to model human hematopoiesis and immunity for biomedical and pharmaceutical applications. We additionally present a comprehensive pipeline technology to create, interrogate, and harvest living, high-content models of human organ tissues at high throughput. To achieve this, we harness a deeply integrative, multifaceted approach to automation. We first develop a precision robotic manufacturing process and a scalable, multilayered architecture to produce DVD-sized, continuous biomicrofluidic dies with integrated valving at high production yields. We employ this architecture to create a Tissue Disk platform that integrates autonomous fluid control for the longterm culture of 128 independent, fully 3D organ tissue models that can be individually addressed and dynamically manipulated. We interface this high throughput “wetware” platform to an analytical software pipeline that leverages advancements in machine learning to quantify and optimize tissuescale phenomics in a deep, high-dimensional space. To demonstrate the advanced application of the depth and breadth enabled by this pipeline, we iteratively refine a highly specialized and true-to-life model of the vascularized pancreatic islet niche to study beta cell deterioration and endocrine vascular pathology. Finally, we exploit our ability to extract hundreds of these long-lasting, realistic tissues from each Tissue Disk, alive and whole, for regenerative transplantation in a mouse model with a process that we call “organ from a chip.”
Georgescu, Andrei, "Large Scale Integration Of Microengineered Tissue Models For High-Content, High-Throughput Analysis Of Complex Human Physiological Systems" (2021). Publicly Accessible Penn Dissertations. 4600.