To control gene expression, transcription factors (TFs) bind to DNA and recruit regulatory cofactors (COFs) that mediate diverse effects on chromatin, such as DNA and histone modifications. To better help understand TF and COF interactions, our lab has been developing the nuclear extract protein-binding microarray (nextPBM) approach that allows high-throughput characterization of COF recruitment to DNA. An extension of the nextPBM, the human TF array (hTF array), allows COF recruitment to be profiled to binding sites of hundreds of human TFs, providing a method to screen TF-COF complexes forming in cell nuclear extracts. DNA (cytosine-5)-methyltransferase 3 alpha (DNMT3A) and (cytosine-5)-DNA-methyltransferase 3 beta (DNMT3B), two isoforms of the methyltransferase DNMT3, are COFs that are known to be implicated in B-cell lymphoma. hTF experiments for DNMT3A recruitment in B-cell lymphoma cells revealed a number of interactions with known B-cell TFs. In this work, I have attempted to validate the interactions seen on the hTF array by testing TF interactions with both the native DNMT3A/B proteins and the central TF-interacting ATRX-DNMT3-DNMT3L (ADD) domains of DNMT3A/B through the use of an immunoprecipitation (IP) assay in RC B-cell lymphoma cells. I successfully cloned and purified the ADD domains of DNMT3A and B as GST epitope-tagged constructs. As preliminary hTF experiments predicted interactions of DNMT3A with Interferon Regulatory Factor (IRF) proteins, I tested whether the DNMT3 ADD domains interacted with IRF3 present in B-cell lymphoma cell extracts, but was unable to identify an interaction. To examine whether native DNMT3A interacted with the IRF proteins, I performed a native IP using a DNMT3A native antibody. However, I was unable to confirm this interaction by native extract co-IP. Future work to validate DNMT3A/B-TF interactions, and TF-COF interactions identified by the hTF platform more generally, can help us understand at a molecular level what is facilitating transition between normal and cancerous states.

The epithelial-to-mesenchymal transition (EMT) is a critical mechanism during the process of normal embryonic development and wound healing that can be pathologically re-activated during cancer progression. We hypothesized that comparing the transcriptional programs of multiple EMT-driving transcription factors (EMT-TFs) would identify a common set of critical EMT effectors. After elucidating this common transcriptional program, the commonly upregulated RNA binding protein RBMS3 was chosen as a target for functional validation. RBMS3 was both necessary and sufficient for EMT and breast cancer progression, demonstrating the validity of focusing on common EMT-associated effectors. Finally, by evaluating the associations of multiple EMT-TFs with the tumor microenvironment in several solid tumor types, ZEB1 and a ZEB1-regulated transcription program was identified as uniquely associated with immune suppression and poor prognosis. In conclusion, this study significantly advances both the understanding of the mechanisms underlying EMT and the distinct associations of different EMT-TFs with tumor biology and the tumor microenvironment.

In nervous system development, as well as in disease and injury, neurons die through programmed cell death, leaving behind cell corpses which must be removed. The clearance of these corpses is accomplished through phagocytosis, or cell eating. Phagocytosis consists of the recognition, internalization, and degradation of external material. In the nervous system, glial cells act as phagocytes, engulfing dead neurons and debris to ensure proper morphology and tissue homeostasis. Glial phagocytosis has been implicated in several neurological diseases. In humans, increased numbers of phagocytic glia are observed in conditions like Alzheimer’s disease, Parkinson’s disease, and traumatic brain injury. In vitro, glia have been shown to clear protein aggregates like those found in neurodegenerative disease. Moreover, variants of genes implicated in glial phagocytosis have been identified as risk factors for neurodegenerative diseases. However, how phagocytosis defects might cause or worsen neurodegeneration remains unknown. To untangle the links between glial phagocytosis and neurodegeneration, we used the fruit fly Drosophila melanogaster, whose complex nervous system harbors phagocytic glia analogous to those in humans. We analyzed mutant flies lacking the phagocytic receptor Draper and found that they show an accumulation of neuronal cell corpses, which result from developmental programmed cell death and persist throughout the organism’s life. We also found that flies lacking glial Draper display age-dependent neurodegeneration. To determine how phagocytic defects lead to neurodegeneration in the draper mutant, we investigated the hypothesis that persisting cell corpses in the brain lead to chronic increased immunity, resulting in neurodegeneration. This hypothesis stems from the findings that persisting cell corpses in other tissues cause inflammation, and that neuroinflammation is thought to worsen neurodegeneration. We measured activation of the immune pathway Imd in aging draper mutants and found that the antimicrobial peptide attacin A is highly overexpressed in fat body. We then suppressed the Imd pathway by knocking down Relish in glia and fat body in draper mutants and found that neurodegeneration was reduced, indicating that immune activation promotes the neurodegeneration in draper mutants. Taken together, these findings indicate that phagocytic defects lead to or exacerbate neurodegeneration through increased immune signaling, both systemically and in the brain.

The human gastrointestinal (GI) epithelium performs major physiologic functions that are critical to survival, health, and homeostatic equilibrium. While model organisms and in vitro cell culture systems have been widely used to study both normal and disease states of the GI tract, these often fail to fully recapitulate critical features of in vivo intestinal tissue. In recent years, investigators have harnessed the ability to perform directed differentiation of human induced pluripotent stem cells (iPSCs) towards cell types originating from all three embryonic germ layers, most notably a wide variety of endodermal lineages, in an attempt to generate in vitro models that better recapitulate human physiology and key developmental milestones. These iPSC-derived cells contain the exact genetic background of a particular donor or patient and are easily amenable to gene-editing, making them particularly advantageous in comparison to non-human model organisms or in vitro cell culture systems often derived from malignant tissue. Here, we report the efficient generation of iPSC-derived mesenchyme-free human intestinal organoids (HIOs) that can be primed towards colonic or proximal intestinal lineages in serum-free defined conditions. By generating a novel CDX2-eGFP iPSC knock-in reporter line to track the emergence of hindgut progenitors, we follow the kinetics of CDX2 expression throughout directed differentiation, enabling the purification of intestinal progenitors. We employ these mesenchyme free HIOs to highlight cystic fibrosis transmembrane conductance regulator (CFTR) dysfunction using cystic fibrosis (CF) patient-derived iPSC lines before and after correction of the CFTR mutation. We also demonstrate that these HIOs represent a powerful tool to model pathogen-mediated GI illness, characterizing the intestinal epithelial host response to infection by the coronavirus SARS-CoV-2 as well as two filoviruses, Ebola (EBOV) and Marburg (MARV). Finally, we report the generation of a clinically relevant library of iPSCs derived from patients with Crohn’s Disease (CD), including successful directed differentiation of these lines to a relevant immune cell type, as a proof of concept for their use in CD in vitro modeling. Taken together, our results provide a comprehensive and reductive iPSC-based model to study disease states of the intestinal epithelium, ranging from enteric viral infection to mendelian disorders such as CF and autoimmune conditions such as inflammatory bowel disease (IBD), highlighting the potential of organoids as a powerful tool for disease modeling and therapeutics development.

No description available.

Necrosis, originally considered chaotic, has been found to be regulated by distinct molecular pathways. To identify novel regulators of plasma membrane rupture, a hallmark of necrotic cell death, a genome-wide shRNA loss-of-function screen was performed. We identified SNARE complex members, mediators of canonical membrane fusion events, as necessary for Ca2+ and ROS-induced plasma membrane rupture andnecrosis. We targeted N-ethylmaleimide sensitive factor (NSF) due to its requirement in SNARE recycling, lack of gene homologs, and redox sensitivity. Deletion of Nsf protected against membrane rupture induced by various necrotic stimuli, yet did not influence apoptosis, suggesting specificity in programmed necrosis. We discovered that NSF localizes to sites of membrane blebbing and rupture. Additionally, SNARE-binding, ATPase activity, and redox modification are necessary for NSF’s role in necrosis. We generated conditional Nsf knockout mice and induced skeletal muscle injury via cardiotoxin injection. Loss of NSF protected against sarcolemmal rupture and myocyte death. Complementarily, cardiomyocyte-specific deletion of Nsf protected against ischemia-reperfusion injury. The SNARE complex regulator, NSF, is a key mediator of membrane rupture in necrotic cell death and a promising therapeutic target for numerous diseases. / Biomedical Sciences

No description available.

No description available.

In mammalian cells, the formation of a bipolar spindle is an essential process as any mistake in the segregation of chromosomes can result in aneuploidy, an outcome that can be detrimental to the organism. Microtubules are the key structures required for the success of this operation, as they are the main constituents of mammalian bipolar spindles. Centrosomes also play an important role, being the primary source of microtubules. Although centrosomes are dispensable for the formation of bipolar spindles, the fidelity of this process decreases when they are experimentally removed. One way to explain how a bipolar spindle assembles has been the ‘search and capture’ model. According to this model, microtubules emanating from centrosomes search the cytoplasm for kinetochores, which capture microtubules laterally. Once captured, the chromosomes move towards the spindle equator, ultimately resulting in a bipolar spindle. In the research presented here, our aim is to understand the role of microtubule sources other than centrosomes in centrosome-containing mammalian cells. We use different techniques and manipulations to bypass the presence of the centrosomes in order to identify the origin of these sources and their importance. We mainly use LLCPK1, pig kidney epithelial, cells stably expressing GFP tagged proteins, such as alpha-tubulin, to follow microtubules and their associated proteins in live cells. Two main sources of microtubules other than centrosomes have been documented: peripheral microtubules and kinetochores. Peripheral, non-centrosome-associated microtubules were originally thought to disassemble at the beginning of mitosis. However, we found that they form bundles and contribute to the forming spindle. Our model for spindle assembly, which incorporates these peripheral microtubules, and the ‘search and capture’ model are not mutually exclusive; instead, they compliment each other in the formation of bipolar spindles. Lastly, we develop a spindle assembly assay in which centrosomes and kinetochores can be observed separately in mammalian cells. The data documented here demonstrate that kinetochores also contribute to the formation of bipolar spindles by nucleating and organizing microtubules. In addition, TPX2, a microtubule associated protein, is revealed as one of the requirements for microtubule formation and organization at the kinetochores.

The zebrafish protein Mosaic eyes (Moe), and the mammalian orthologue Erythrocyte protein band 4.1-like 5 (Epb4.1l5), are FERM (for Protein 4.1, Ezrin, Radixin, Moesin) domain containing proteins that have important roles during embryonic development. Zebrafish with loss-of-function mutations in moe exhibit defects in retinal lamination, brain ventricle formation, and heart and body morphology. The mammalian epb4.1l5 locus encodes at least two alternately spliced transcripts, the protein products (Epb4.1l5long and Epb4.1l5short) of each contain the FERM domain, but have unique C’ termini. Mice with loss-of-function mutations in epb4.1l5 have early developmental defects in germ layer morphogenesis during the Epithelial to Mesenchymal Transition (EMT), and these mutants fail to properly form a gut and neural tube. Our lab has shown that Moe functions in rod photoreceptors as a negative regulator of outer segment size, and directly binds to members of the Crumbs (Crb) family of proteins, which are apical polarity determinants. Little is known of how Moe, or the Epb4.1l5 isoforms, perform their roles in cell or tissue morphogenesis. The collective aim of my studies was to elucidate the role of Moe and Epb4.1l5 isoforms in development and retinal function. I show that Moe and the apical polarity determinant Crb2a require reciprocal protein function for their respective localization at the eye and brain ventricle surface in zebrafish. I have identified multiple tissues and developmental stages wherein Moe and Crb proteins colocalize, and thus are likely to interact in vivo. I have identified specific sequences shared by Moe and Epb4.1l5 long that are important for function in specific tissues during zebrafish development, and I have investigated morphological, ultrastructural, and behavioral consequences of losing Moe (Epb4.1l5long) function in the patterned zebrafish retina. I describe the interaction of Moe with the Ca2+ binding protein Calmodulin (CaM). I identify three novel binding partners of Epb4.1l5long; Casein Kinase II, Moesin and Radixin, in mammalian retina and retinal pigmented epithelial (RPE) tissue homogenate. And lastly, I integrate the results of my studies to provide a model for Moe function, wherein Moe, together with Moesin and Radixin, link the vesicular transport Crb proteins to the actin cytoskeleton.