Browsing by Subject "Signaling"

The vasculature has a crucial function in normal physiology, enabling the transport of oxygen and nutrients to cells throughout the body. In turn, endothelial cells, which form the inner-most lining of blood vessels, are key regulators of vascular function. In addition to forming a barrier which separates the circulation from underlying tissues, endothelial cells respond to diverse extracellular cues and produce a variety of biologically-active mediators in order to maintain vascular homeostasis. Disruption of normal vascular function is a prominent feature of a variety of pathological conditions. Thus, elucidating the signaling pathways regulating endothelial function is critical for understanding the role of endothelial cells in both normal physiology and pathology, as well as for potential development of therapeutic interventions.

In this dissertation, we use a combination of pharmacological inhibition and knockdown studies, along with generation of endothelial conditional knockout mice, to demonstrate an important role of the Abelson (Abl) family of non-receptor tyrosine kinases (Abl and Arg) in vascular function. Specifically, loss of endothelial expression of the Abl kinases leads to late-stage embryonic and perinatal lethality in conditional knockout mice, indicating a crucial requirement for Abl/Arg kinases in normal vascular development and function. Endothelial Abl/Arg-null embryos display focal regions of vascular loss and tissue damage, as well as increased endothelial cell apoptosis. An important pro-survival function for the Abl kinases is further supported by our finding that either microRNA-mediated Abl/Arg depletion or pharmacological inhibition of the Abl kinases increases endothelial cell susceptibility to stress-induced apoptosis in vitro. The Abl kinases are activated in response to treatment with the pro-angiogenic growth factors vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF). We show that both VEGF- and bFGF-mediated endothelial cell survival is impaired following Abl kinase inhibition.

These studies have uncovered a previously unappreciated role for the Abl kinases in the regulation of the angiopoietin/Tie2 signaling pathway, which functions to support endothelial cell survival and vascular stability. Loss of Abl/Arg expression leads to reduced mRNA and protein levels of the Tie2 receptor, resulting in impaired activation of intracellular signaling pathways by the Tie2 ligand angiopoietin-1 (Angpt1), as well as decreased Angpt1-mediated endothelial cell survival following serum-deprivation stress. Notably, we found that the Abl kinases are activated following Angpt1 stimulation, suggesting a unique dual role for Abl and Arg in Angpt/Tie2 signaling, potentially modulating Tie2 downstream signaling responses, as well as regulating Tie2 receptor expression.

Further, we show an important contribution of the Abl family kinases to the regulation of endothelial permeability responses both in vitro and in vivo. The Abl kinases are activated in response to a diverse group of permeability-inducing factors, including VEGF and the inflammatory mediators thrombin and histamine. We show that inhibition of Abl kinase activity, using either the ATP-competitive inhibitor imatinib or the allosteric inhibitor GNF-2, protects against disruption of endothelial barrier function by the permeability-inducing factors in vitro. VEGF-induced vascular permeability similarly is decreased in conditional knockout mice lacking endothelial Abl expression, as well as following treatment with Abl kinase inhibitors in vivo. Mechanistically, we show that loss of Abl kinase activity is accompanied by activation of the barrier-stabilizing GTPases (guanosine triphosphatases) Rac1 and Rap1, as well as inhibition of agonist-induced Ca2+ mobilization and generation of acto-myosin contractility.

Taken together, these results demonstrate involvement of the Abl family kinases in the regulation of endothelial cell responses to a broad range of pro-angiogenic and permeability-inducing factors, as well as a critical requirement for the endothelial Abl kinases in normal vascular development and function in vivo. These findings have implications for the clinical use of Abl kinase inhibitors.

The bone morphogenetic protein (BMP) signaling pathways have important roles in embryonic development and homeostasis. BMPs have been shown to pattern the dorsoventral axis in zebrafish (Danio rerio) early during embryonic development by establishing a dorsal-to-­ventral ligand gradient. During tumorigenesis, BMPs primarily function as tumor promoters, as an increase in BMP expression is associated with an increase in invasion, migration, epithelial-­to-­mesenchymal transition (EMT), proliferation and angiogenesis.

Although it is clear that BMPs play multiple roles in these biological events, the precise mechanism by which BMPs mediate these functions is not fully understood. Canonically, BMP ligands signal through cell surface receptor complexes that phosphorylate transcription factors, Smad1, Smad5 and Smad8, which mediate BMP-­ specific gene transcription. While studying BMP signaling during cancer progression, we determined that BMPs unexpectedly signal through the canonical TGF-β-­responsive transcription factors, Smad2 and Smad3.

We determined that BMP-­induced Smad2/3 signaling occurs preferentially in embryonic cells and transformed cells. BMPs signal to Smad2/3 by stimulating complex formation between the BMP binding TGF-­β superfamily receptors, ALK3/6, and the Smad2/3 phosphorylating receptors, ALK5/7. BMP signaling through Smad1/5/8 induces MET, while Smad1/5 and Smad2 mediate dorsoventral axis patterning in zebrafish embryos and Smad3 facilitates invasion.

Taken together, our data provides evidence that BMP-­induced Smad2 and Smad3 phosphorylation occurs through a non-­canonical signaling mechanism to mediate multiple biological events. Thus, the signaling mechanisms utilized by BMPs and TGF-­β superfamily receptors are broader than previously appreciated.

G-protein coupled receptors (GPCRs) are pivotal to multiple physiological processes, including neuronal functions such as neurotransmission and memory formation. Consequently, aberrant GPCR signaling has been implicated in complex psychiatric disorders. Our ability to develop efficient therapeutic strategies depends on our understanding of the molecular factors that govern GPCR activity. We performed a genome-wide CRISPR interference screen using a fluorescent cAMP-dependent transcriptional reporter and identified 95 novel regulators that had no previous connection to the GPCR/cAMP pathway. We functionally validated eight regulators and showed that they control distinct steps of the pathway. One of these, RNA binding motif 12 (RBM12), is a novel potent negative regulator of the GPCR/cAMP pathway. Truncating RBM12 mutations are implicated in highly penetrant schizophrenia and defective brain development, yet its cellular function is unknown. To investigate its role in GPCR/cAMP signaling, we performed CRISPR-based gene editing in two complementary cellular models: experimentally tractable HEK293 cells and physiologically relevant human induced pluripotent stem cell-derived neurons. We found that RBM12 deficiency leads to hyperactive cAMP accumulation, PKA activity, and downstream CREB-dependent transcriptional responses in response to the activation of stimulatory GPCRs with key functions in neurobiology (beta-adrenergic, dopamine, and adenosine). We observed that the cAMP signaling and PKA activity are independently subject to regulation by RBM12. This signaling hyperactivity is conserved in the two cellular model systems. We further show that schizophrenia-associated truncating RBM12 mutations (c.2377G>T and c.2532delT) failed to rescue the signaling hyperactivation due to possible loss-of-function and protein stability defect. In agreement with its activity as an RNA-binding protein, we show that RBM12 depletion led to altered expression of genes with established roles in the signaling pathway, including an upregulation of cyclases and the protein kinase A catalytic subunits as well as a downregulation of phosphodiesterases. In addition, we observed aberrant gene expression in neurogenesis, neurodevelopment, and differentiation pathways. Together, these experiments provide critical insights into this uncharacterized gene and its function as a novel regulator of GPCR signaling. By uncovering the novel regulatory role of a schizophrenia-risk gene in GPCR signaling, we expand our understanding of the molecular basis of neuropsychiatric disorders and enable the identification of novel druggable targets.

Diacylglycerol (DAG), a lipid messenger generated upon T cell receptor (TCR) engagement, mediates signaling through the IKK/NF-κB and Ras/ERK pathways. Further downstream of the Ras/ERK pathway are mammalian target of rapamycin (mTOR) and MAP kinase signal integrating kinases Mnk1 and Mnk2. While mTOR acts as a critical regulator of T cell metabolism, homeostasis and function, Mnk1 and Mnk2 phosphorylate the initiation factor eIF4E that plays an important role in cap-dependent mRNA translation. Diacylglycerol kinases (DGKs) terminate DAG-mediated signals by phosphorylating DAG into phosphatidic acid. T cells that lack both α and ζ isoforms of DGK accumulate excess DAG upon activation, resulting in hyper-activation of the IKK/NF-κB, Ras/ERK and mTOR pathways, hypersensitivity to TCR stimulation, and loss of self-tolerance. Here, we have examined the mechanisms by which dysregulated DAG-mediated signaling affects T cell function. To this end, we studied the effects of hyper-activating individual DAG-mediated pathways (IKK/NF-κB and TSC/mTOR) on T cell function. We also examined the role of ERK-activated kinases Mnk1 and Mnk2 in T cell function.

Using mice with T cell-specific expression of a constitutively active form of IKKβ (`IKK' mice), we found that uncontrolled IKKβ/NF-κB signaling promotes T cell apoptosis and attenuates responsiveness to TCR stimulation. Defective IL-2 production and increased FasL expression contributed to enhanced IKK T cell apoptosis. Impaired IKK T cell activation and proliferation were associated with defects in TCR signaling, and upregulation of the cell surface inhibitory receptor PD1. In vivo, IKK T cells mounted a compromised antigen-specific CD8 T cell response with curtailed expansion and exaggerated contraction phases. Notably, expression of transcriptional repressor Blimp1 (a regulator of T cell exhaustion) was increased in IKK T cells, and conditionally deleting Blimp1 was able to largely restore responsiveness to TCR stimulation.

Investigating Mnk1/2 double knockout (DKO) mice, we found that Mnk1 and Mnk2 are dispensable for T cell development and function, but important for the pathogenesis of experimental autoimmune encephalomyelitis (EAE). TCR engagement activated Mnk1/2 in a Ras/ERK-dependent manner in primary T cells, and was inhibited by DGK α and ζ. Mnk1/2 deficiency did not affect the development of conventional αβ T cells, regulatory T cells, or invariant NKT cells. Mature T cells from DKO mice showed normal activation and CD4 TH differentiation ex vivo, but DKO mice developed lower clinical scores than WT counterparts in an EAE model, correlating with a smaller pool of MOG-reactive IL-17-producing and IFNγ-producing CD4 cells. These results suggest that Mnk1/2 may play a minimal role in T cell development and function but may control non-T cell lineages to regulate TH1 and TH17 differentiation in vivo.

To determine the effect of constitutive mTOR complex 1 activity on anti-bacterial CD8 responses, we investigated mice with T cell-specific deletion of TSC1, a suppressor of mTOR complex 1 activity. Using an established model system of transgenic (OT1) CD8 cell adoptive transfer and challenge with Listeria monocytogenes expressing a cognate antigen, we found that TSC1 deficiency impairs antigen-specific CD8 responses. Fewer TSC1-deficient OT1 cells were present in the peripheral blood and spleen at the peak of the response and fewer memory cells were found at later time points, in individual and competitive adoptive transfer experiments with WT counterparts. Weak expansion of TSC1-deficient cells was correlated with defects in survival and proliferation in vivo, while exaggerated contraction was associated with an increased ratio of SLECs to MPECs in the effector cell population. This perturbation in effector-memory differentiation was concomitant with enhanced T-bet expression and decreased Eomes expression among activated TSC1 KO cells. Upon competitive adoptive transfer with WT counterparts and antigen re-challenge, TSC1-deficient memory cells showed moderate defects in expansion but not cytokine production. Taken together, these findings provide direct evidence of a CD8 cell-intrinsic role for TSC1 in regulating antigen-specific primary and memory responses.

In sum, findings from these studies provide deeper insight into the regulation of T cell function by DAG-mediated pathways, and may have implications for the design of immune-modulation strategies during vaccination, autoimmunity and cancer immunotherapy.

Enterotoxigenic Escherichia coli (ETEC), the leading cause of traveler's diarrhea and childhood mortality due to diarrhea in the developing world, has been shown to secrete heat-labile enterotoxin (LT) in association with outer membrane vesicles. However, studies on the effect of LT have been performed using soluble LT, which is not its physiologically relevant presentation context. The effect of LT associated with vesicles and its trafficking within human intestinal epithelial cells were compared with soluble LT. Cytokine responses and trafficking of standardized samples of soluble LT and vesicle-associated LT were evaluated in polarized intestinal epithelial cells. Using real-time PCR, immunoblotting, and ELISAs, we found that compared to soluble LT, vesicle-bound LT showed delayed kinetics in the activation of LT. Vesicles containing LT or not also produced cytokines through different signaling pathways than soluble LT. We found that this difference in signaling was due to different trafficking within the cell. Interestingly, not all LT associated with vesicles is active within cells. Vesicle-associated LT must bind to the host receptor GM1 in lipid rafts to be active within cells. This suggests that although vesicles can deliver large amounts of LT to a cell, much of the LT would be inactive and not produce a physiological response. To test this hypothesis, we attempted to develop animal models for ETEC-induced diarrhea. Although the models were largely unsuccessful, the mouse model appears promising for determining the physiological response of a host to LT as fluid accumulation was observed in response to vesicles containing LT. The results in this thesis provide further understanding of the mechanism of LT-induced diarrhea and emphasize the importance of study toxins in their natural context.

β-arrestins (βarrs) interact with G protein-coupled receptors (GPCRs) to desensitize G protein signaling, initiate signaling on their own, and mediate receptor endocytosis.  Using a panel of GPCRs believed to couple differently to βarrs we demonstrate how distinct conformations of GPCR–βarr complexes are specialized to perform different subsets of these cellular functions. Our results thus provide a new signaling paradigm for the understanding of GPCRs, whereby a specific GPCR–βarr conformation mediates receptor desensitization, while another drives internalization and some forms of signaling.

In addition, some GPCRs activate G proteins from within internalized cellular compartments resulting in sustained signaling. We used a variety of biochemical, biophysical, and cell-based methods to demonstrate the existence, functionality, and architecture of internalized receptor “mega-complexes” composed of a single GPCR, βarr, and G protein. EM of purified ‘megaplexes’ reveals that a single receptor binds simultaneously through its core region with G protein and with βarr in the tail conformation. Thus, the two GPCR–βarr conformations carry out distinct cellular functions.

Animals must coordinate development with fluctuating nutrient availability. Nutrient availability governs post-embryonic development in Caenorhabditis elegans: larvae that hatch in the absence of food do not initiate post-embryonic development but enter "L1 arrest" (or "L1 diapause") and can survive starvation for weeks, while rapidly resume normal development once get fed. Insulin-like signaling (IIS) has been shown to be a key regulator of L1 arrest and recovery. However, the C. elegans genome encodes 40 insulin-like peptides (ILPs), and it is unknown which peptides participate in nutritional control of L1 arrest and recovery. Work in other contexts has identified putative receptor agonists and antagonists, but the extent of specificity versus redundancy is unclear beyond this distinction.

We measured mRNA expression dynamics with high temporal resolution for all 40 insulin-like genes during entry into and recovery from L1 arrest. Nutrient availability influences expression of the majority of insulin-like genes, with variable dynamics suggesting complex regulation. We identified 13 candidate agonists and 8 candidate antagonists based on expression in response to nutrient availability. We selected ten candidate agonists (daf-28, ins-3, ins-4, ins-5, ins-6, ins-7, ins-9, ins-26, ins-33 and ins-35) for further characterization in L1 stage larvae. We used destabilized reporter genes to determine spatial expression patterns. Expression of candidate agonists was largely overlapping in L1 stage larvae, suggesting a role of the intestine, chemosensory neurons ASI and ASJ, and the interneuron PVT in systemic control of L1 development. Transcriptional regulation of candidate agonists was most significant in the intestine, as if nutrient uptake was a more important influence on transcription than sensory perception. Scanning in the 5' upstream promoter region of these 40 ILPs, We found that transcription factor PQM-1 and GATA putative binding sites are depleted in the promoter region of antagonists. A novel motif was also found to be over-represented in ILPs.

Phenotypic analysis of single and compound deletion mutants did not reveal effects on L1 recovery/developmental dynamics, though simultaneous disruption of ins-4 and daf-28 extended survival of L1 arrest without enhancing thermal tolerance, while overexpression of ins-4, ins-6 or daf-28 shortened L1 survival. Simultaneous disruption of several ILPs showed a temperature independent, transient dauer phenotype. These results revealed the relative redundancy and specificity among agonistic ILPs.

TGF- β and steroid hormone (SH) signaling have been reported to control the dauer formation along with IIS. Our preliminary results suggest they may also mediate the IIS control of L1 arrest and recovery, as the expression of several key components of TGF-β and SH signaling pathway genes are negatively regulated by DAF-16, and loss-of-function of these genes partially represses daf-16 null phenotype in L1 arrest, and causes a retardation in L1 development.

In summary, my dissertation study focused on the IIS, characterized the dynamics and sites of ILPs expression in response to nutrient availability, revealed the function of specific agonistic ILPs in L1 arrest, and suggested potential cross-regulation among IIS, TGF-β signaling and SH signaling in controlling L1 arrest and recovery. These findings provide insights into how post-embryonic development is governed by insulin-like signaling and nutrient availability.

In my first chapter, I present a model in which sellers can signal the quality of an asset both by retaining a fraction of the asset and by choosing the liquidity of the market in which they search for buyers. Although these signals may seem interchangeable, I present two settings which show they are not. In the first setting, sellers have private information regarding only asset quality, and I show that liquidity dominates retention as a signal in equilibrium. In the second setting, both asset quality and seller impatience are privately known, and I show that both retention and liquidity operate simultaneously to fully separate the two dimensions of private information. Contrary to received theory, the fully separating equilibrium of the second setting may contain regions where market liquidity is increasing in asset quality. Finally, I show that if sellers design an asset-backed security before receiving private information regarding its quality, then the optimality of standard debt is robust to the paper's various settings.

In my second chapter, I explore the question of how informative bank stress tests should be. I use Bayesian persuasion to formalize stress tests and show that regulators can reduce the likelihood of a bank run by performing tests which are only partially informative. Optimal stress tests give just enough failing grades to keep passing grades credible enough to avoid runs. The worse the state of the banking system, the more stringent stress tests must be to prevent runs. I find that optimal stress tests, by reducing the probability of runs, reduce the optimal level of banks' capital cushions. I also examine the impact of anticipated stress tests on banks' ex ante incentive to invest in risky versus safe assets.

Fungi have a genetically controlled sex determination system, which is governed by a small, sex-specific region in the genome called the mating-type locus (MAT). In the basidiomycetous yeast Cryptococcus neoformans, the pathogen that causes cryptococcal meningitis and cryptococcosis, sex has been associated with virulence. To further understand how sex is genetically regulated in C. neoformans, we focused our studies on the evolution of the MAT locus and molecular dissection of the pheromone signaling pathway that controls sexual development. Two MAT-linked meiotic recombination hotspots that likely drove the assembly and rearrangement of MAT were identified. Fine mapping through the integration of genetic markers established that two hotspots, one on each side of the MAT locus, are located in an ~10 kb and ~5 kb region. Plotting the G + C content along MAT and the flanking regions revealed a strong association between the location of these two hotspots and a high G + C content. By deletion and insertion of the G + C rich region, we demonstrated that the high G + C rich region is required but not sufficient to induce recombination. On the other hand, to provide direct experimental evidence to support the previously proposed model for the evolution of MAT, we sought to recapitulate the ancestral tetrapolar, and the intermediate tripolar mating systems of C. neoformans by manipulating the MAT structure to model a tetrapolar system. In the two modified "a" and "α" strains, the sex-determining genes SXI1α or SXI2a residing at the MAT locus were disrupted and the wild-type allele of these two genes was then reintroduced at another genomic location (URA5) that is unlinked to MAT. Our results show that C. neoformans can complete the sexual cycle with a tetrapolar mating configuration and the transitional tripolar state might be under strong negative selection pressure, which could have facilitated the transition from a tripolar state to the final bipolar mating system.

The MAT locus is the major determinant of the sexual identity of a cell, but several signaling pathways, including the pheromone signaling pathway, are required to regulate mating and sexual development. Many components of the pheromone signaling pathway have been identified; however, it is less clear what lies upstream of the MAPK cascade. To address this question, we studied the role of two Gα subunits (Gpa2, Gpa3) in mating and concluded that they share both redundant and divergent roles in mating. gpa2 gpa3 double mutants, but neither gpa2 nor gpa3 single mutants, are sterile in bilateral crosses. In their GTP-bound form, they signal in opposition: Gpa2 promotes mating whereas Gpa3 inhibits. Furthermore, we also studied the functions of a novel upstream component Cpr2, a pheromone receptor-like gene, in pheromone signaling and sexual development. All lines of evidence suggest that Cpr2 is a constitutive ligand-independent receptor that, when expressed, engages the same G-proteins and activates the same pheromone signaling pathway as the canonical ligand-activated pheromone receptors. Expression of Cpr2 is induced post cell fusion during mating, and likely introduces a positive feedback loop to allow a self-perpetuating signaling state to enable efficient mating. Cells lacking this receptor are fertile, but produce abnormal filamentous structures. Overexpression of CPR2 in a or α cells strongly enhances fruiting, an alternative same-sex mating process in C. neoformans. Therefore, Cpr2 establishes a new paradigm for a naturally occurring constitutively active GPCR that governs cell fate in fungi.

Finally, we described a sex-induced silencing (SIS) phenomenon in C. neoformans. Using genetic approaches, we showed that SIS is triggered by a tandem insertion of a transgene during the sexual cycle. Interestingly, only a proportion of progeny carrying the transgene are silenced. Gene deletion, RIP, or DNA methylation do not contribute to SIS but the RNAi machinery is required. In conclusion, these studies provide further understanding of sex in C. neoformans from different perspectives, which invites comparisons to other fungal and even more broadly, eukaryotic pathogens to address the role of sex in evolution.

Sea urchin larvae possess a beautiful, intricately patterned, calcium-carbonate skeleton. Formation of this complex structure results from two independent processes within the developing embryo: specification of the mesenchymal cells that produce the skeletal rods, and patterning inputs from the ectoderm, which secretes signals directing the growth and shape of the skeleton. To understand patterning of the skeleton therefore, the specification events behind these two processes must be understood separately, and then connected in order to understand how ectoderm signaling directs skeletal growth. While the former processes of mesenchyme specification and mineralization are under study elsewhere, the means by which ectodermal cues directing skeletal growth are activated and localized is not known. Using molecular genetics, including gene knock downs and mis-expression, as well as microsurgical manipulations of early cleavage embryos, I show how a previously undescribed territory within the ectoderm, the border ectoderm (BE) is specified with short range signaling inputs. Then, experiments show that the BE provides signals that initiate, and contribute to the propagation of skeletogenesis. From this dataset, and from biological experiments I outline a model for how the BE patterns and contributes to the directed growth of the skeleton. I also discuss challenges to this model that need to be addressed in future research. In principle, the mechanism proposed herein depends on the integration of information from both the primary and secondary embryonic axes. It requires both short-range signaling by Wnt5 from the endoderm to establish the BE fate, and TGFß signaling from the oral and aboral ectoderm which subdivides the BE into four territories. These findings demonstrate that the short-range signaling cascade that subdivides the embryo into first mesoderm and then endoderm also specifies ectodermal fates. Ultimately, this research paves the way for understanding how the larval skeleton is patterned during embryogenesis and may provide a paradigm for understanding other, more complex, developmental problems.

Inositol Polyphosphate Multi-Kinase (IPMK, also known as ArgRIII, Arg82, and IPK2) is a central component of the inositol signaling system, catalyzing the phosphorylation of at least four different inositol polyphosphate species in vivo with in vitro activity observed for three more. Each of these IP species is sterically unique and the phosphorylation target varies between the 6'-, 3'-, or 5'-hydroxyls, classifying IPMK as a 6/3/5-kinase. The products of IPMK have been linked to multiple processes including cell cycle regulation, transcriptional control, telomere length regulation, mRNA export and various phenotypes including mouse embryonic and fly larvae development, and stress responses in plants and yeast. Linking specific IP species and cellular processes has been complicated by the inability to distinguish between the different effects of the various IP species generated by IPMK. Deletion of IPMK affects the IP populations of all its various substrates and products and therefore the role of a single IP species cannot be tracked. The goals of this work were to address the question of substrate selectivity and develop new tools to probe inositol signaling in vivo through a combination of structural, enzymatic, and genomic techniques.

The structure of Arabidopsis thaliana IPMK is reported at 2.9Å resolution and in conjunction with a new model of inositol selectivity has been used to design constructs with altered substrate profiles. In vitro and in vivo experiments have confirmed that IPMK identifies substrate inositol polyphosphate species through a recognition surface that requires phosphate groups occupy specific pockets and rejects those with axial phosphate groups in specific regions. In vivo experiments have linked specific inositol polyphosphate species to nitrogen metabolism and temperature sensitivity in yeast and established the potential for these constructs to be used to probe signaling in other organisms.

Neuroblastoma is a pediatric cancer arising from undifferentiated neural crest-derived precursor cells. Treatment strategies for neuroblastoma aim to promote neuroblast differentiation, however current therapies available are only modestly effective. The tumor stroma contributes to the suppression of tumor growth by releasing soluble factors that act to promote neuroblast differentiation, though the precise factors released and their mechanism of action in neuroblastoma remains unclear. Here, we identify a novel component of the differentiating stroma secretome and harness stroma biology to inform the use of a combination therapy for neuroblastoma treatment.

HBEGF expression is decreased in neuroblastoma compared to benign disease, correlating to an increase in mortality. HBEGF protein is expressed only in stromal compartments of tumor specimens, with tissue from late-stage disease containing very little stroma or HBEGF. Addition of soluble HBEGF to neuroblastoma cell lines leads to increased neuroblast differentiation and decreased proliferation. Heparan sulfate proteoglycans (HSPGs) and heparin derivatives further enhance HBEGF-induced differentiation by forming a complex with the epidermal growth factor receptor (EGFR), leading to activation of the ERK1/2 and STAT3 pathways and upregulation of the inhibitor of DNA binding 1 transcription factor.

Expression of the type III TGF-β receptor (TβRIII), an HSPG, is epigenetically regulated in neuroblastoma cells via direct binding of the N-Myc transcription factor to Sp-1 sites on the TβRIII promoter. Analysis of patient microarray data demonstrate that other members of the differentiating stroma secretome, including HBEGF and EGFR, are positively correlated with TβRIII expression, suggesting that these proteins may be co-regulated. Treatment with inhibitors aimed at blocking N-Myc function, including inhibitors of histone deacetylases, DNA methyltransferases (DNMTs), and aurora kinase A (AurkA) can promote neuroblast differentiation and decrease proliferation. The combination of the DNMT inhibitor decitabine with the AurkA inhibitor MLN8237 enhances differentiation and reduces proliferation compared to either agent alone. Importantly, the combination of clinically achievable doses of these targeted agents dramatically reduces tumor growth in orthotopic xenograft models of neuroblastoma, identifying a novel combination therapy that may benefit children with this disease.

In conclusion, these studies dissect the tumor microenvironment to identify an important member of the differentiating stroma secretome, while also revealing a combination therapy for clinical development that has the potential to decrease adverse side effects and increase effectiveness of neuroblastoma treatment.

The endothelial receptor tyrosine kinase (RTK) Tie1 was discovered over 20 years ago, yet its precise function and mode of action remain enigmatic. To shed light on Tie1’s role in endothelial cell biology, we investigated a potential threonine phosphorylation site within the juxtamembrane domain of Tie1. Expression of a non-phosphorylatable mutant of this site (T794A) in zebrafish (Danio rerio) significantly disrupted vascular development, resulting in fish with stunted and poorly branched intersomitic vessels. Similarly, T794A-expressing human umbilical vein endothelial cells formed significantly shorter tubes with fewer branches in three-dimensional Matrigel cultures. However, mutation of T794 did not alter Tie1 or Tie2 tyrosine phosphorylation or downstream signaling in any detectable way, suggesting that T794 phosphorylation may regulate a Tie1 function independent of its activity as a kinase. Although T794 is within a consensus Akt phosphorylation site, we were unable to identify a physiological activator of Akt that could induce T794 phosphorylation, suggesting that Akt is not the physiological Tie1-T794 kinase. However, the small GTPase Ras-related C3 botulinum toxin substrate 1 (Rac1), which is required for angiogenesis and capillary morphogenesis, was found to associate with phospho-T794 but not the non-phosphorylatable T794A mutant. Pharmacological activation of Rac1 induced downstream activation of p21-activated kinase (PAK1) and T794 phosphorylation in vitro, and inhibition of PAK1 abrogated T794 phosphorylation. Our results provide the first demonstration of a signaling pathway mediated by Tie1 in endothelial cells, and they suggest that a novel feedback loop involving Rac1/PAK1-mediated phosphorylation of Tie1 on T794 is required for proper angiogenesis.

Cleaner shrimp are small, brightly-colored tropical crustaceans that attract reef fish clients to set locations on a reef called cleaning stations and then "clean" them by removing ectoparasites from their scales, gills, and mouths. Because clients benefit from the removal of potentially harmful parasites, and shrimp benefit from a meal, this interaction is considered mutualistic. The evolution of cleaner-client relationships is especially paradoxical, however, given that crustaceans make an easy meal for many reef fish. How did such an interaction arise and evolve? Additionally, how do cleaner shrimp and clients find and recognize each other, and why doesn't the client eat the cleaner? One hypothesis is that mutualistic partners evolve signals that identify them as beneficial partners. For the cleaner shrimp-client fish mutualism, it has been suggested that these signals are visual, and function to identify cleaners as helpers, rather than food, and perhaps to identify clients as seeking cleaning rather than a meal.

The broad goal of this dissertation was to examine the cleaner shrimp-client fish mutualism by combining visual physiology, visual ecology, and animal behavior. Throughout, I focus on visual acuity-the ability to perceive detail-an underexplored aspect of visual capability which should represent an important selective force on signals, as it determines what visual information can and cannot be resolved. To begin, in Chapter 2, I characterized the visual capabilities of cleaner shrimp for the first time, in particular examining spectral sensitivity, visual acuity, and temporal resolution of three species from the three primary genera where cleaning has arisen. This revealed that these cleaner shrimp have monochromatic, coarse vision, and thus that their color patterns likely do not serve an intraspecific signaling purpose, but rather may be part of signals directed at client fish. Thus, Chapter 3 examined visual acuity in fish by performing a literature synthesis of known visual acuity values across fishes and examining how acuity relates to certain aspects of morphology and ecology. Here, I found that acuity is higher in fish with larger eyes and in fish that live in spatially complex habitats.

Having examined the visual capabilities of both parties in the cleaner-client mutualism, I then explored signaling between cleaners and clients. In chapter (4), I used sequential analysis to demonstrate that certain stereotyped motions by cleaner shrimp are signals directed at clients, and provide the first evidence of potential signals on the part of client fish as well. Additionally, I developed an R package, AcuityView (Appendix A) which displays images with only the spatial information available to a receiver of given acuity from a given distance, and used it to show that cleaner and client signals are visible to their intended receivers. Lastly, in Chapter (5), I outlined what we know about acuity across species, provide primers on the anatomical basis for acuity and what factors can make acuity higher, and put forward specific predictions and hypotheses about how receiver acuity may influence signal form.

A number of puzzles exist regarding the role of domestic politics in the likelihood of international conflict. In particular, the sources of incomplete information remain under-theorized and the microfoundations deficient. This study will examine the role that the unity of the government and the views of the government towards the use of force play in the targeting of states. The theory presented argues that divided dovish governments are particularly likely to suffer from military challenges. In particular, divided governments have difficulty signaling their intentions, taking decisive action, and may appear weak. The theory will be tested on a new dataset created by the author that examines the theory in the context of international territorial disputes. A number of significant findings emerge from the data. First, divided governments are significantly more likely to face challenges and increasing government unity markedly decreases the likelihood of a challenge. Second, unified hawkish governments were far less likely to be the targets of challenges than divided dovish governments. Finally, the substantive impact of key variables was much greater than that of the control variables. The causal process postulated in the theory is examined in eight case studies, two for each of the four government types. These case studies show that government type not only impacts the signaling of democratic states but also influences the perceptions of potential challengers.