Microscopy image processing is critical for precision oncology and immunotherapy, two approaches in cancer treatment often combined to enhance patient outcomes. Numerous scientists have studied the effects of drugs on the immune system and tumors. To quantify the impact of various drugs on different immune and cancer cell types, medical researchers conduct ex vivo assays. In these assays, patient-derived live cells are placed in an artificial microenvironment where drug responses are monitored over time. Brightfield microscopy captures one image every half hour of numerous patient-derived cells in an ex vivo reconstruction of the tumor microenvironment treated with 31 drugs for up to six days. These images are used to quantify the response of different cells to various drugs. However, tracking thousands of cells in low-resolution, low-frame-rate images remains challenging. Existing digital image processing algorithms require high-resolution images involving very few cells from homogeneous cell line populations. In this work, we propose three novel frameworks to track cancer cells, capture their behavior, and quantify cell viability to inform clinical decisions in a high-throughput manner.

The cover of the heart, or epicardium, consists of a single layer of mesothelial cells. During cardiac development, epicardial cells undergo Epithelial-to-Mesenchymal Transition (EMT) to form multipotent precursors known as epicardial-derived cells (EPDC). The EPDC migrate into myocardial tissue (containing cardiomyocytes) and subsequently differentiate into fibroblasts, myofibroblasts, and smooth muscle cells. In adult hearts, a similar process of epicardial cell proliferation, migration, and differentiation occurs after myocardial infarction (MI, heart attack). EPDC differentiation into vascular endothelial cells or cardiomyocytes is rare and not well understood. Recently, we observed that running (exercise) in mice promotes differentiation of EPDC into microvascular endothelial cells (CD31+). After running, EPDC appear to generate endothelial cells and not other cardiac cell types. Of interest, running promotes cardiac hypertrophy that requires additional perfusion (blood flow) and may therefore stimulate the contribution of EPDC to capillaries. We hypothesized that running exercise induces gene expression in epicardial cells that promotes endothelial specification. To test our hypothesis, we developed an efficient method to directly isolate primary adult epicardial cells from the heart cover based on their expression of integrin-β4 (CD104). After 2 hours of protease digestion, we used Magnetic-Activated Cell Sorting with antibodies against CD104 (CD104 MACS) to obtain undifferentiated epicardial cells; this was confirmed by expression of Keratin-18, an epicardial-specific protein in the heart. By cDNA microarray assays and bioinformatics analysis, we compared the gene expression profile of epicardial cells isolated from running-conditioned mice with that of age-matched controls (non-runners). Our data suggest that extracellular matrix remodeling in the heart is mediated, in part, by epicardial cells during running. Furthermore, we identify epicardial gene expression for cell signals/pathways and transcription factors that may enhance vascular perfusion after MI through promoting angiogenesis or endothelial specification of epicardial derivatives.

Amoeboid sperm from Ascaris contain an unusual cytoskeleton, composed of major sperm protein (MSP). In spermatozoa MSP filaments are organized into fiber complexes that extend from the leading edge to the base of the pseudopod and flow rearward at the same rate as the cell crawls forward. The tight coupling between cytoskeletal flow and motility indicates a key role for the MSP cytoskeleton in locomotion. I have studied both the regulation and structure of the MSP filament system in Ascaris sperm. Changes in intracellular pH (pH$\sb{\rm i}$) are involved in controlling the assembly status of MSP in vivo. Spermatozoa established a pseudopodial pH gradient with pH$\sb{\rm i}$ 0.15 units higher at the leading edge, where fiber complexes assemble, than at the base where disassembly occurs. Agents that abolished this gradient, such as weak acids, caused the cell to stop crawling and the fiber complexes to disassemble. The correlation between elevated pH$\sb{\rm i}$ and MSP assembly and low pH$\sb{\rm i}$ and depolymerization also was observed in developing sperm, indicating that pH regulation of the cytoskeleton also occurs during spermatogenesis. To understand cytoskeletal dynamics in greater detail, I studied the structure of MSP polymers. Self-association of monomers into helical assemblies is an intrinsic property of MSP. This feature is clearly displayed in the number of polymorphic forms MSP assumes in vivo and in vitro. We have characterized 4 distinct assemblies--subfilaments, filaments, macrofibers, and fiber complexes. Individual filaments formed either in vivo or in vitro are composed of two subfilaments wrapping around each other. These filaments coil around one another to form larger helical assemblies, macrofibers in vitro and fiber complexes in vivo. Filaments do not require accessory proteins to interact. Self-association of MSP filaments into / helical superstructures may have important implications for the role of MSP in sperm motility. This property could allow the filaments to form fiber complexes without requiring accessory proteins and, thus, create a simple framework to propel locomotion. / Source: Dissertation Abstracts International, Volume: 54-12, Section: B, page: 6023. / Major Professor: Thomas Michael Roberts. / Thesis (Ph.D.)--The Florida State University, 1993.

Two isozymes of creatine kinase (CK) were partially purified from isolated intestinal epithelial cells and identified as the non-muscle cytoplasmic isoform (B-CK) and the non-muscle mitochondrial isoform (Mi-CK). In intestinal epithelia, mitochondria (and Mi-CK) are excluded from the dense cytoskeletal web underlying the microvillar brush border. There are two myosin II subsets in this domain. The first cross-links the descending microvillar rootlets while the second is organized within a circumferential contractile ring. In glycerinated cells and isolated brush borders both interrootlet myosin solubilization and ring contraction were supported by the addition of creatine phosphate (Cr-P) to the endogenous B-CK-based ATP-regenerating system. When an exogenous ATP hydrolysis system was added (hexokinase and 25 mM glucose) both solubilization and contraction were supported in glycerinated cells, but only contraction was supported in isolated brush borders. Immunofluorescent imaging of B-CK revealed a marked loss in isolated brush borders compared to that of glycerinated cells, indicating that the interrootlet myosin-coupled B-CK subset had been extracted. These differential B-CK-myosin interactions were exploited to demonstrate ATP compartmentation between B-CK and ring myosin. Comparisons of epithelia from duodenum, jejunum, and ileum reveal that both contractile ring and interrootlet myosin-B-CK interactions are maintained despite the lower endogenous B-CK concentrations observed in the jejunum and ileum. / Source: Dissertation Abstracts International, Volume: 53-07, Section: B, page: 3260. / Major Professor: Thomas C. S. Keller, III. / Thesis (Ph.D.)--The Florida State University, 1992.