Yes-associated protein (YAP) is a co-transcription factor that acts downstream of the evolutionarily conserved Hippo pathway. Canonically, this pathway regulates tissue growth in flies and mammals, by controlling the nuclear localization of YAP. Interestingly, in addition to the conserved functions of this pathway, some of the mammalian orthologs of pathway components (e.g. MST, RASSF1, WW45, and LATS) have been shown to localize to the nucleus and alterations in their expression induces alterations in mitotic processes, suggesting additional roles for these proteins in mitosis. In this thesis, I have uncovered a role for the Hippo pathway effector protein, YAP, in cytokinesis. YAP was found to localize to the central spindle and cytokinetic midbody and biochemical analysis demonstrated that YAP is phosphorylated by the mitotic regulatory kinase CDK1 during mitosis. Time-lapse microscopy of cells in which YAP was downregulated by shRNA revealed that reduction in YAP expression causes a delay in abscission and induces a cytokinesis phenotype associated with increased contractile force, membrane blebbing and bulges, and abnormal spindle orientation; consequently, this leads to an increased frequency of multinucleation, micronuclei, and aneuploidy. Expression of or expression of a variant of YAP that could not be phosphorylated at the mitotic phosphoacceptor sites induced a phenotype similar to that of YAP knockdown, suggesting that mitotic YAP phosphorylation is critical for YAP’s function in cytokinesis. Reduction in YAP expression also disrupted the localization of ECT2, MgcRacGap, Anillin, and RHOA, proteins important for cleavage furrow function during cytokinesis, Reduction of YAP also increased levels of phosphorylated myosin light chain, which activates myosin II contractile activity. These findings suggest that YAP is required for proper coordination of these contractile processes involved in cytokinesis. In addition, the YAP mitotic phosphorylation sites are required for interaction with the scaffold polarity protein PATJ, and PATJ co-localizes with YAP at the cytokinesis midbody. PATJ knockdown induces cytokinesis defects and spindle orientation alterations similar to those detected in YAP- depleted cells or cells expressing a non-phosphorylatable mutant of YAP. This study reveals an unanticipated role for YAP during mitosis and implicates YAP in processes that control the proper organization of cytokinesis machinery through interaction with the polarity protein PATJ. Thus, these studies demonstrate a previously unanticipated role for YAP that is independent of its activity as a transcriptional coactivator. In addition, although YAP is known to function as a potent oncogene, our findings indicate that YAP may also act as a tumor suppressor in certain contexts since loss of YAP could lead to genetic alterations associated with defective cytokinesis. These studies add to the complexity of YAP regulation in cancer as well as in normal development and provide a framework for future studies in a new area of Hippo pathway biology. / Medical Sciences

The transcription factor Sox2 is essential for the establishment and maintenance of multiple stem cell populations and its coding region is amplified in certain carcinomas. However, Sox2’s role in stomach homeostasis and cancer is poorly understood. In this thesis, I used mouse genetics to investigate the expression pattern and function of Sox2 in the adult stomach during normal tissue homeostasis and tumorigenesis. Using a genetic lineage tracing system, I found that Sox2 expression marks a gastric stem cell population capable of self renewal and differentiation throughout the lifetime of a mouse, raising the key question of whether Sox2 itself is required for adult stomach function. Using a combination of novel mouse models, I examined the consequences of Sox2 loss of function on stomach regeneration as well as the susceptibility of Sox2+ cells to transformation. Surprisingly, I found that Sox2 itself is dispensable during stomach homeostasis, although Sox2-expressing cells readily give rise to Wnt-driven adenomas. To gain insight into the molecular function of Sox2, I performed ChIP-Seq analysis which revealed that the majority of Sox2 targets in mouse gastric stem and progenitor cells are related to tissue-specific functions such as endoderm development, Wnt signaling and gastric cancer while only a small set of genes overlaps with targets occupied by Sox2 in other stem cell populations. SOX2 has been described as an amplified oncogene in several types of human cancers derived from the foregut endoderm including lung and esophageal squamous cell carcinomas. Unexpectedly, I found that Sox2 loss enhances stomach tumor formation and organoid growth in an Apc/Wnt depdendent adenoma mouse model. Using a reporter assay, I further showed that altered Sox2 levels modulate Tcf/Lef-dependent transcription, providing a molecular explanation for the observed proliferation phenotypes. In summary, my genetic and molecular studies offer insight into how Sox2 regulates stomach tissue homeostasis and cancer and evidence that Sox2’s mode of action is context and tissue specific. / Medical Sciences

One of the most commonly mutated genes found in cancer is the tumor suppressor p53. p53 is a transcription factor capable of inducing cell-cycle arrest, apoptosis, senescence, and other cellular processes thought to halt the progression of a nascent cancer. As part of a stress signaling pathway, p53 is acutely activated by ionizing radiation and the formation of DNA double-strand breaks. The appearence of this DNA damage causes the concentration of p53 within the nucleus to fluctuate and pulse regularly, which can be observed in single cells using fluorescence time-lapse microscopy. From the time this was first discovered, the connection between these p53 dynamics and p53 function has been speculated upon. A key insight into this connection came from a Lahav Lab publication that demonstrated the act of pulsing, itself, controls p53-dependent transcription and cell fate. The mechanisms and molecular details behind this relationship are now an area of intense study. Another area of high interest is the broader characterization of p53 dynamics in different time-scales, genetic backgrounds, and stresses. These lines of research each depend upon single-cell measurements that are often time consuming, noisy, and yield small sample sizes. The ongoing development of experiemental and computational tools for single-cell biology is needed to overcome these limitations. In the publication referenced earlier, a novel method was created to measure p53 dynamics and gene expression in the same cell. In a seperate study characterizing p53 dynamics over long time-scales, semi-automated tracking software aided in the discovery of new p53 dynamics: sustained elevation of p53 levels that follow a period of pulsing. Population measurements showing similarly elevated p53 levels on the same time-scale are shown to depend on the late induction of the p53-target PIDD. / Systems Biology

The myeloid differentiation primary response gene 88 (MyD88) is an adaptor protein proximally downstream of the Toll-like receptor (TLRs) and the IL-1 receptor family (IL-1R, IL- 18R and IL-33R). The TLR-MyD88 signaling pathway enables the innate immune system to sense inflammation and promote adaptive immune responses. TLRs are also expressed on T cells, and we have previously demonstrated that T cell intrinsic MyD88 signaling promotes T cell survival and is required for optimal responses to pathogens such as T. gondii and LCMV. Surprisingly however, we observed that mice with a targeted T cell deletion of MyD88 (MyD88fl/fl x CD4-Cre, termed MyD88-ΔT) reject bm12 cardiac allografts, and bm12 skin allografts when treated with αCD154 and rapamycin, at a higher frequency than wild type (WT) mice. As Tregs are critical for graft prolongation following costimulatory blockade in this model, we hypothesized a “defect” in MyD88-ΔT Tregs in the above model. As MyD88-ΔT mice do not allow us to differentiate the effect of MyD88 deletion specifically on Tregs versus the T cell compartment as a whole, we created mice with a targeted deletion of MyD88 specific to Tregs (MyD88fl/fl x FoxP3-Cre-YFP, MyD88-ΔTreg). MyD88-ΔTreg mice, similar to MyD88-ΔT mice have a similar Treg frequency when compared to WT mice and have no observable signs of autoimmunity. We found that MyD88-ΔTreg mice failed to accept bm12 skin allografts at the same frequency as WT control mice when given CoB, suggesting that the phenotype we observed in the MyD88-ΔT mice was likely caused by MyD88 deficient Tregs. As T cell intrinsic MyD88 is required for T cell survival during the response to pathogens, we thus hypothesized that the inability for MyD88 deficient Tregs to promote long-term allograft survival was due to an inability of these cells to survive. Surprisingly, through a series of in vivo and in vitro experiments, we observed that MyD88 deficient Tregs survive as well as WT Tregs under competitive conditions with WT Tregs, during activation with αCD3 and αCD28, as well as in the allograft 21 days post transplantation. In addition, MyD88 deficient Tregs survived as well as WT Tregs when activated and cultured with rapamycin and αCD154. In an attempt to reconcile the inability of MyD88-deficient Tregs to promote graft survival with their surprising ability to survive as well as WT Tregs, we next assessed the function of MyD88-deficient Tregs. In an in vitro suppression assay, MyD88-deficient Tregs suppressed effector T cell proliferation as well as WT Tregs at multiple Treg:T effector ratios. In addition, MyD88-deficient Tregs expressed similar levels of the Treg functional proteins CTLA-4, PD-1, Granzyme B, Lag-3, GITR, CD39, CD73 and CD28 as WT Tregs, suggesting that the function of MyD88-deficient Tregs may only be impaired during the response to alloantigen. Together, these data define MyD88 as having a divergent requirement for cell survival in non-Tregs and Tregs, and a yet-to-be defined functional requirement in Tregs during the response to alloantigen. / Medical Sciences

Hematopoietic stem cells (HSCs) sustain continuous turnover and maintenance of all blood lineages through organismal lifespan. The extensive regenerative potential of HSCs has been harnessed in transplantation medicine to enable curative therapies for numerous life-threatening conditions that require hematological reconstitution. However, the rarity of HSCs combined with the limited availability of immunologically matched donors have constrained the utility of HSC transplantation whose success and safety depend critically on the quantity donor HSCs; therefore, ex vivo expansion of HSCs has been a highly sought after goal in HSC research. In this thesis, I present a hypothesis driven approach toward identifying cocktails of small molecules that enable ex vivo maintenance and propagation of mouse and human HSCs. Specifically, using HSCs isolated from Fgd5ZsGreen HSC-specific fluorescent reporter mice along with previously identified immunophenotypic markers, I conducted a small scale combinatorial chemical screen of developmental signaling modulators to determine a condition that would preserve immunophenotypic HSCs ex vivo. The screen led to the discovery that murine HSCs can be maintained for at least 14 days ex vivo when the basal media was supplemented with cytokines and the minimal combination of a small molecule inhibitor of TGF-β signaling and two epigenetic inhibitors, namely LSD1 inhibitor and HDAC inhibitor, which were selected based on reports that they may derepress Notch target gene expression. Additionally, metabolic optimizations led to the identification of putrescine as a critical culture supplement for promoting HSC propagation. The three chemicals identified in the murine screen were conserved in their ability to promote the ex vivo preservation of primary human HSCs. However, as presence of the two epigenetic inhibitors caused substantial growth suppression, alternative, more specific means to activate the Notch pathway were sought. To this end, I hypothesized that inhibition of IKAROS transcription factor, a repressor of Notch target gene expression, would be able to derepress Notch pathway activity and hence replace the growth suppressive epigenetic inhibitors. Indeed, substitution of the epigenetic inhibitors with an IKAROS inhibitor rescued immunophenotypic HSC propagation, and additional supplementation with UM171, a recently identified small molecule enhancer of human HSC expansion, further improved HSC yield as well as the durability of immunophenotypic preservation over prolonged culture; in sum, three compounds, namely a TGF-β inhibitor, pomalidomide, and UM171, were found to be necessary for robust ex vivo maintenance and propagation of human HSCs. At the time of writing, xenotransplantation is underway to assess the in vivo function of ex vivo cultured human HSCs. Collectively, this body of work contributed to identifying chemically defined ex vivo culture conditions supportive of murine and human HSCs while underscoring the importance of combinatorial pathway modulation for maintaining HSC immunophenotype and function. Development of effective HSC culture conditions should not only benefit clinical medicine but also facilitate interrogation and manipulation of HSCs in vitro. / Medical Sciences

Extreme cell polarity places unique demands on long-range transport mediated by the microtubule-based motors dynein and kinesin. Accordingly, deficient protein homeostasis in highly polarized neurons, a process to which these motors contribute through spatial organization of damaged proteins, is closely associated with neurodegenerative disease and cellular aging. Despite this relationship, relatively little is understood regarding how dynein and kinesin meet the challenges of extreme cell polarity to achieve proper protein homeostasis. Furthermore, mechanistic understanding of such processes is limited by a lack of a model system capable of addressing questions both in vivo and in vitro. In this thesis, I use the tractable filamentous fungus Aspergillus nidulans to show that microtubules and dynein mediate the assembly of periodic deposition sites of deleterious aggregated proteins in highly polarized fungal cells as a mechanism of cytoplasmic quarantine. Overwhelming this spatial sequestration pathway yields impaired cellular fitness and global impediments of microtubule-based transport processes. In addition, I demonstrate that the activity of purified A. nidulans dynein and kinesin motors can be observed in vitro, providing a foundation for complementary reconstitution and fine mechanistic study of processes such as aggregate transport. Collectively, these results demonstrate and expand the versatility of filamentous fungi in studying microtubule-based transport in highly polarized cells. / Medical Sciences

Breast cancer is a highly heterogeneous disease, having not only several intrinsic sub-types but also significant sub-clonal heterogeneity within tumors. Intra-tumor heterogeneity can have profound impact on tumor evolution, disease progression and response to therapy. Furthermore, these phenomena can also be influenced by interactions of cancer cells with those of the microenvironment, thereby adding an extra layer of complexity to the study of tumor biology. To investigate the impact of sub-clonal heterogeneity on tumor phenotypes, we developed a heterogeneous mouse xenograft model of breast cancer. Our model revealed that tumor growth can be driven by a minor clone, expressing IL11, in a non-cell autonomous fashion mediated through the microenvironment. We also found that non-cell autonomous driving and clonal interference stabilizes sub-clonal heterogeneity, thereby enabling inter-clonal interactions leading to new phenotypic traits. Utilizing the same model, we identified cooperative interactions between IL11- and FIGF- expressing sub-clones that enhance the metastatic behavior of the tumor as a whole. We found that metastatic cooperation between these two populations result in larger and heterogeneous lung metastasis. Using expression profiles from primary tumors and corresponding metastatic lesions, we identified several key immune-regulatory and extracellular matrix (ECM) remodeling pathways that promote metastasis in our model system. Lastly, we examined heterotypic interactions between tumor cells and cancer associated fibroblasts (CAFs) to understand the mechanism of resistance to lapatinib. Using a 3D co-culture model, we identified significant sub-type-specific changes in gene expression, metabolic, and therapeutic sensitivity profiles of breast cancer cells induced by CAFs. We identified JAK2/STAT3 pathway and CAF-secreted hyaluronan as major factors contributing to CAF-mediated protection. We also found that close spatial proximity to CAFs impacts therapeutic responses by affecting proliferation rates of cancer cells. In summary, we have used in vitro and in vivo models systems to identify key interactions within populations of tumors cells, as well as between tumor microenvironmental components and cancer cells, to identify mechanisms that influence tumorigenesis, metastasis and drug response. We believe that these findings will increase our understanding of breast cancer heterogeneity and enable us to design better therapeutic regimens to eradicate the disease. / Medical Sciences

Diabetes mellitus is a complex and heterogeneous group of metabolic diseases characterized by a disruption in blood glucose homeostasis. This could result from an autoimmune destruction of insulin-producing pancreatic beta cells or insulin resistance with subsequent beta cell dysfunction. Thus, a major goal of diabetes research is developing strategies to replenish beta cells. One emerging and promising strategy is harnessing pancreatic plasticity—the ability of pancreatic cells to undergo cellular interconversions— and numerous studies implicate this phenomenon in diverse states of physiological stress or pancreatic injury. In this study, we investigated the effect of inflammatory cytokine stress on the differentiation potential of ductal cells in a human cell line in vitro; on mouse ductal cells by pancreatic intraductal injection in vivo; and throughout the progression of autoimmune diabetes in the non-obese diabetic (NOD) mouse model. We demonstrate that inflammatory cytokine insults stimulate epithelial-to-mesenchymal transition (EMT) as well as the endocrine program in human pancreatic ductal cells via STAT3-dependent NGN3 activation. Furthermore, we show that inflammatory cytokines activate ductal cell proliferation as well as ductal-to-endocrine cell reprogramming in vivo independently of hyperglycemic stress. Together, our findings provide novel evidence of inflammatory cytokines in directing ductal- to-endocrine cell differentiation and have significant implications for diabetes and cytokine- driven beta cell regeneration. / Medical Sciences

Enhancers are cis regulatory elements that play key roles in the control of cell-type specific gene expression programs. In cancer, enhancer deregulation plays a key role in maintaining gene regulatory programs that underlie an oncogenic state. This dissertation focuses on understanding and modulating aberrant enhancer activity to identify potential vulnerabilities in human cancers. These studies were empowered by evolving technologies in genome-wide measurements of enhancer factors, computational approaches, and chemical and genetic tools to disrupt enhancer function. In high-risk pediatric neuroblastoma, the transcription factor MYCN is frequently amplified and treatment options for these patients are largely ineffective thus establishing the need for improved therapeutic options. To identify previously unrecognized dependencies in neuroblastoma, we generated genome-wide maps of the active enhancer gene regulatory landscape leading to the identification of ID1 as an uncharacterized dependency in neuroblastoma. These results outline a strategy to identify alternative therapeutic avenues based on a holistic understanding of aberrant enhancer activity. While MYCN amplification is the defining feature of high-risk neuroblastoma, a detailed mechanistic understanding of oncogenic transcriptional rewiring has been stalled by a lack of genome-wide binding data. Here we present the dynamic and temporally resolved landscape of genome-wide MYCN occupancy in neuroblastoma. We find that deregulated MYCN binding at enhancers (termed enhancer invasion) is critical to maintaining the oncogenic station and identify the lineage specific transcription factor TWIST1 as a key collaborator and synthetic lethality of oncogenic MYCN. These data suggest that MYCN enhancer invasion shapes transcriptional amplification in neuroblastoma to promote tumorigenesis. The development of small molecule inhibitors of the bromodomain and extra-terminal (BET) family of proteins provides a pharmacological strategy to inhibit enhancer activity. The efficacy of BET inhibition in several cancers has prompted efforts to predict and understand mechanisms of resistance to BET inhibition. Here, we use a newly developed class of small molecules to pharmacologically induce targeted degradation of the BET family. In triple negative breast cancer, we demonstrate that targeted BET family degradation effectively overcomes BET inhibitor resistance. These studies suggest BET degradation as a strategy to overcome BET inhibitor resistance and further disrupt and dissect enhancer activity in cancer. / Medical Sciences

Mitochondria are crucial to the cell and perform numerous functions including generating cellular ATP, buffering calcium, and creating macromolecules. Mitochondria contain their own DNA and as such, cannot be synthesized de novo. Additionally, both nuclear-encoded and mitochondrial-encoded proteins work in concert in order for mitochondria to function. During cell division, the cell goes through dramatic morphological changes to ensure the appropriate amount of DNA is inherited into daughter cells. Given the importance of mitochondria, our study sought to understand the underlying mechanisms that ensure proper mitochondrial inheritance. Organelles are inherited by two mechanisms: active and passive. Through our studies we found that mitochondria go through phases of both active and passive regulation. Initially, mitochondria undergo a release from microtubules, actin, and the endoplasmic reticulum, which allows mitochondria to passively float throughout the cytoplasm. As the cell enters cytokinesis, an active phase occurs that can compensate for some, but not all, asymmetry prior to cytokinesis. The detachment from microtubules, actin, and ER is regulated during cell division, and we discovered the mechanism behind mitochondrial release from microtubules. Dynein and kinesin motors move mitochondria along microtubules and also attach them to the cytoskeleton. During mitosis, motors shed from the mitochondrial surface, thus releasing mitochondria from microtubules. CDK1 releases dynein through phosphorylation, and Aurora A kinase releases kinesin. When exogenously expressed motors are recruited to mitochondria, it results in asymmetric inheritance of mitochondria, a delay in mitotic progression, and cytokinesis failure. Since mitochondria are initially positioned passively, we were able to manipulate mitochondrial positioning before the onset of mitosis, which persisted into the duration of mitosis. Overall, our study elucidates the mechanism behind mitochondrial inheritance. Undoubtedly, the mechanisms are important for normal inheritance into daughter cells. It remains an open question of whether or not directed inheritance can take place through these same mechanisms. / Medical Sciences