Search

Tuberculosis (TB) is a global health concern that affects approximately 10 million people worldwide (WHO, 2018). Current in vitro TB models are low throughput and/or lack caseation, which impairs drug effectiveness in humans. Here, we report the development of THP-1 human monocyte/macrophage spheroids housing mycobacteria (TB spheroids). These TB spheroids have a necrotic core co-localized with mycobacteria and are hypoxic. TB spheroids exhibit higher levels of pro-inflammatory factor Tumor necrosis factor �� (TNF��) and growth factors such as Granulocyte colony stimulating factor (G-CSF) and Vascular endothelial growth factor (VEGF) when compared to non-infected controls. TB spheroids show high levels of lipid deposition, characterized by MALDI mass spectrometry imaging. TB spheroids infected with strains of differential virulence, Mycobacterium tuberculosis (Mtb) HN878 and CDC1551 vary in response to isoniazid and rifampicin, first line anti-TB agents. We then adapted the spheroid model to form peripheral blood mononuclear cells (PBMCs) and lung fibroblasts (NHLF) 3D co-cultures. These results demonstrate that spheroids can be applied for TB disease modeling and studying the granuloma biology.In this thesis, we demonstrate the application 3D TB spheroids to study anti-TB drug responses using two orthogonal scaled screening approaches. First, we generated methods to collect single spheroid level data from the Aggrewell 400��� spheroid forming system to understand inter-spheroid variability under the same drug conditions. Here, we tracked the mean fluorescent intensity of BCG engineered with a mCherry reporter to assess the drug response of individual spheroids to isoniazid and rifampicin. This method aids in studying heterogeneity in single drug response across multiple spheroids within the same grid. As an alternative approach, we also developed 3D TB spheroid assays in conventional 384 well formats which requires lower working volumes, making it adaptable for high throughput chemical library screens. Using this 384 well assay, we demonstrated that isoniazid, a drug that penetrates freely into the caseum, does not show differences in drug response between 2D and 3D assay formats. However, rifampicin showed lesser reduction in BCG mCherry in the 3D assay format when compared to 2D. These results collectively indicate that the 3D TB spheroids can be used for studying drug response in vitro. Finally, we evaluated if the 3D TB spheroid model can be used to for modeling co-infection with Human Immunodeficiency Virus (HIV). In TB patients, HIV remains one of the largest contributors to the reactivation of latent tuberculosis (TB). Current 3D aggregate models have not been adopted to study HIV-TB interactions. Therefore, we developed a dual infection 3D TB spheroid model by coinfecting THP1 macrophages with BCG mCherry and a pseudotype HIV using a 96 well plate spheroid forming assay. In this HIV-TB co-infection model, we found an increase in BCG mCherry growth within the 3D spheroids in the presence of HIV pseudotype. The degree of disruption of the spheroid formation was directly proportional to the viral load. In addition to this, we also observed an increase in cytokines, Interleukin 1- receptor antagonist (IL-1Ra), Interleukin 8 (IL-8), Interferon gamma-induced protein 10 (IP10), Macrophage inflammatory protein (MIP 1�� and ��). These host-pathogen interaction studies can aid in the discovery of therapeutic compounds for HIV-TB co-infections in the future. Collectively, these studies enabled us to develop TB mimetic in vitro 3D models that display key granuloma features such as central necrosis/caseum, lipid deposition, cytokine profiles, peripheral fibrosis and show drug response trends similar to the results reported in vivo. From an assay development standpoint, we have been able to establish the spheroid model in a high volume, high throughput Aggrewell 400��� spheroid plate which generates around thousand spheroids within a single grid. We have also been able to scale down the assay to low volume, high throughput 96 and 384 well formats These results provide a comprehensive approach that can be applied to high throughput screening to identify anti-TB compounds in the future.