Stroke or cerebral vascular accident (CVA) is the 4th leading cause of mortality and the principle cause of long-term disability in the United States. Unfortunately, current reperfusion-based treatments (e.g. thrombolysis, tPA) cannot be administered to the majority of patients presenting with ischemic stroke. Accordingly, new treatments for ischemic stroke are desperately needed. Reactive astrocytes perform key roles in tissue repair and remodeling following stroke such as preservation and repair of the blood-brain barrier, modulation of immune cell invasion, glutamate uptake and neuroprotection, and glial scar formation. The proliferative subpopulation of reactive astrocytes found immediately adjacent to the infarct core after stroke (known as the peri-infarct area) is particularly important for protecting the brain parenchyma from ischemic damage and inflammation. Defining the signaling network that controls reactive astrocyte formation and function has potential to provide new treatment strategies for patients ineligible for reperfusion therapy. Notch1 signaling is required for the proliferation of peri-infarct reactive astrocytes after stroke. To identify downstream targets and potential functional effectors of Notch1 signaling in reactive astrocytes, we developed an ex vivo forward signaling screen. To generate large quantities of adult reactive astrocytes, we employed adult Reactive astrocyte-derived Neural Stem Cells (Rad-NSCs) isolated from the peri-infarct area of mice after stroke. Astrocytes re-differentiated from Rad-NSCs (AstroRad-NSC) were then exposed to immobilized Jagged-1, a Notch1 ligand. In response to Jagged-1, many genes involved in reactive astrocyte-mediated tissue protection, metabolic regulation, angiogenesis and glial scar formation were up-regulated. Of special interest, several genes for proteins that regulate with glutamate uptake and metabolism were increased by Jagged-1/Notch signaling, including the glial-specific GLutamate-ASpartate Transporter (GLAST). With loss-of-function experiments, we determined that deletion of Notch1 decreased GLAST transcript and protein levels in cultured AstroRad-NSC. Furthermore, we isolated reactive astrocytes directly from cerebral cortex after stroke and confirmed the effects of Notch1 on GLAST in vivo. Our results suggest that treatments designed to stimulate Notch1 signaling after stroke may promote glutamate uptake, thereby decreasing excitotoxicity and neuronal cell death. Binding of Endothelin peptides to the type B Endothelin receptor (ETBR) has been shown to alter cell proliferation. Investigating a possible relationship between Jagged-1/Notch1 and Endothelin signaling in reactive astrocytes, we determined that Notch1 signaling regulated ETBR indirectly, by activating STAT3, an unidentified transcriptional activator of ETBR. Using inducible transgenic astrocyte-specific conditional knockout (cKO) mice (GFAP-ETBR-cKO), we found that specific deletion of ETBR in reactive astrocytes phenocopied the defect in reactive astrocyte proliferation observed in our previous work with GFAP-Notch1-cKO mice. Notably, the Notch1-STAT3-ETBR axis we identified is likely to control reactive astrocyte proliferation in most, if not all, forms of CNS injury. The experimental results presented in this doctoral dissertation provide novel insight into signaling mechanisms that may someday be exploited to improve care for patients with stroke and other forms of CNS injury or disease.

Proteolytic processing of transmembrane receptors and ligands can have dramatic effects on cell signaling and subsequent cellular responses. Previous studies demonstrated that a disintegrin and metalloproteinase 10 (ADAM10) may cleave numerous B cell-expressed receptors, including the low affinity IgE receptor (CD23). However, lethality of ADAM10-deficient embryos has limited examination of these cleavage events in lymphocytes. To investigate their role in B cell development and function, we generated B cell-specific ADAM10 knockout mice. Intriguingly, deletion prevented development of the entire marginal zone B cell (MZB) lineage. Further analysis revealed that ADAM10 is required for S2 cleavage of the Notch2 receptor and initiation of Notch2 signaling, which is required for MZB development. Additionally, cleavage of CD23 was dramatically impaired in ADAM10-deficient B cells. This finding and results of ex vivo cleavage assays demonstrated that ADAM10 is the principal in vivo sheddase of CD23. Previous studies have demonstrated that Notch signaling and CD23 cleavage regulate antibody production. Accordingly, deletion of ADAM10 profoundly inhibited germinal center formation, and T-dependent and T-independent antibody responses to immunization, implicating ADAM10 as a novel regulator of adaptive immunity. Additionally, to determine the role of ADAM10 activity in hematopoiesis, we generated transgenic mice (A10Tg) that overexpress the protease on lymphoid and myeloid progenitors. Surprisingly, this markedly suppressed B2 cell development and promoted dramatic expansion of myeloid-derived suppressor cells (MDSCs) via a cell intrinsic mechanism. A10Tg MDSCs inhibited T cell proliferation and adoptive immunotherapy of B16 melanoma, resulting in exacerbated metastatic progression that was prevented by MDSC depletion. Thus, A10Tg mice represent a novel model for the examination of MDSC development and MDSC-mediated immune suppression in a tumor-free environment. Finally, hematopoietic stem cell cultures revealed that ADAM10 overexpression directs myeloid development by dysregulating Notch signaling via uncoupling the highly regulated proteolysis of Notch receptors. Collectively, these findings demonstrate that ADAM10 is a critical regulator of Notch signaling, B cell development, and MDSC expansion. Moreover, they have important implications for the treatment of numerous CD23 and Notch mediated pathologies, ranging from allergy to cancer.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

Many pathways dysregulated in prostate cancer are also involved in epithelial differentiation. To better understand prostate tumor initiation, we sought to investigate specific genes and mechanisms required for normal basal to luminal cell differentiation. Utilizing human prostate basal epithelial cells and an in vitro differentiation model, we tested the hypothesis that regulation of NOTCH3 by the p38 MAPK family (hereafter p38-MAPK), via MYC, is required for luminal differentiation. Inhibition (SB202190 and BIRB796) or knockdown of p38a (also known as MAPK14) and/or p38d (also known as MAPK13) prevented proper differentiation. Additionally, treatment with a gamma-secretase inhibitor (RO4929097) or knockdown of NOTCH1 and/or NOTCH3 greatly impaired differentiation and caused luminal cell death. Constitutive p38-MAPK activation through MKK6(CA) increased NOTCH3 (but not NOTCH1) mRNA and protein levels, which was diminished upon MYC inhibition (10058-F4 and JQ1) or knockdown. Furthermore, we validated two NOTCH3 enhancer elements through a combination of enhancer (e) RNA detection (BruUV-seq) and luciferase reporter assays. Finally, we found that the NOTCH3 mRNA half-life increased during differentiation or upon acute p38-MAPK activation. These results reveal a new connection between p38-MAPK, MYC and NOTCH signaling, demonstrate two mechanisms of NOTCH3 regulation and provide evidence for NOTCH3 involvement in prostate luminal cell differentiation.

Der Notch Signalweg spielt während der Embryonalentwicklung eine zentrale Rolle in der Spezifizierung des Zellschicksales, der Proliferation und der Kommunikation benachbarter Zellen. Die Hey bHLH Transkriptionsfaktoren sind Zielgene des Notch-Signalweges und besitzen wichtige Funktionen in der kardiovaskulären Entwicklung. Hey2 Knockout (KO) Mäuse und Hey1/HeyL Doppelknockout-Mäuse (DKO) sind gekennzeichnet durch eine fehlerhafte Ausbildung der Herzscheidewand und der Herzklappen und durch eine unzureichende Differenzierung während der Blutgefäßentwicklung. Ziel dieser Arbeit war es, neue Zielgene der Hey Proteine zu finden, um ihre Funktion in der Organentwicklung und die Ausprägung der Hey KO Maus-Phänotypen besser verstehen zu können. Dazu wurde als Methode eine Kombination aus Microarray-Analyse und Chromatinimmunpräzipitation (ChIP) gewählt, um gleichzeitig einen Überblick über die regulierten Zielgene und der direkt gebundenen Promotoren zu gewinnen. Als Zellkulturmodell wurden HEK293-Zellen genutzt, die doxyzyklin-induzierbar Flag-markiertes Hey1, bzw. Hey2 Protein überexprimieren. Eine Microarray-Analyse nach Überexpression von Hey1, bzw. Hey2 ergab insgesamt ca. 100 bis zu 5-fach herunterregulierte Zielgene und nur für Hey2 15 Gene, die stärker als 2-fach hochreguliert waren. Eine ChIP mit αFlag-Antikörper zeigte eine direkte DNA-Bindung von Hey1, bzw. Hey2, im proximalen Promotorbereich von 4 herunterregulierten Zielgenen (HEY1, BMP2, KLF10 und FOXC1). Ist jedoch die DNA-bindende basische Domäne des Hey1-Proteins deletiert, bzw. durch Aminosäureaustausche (3 Arginine zu 3 Lysine) vermutlich nicht mehr DNA-bindend, kann eine Herunterregulation der Zielgene nach Überexpression der Hey1-Mutanten nicht mehr festgestellt werden. Ebenso kann eine Bindung der Hey1-Mutanten an die ausgewählten Promotoren von HEY1, BMP2, KLF10 oder FOXC1 mit ChIP nicht mehr nachgewiesen werden. Dies deutet darauf hin, dass die basische Domäne essentiell für die DNA-Bindung und für die Funktion der Hey Proteine ist. Mit ChIP-PET und anschließender Hochdurchsatz-Sequenzierung wurde ein genomweiter Screen der Hey1- und der Hey2-Bindungsstellen in HEK293-Zellen durchgeführt. Für Hey1 wurden 1453 Zielgene, für Hey2 4288 Zielgene bestimmt, wobei 1147 Gene gemeinsame Zielgene von Hey1 und Hey2 waren. Obwohl die Bindungsstellen in 5'- und 3'-Richtung von kodierenden Sequenzen und auch in Exons und Introns lokalisiert waren, waren 55 %, bzw. 49 % aller Bindungsstellen für Hey1, bzw. Hey2 im proximalen Promotorbereich von -0,5 kb und im ersten Exon lokalisiert. Eine in silico Analyse des Bindemotivs deutete auf eine repetitive GC-haltige Sequenz hin, die vermutlich in CpG Inseln lokalisiert ist. Diese Ergebnisse weisen auf eine direkte Regulation der Transkriptionsmaschinerie durch die Hey Proteine hin. Ein Vergleich der Zielgene aus den Microarray-Analysen mit den ChIP-PET Daten zeigte einen hohen Anteil an herunterregulierten Genen mit Bindestellen, die direkt von Hey gebunden waren. Während 60 % der herunterregulierten Hey2 Zielgene in der ChIP-PET Analyse eine direkte DNA-Bindung zeigen, weisen nur 20 % der hochregulierten Gene Bindestellen für Hey2 auf. Dies spricht für eine überwiegende Repressorfunktion der Hey Proteine. Um zu überprüfen, inwieweit die Hey Proteine zelltypspezifisch verschiedene Zielgene regulieren, wurden embryonale Stammzellen (ES-Zellen) generiert, die ebenfalls doxyzyklin-induzierbar Hey1, bzw. Hey2 überexprimieren. Diese ES-Zellen konnten effektiv zu Kardiomyozyten differenziert werden, so dass auch in diesen Zellen eine Hey Überexpression induziert und somit eine Genexpressionsanalyse durchgeführt werden konnte. Microarray Analysen der ES-Zellen und Kardiomyozyten ergaben mehr hoch- als herunterregulierte Gene im Vergleich zu HEK293-Zellen. Die Überlappung an gemeinsam regulierten Zielgenen in HEK293, ES-Zellen und Kardiomyozyten war sehr gering. Nur zwei Hey2-Zielgene wurden gleichzeitig in HEK293 und ES-Zellen stärker als 2-fach reguliert (Hes1, Zic2). Diese geringe Überlappung deutet auf ein enges zelltypspezifische Regulationspotential hin. Eine Genontologie-Analyse aller Zielgene zeigte Interaktionen der Hey Proteine mit verschiedenen Signalwegen (z.B. TGFβ-, Id- oder Wnt-Signalweg), die alle unersetzlich in frühen Entwicklungsprozessen sind. Diese Ergebnisse deuten darauf hin, dass die Hey Proteine zelltypspezifisch die Expression von Genen aus verschiedenen Signalwegen beeinflussen und modulieren können. Weiterhin eröffnen diese Daten neue Möglichkeiten für zukünftige Forschung, um die Rolle der Hey Proteine in der frühen Organentwicklung genauer ergründen. / During embryonic development, the Notch signaling pathway plays a central role in cell fate specification, proliferation and communication between neighboring cells. Hey basic helix-loop-helix (bHLH) transcription factors are targets of the Notch signaling pathway and show crucial functions in cardiovascular development. Hey2 knock out (KO) mice and Hey1/HeyL double knock out mice (DKO) exhibit incomplete formation of the septum and valves in heart development and defects in the differentiation of blood vessels. The aim of this study was to find new target genes of the Hey proteins to further clarify their function during organ development and to get more insight into the phenotypes of the Hey KO mice. Towards this goal, a combination of microarray analysis and chromatin immunoprecipitation (ChIP) was chosen to get an overview of genes directly regulated by Hey proteins in a cell culture model of HEK293 cells overexpressing doxycycline-inducible Flag-tagged Hey1 or Hey2. Microarray analysis revealed approximately 100 target genes that were downregulated up to 5-fold by both Hey1 and Hey2. Interestingly, 15 genes were upregulated more than 2-fold by Hey2. ChIP with αFlag antibody confirmed direct interaction of Hey1 and Hey2 with the proximal promotor regions of 4 downregulated target genes (HEY1, BMP2, KLF10 und FOXC1). Overexpression of mutant Hey1 proteins with deletion of the DNA-binding basic domain or single amino acid exchanges in the basic domain (3 arginine to lysine), failed to downregulate Hey1 target genes. Additionally, ChIP assay demonstrated that the binding of the Hey mutants to the promotor regions of HEY1, BMP2, KLF10 or FOXC1 is abolished, suggesting an essential role for the basic domain in DNA binding and function of the Hey proteins. We then utilized ChIP-PET in conjunction with highthroughput sequencing to perform a genome-wide screen for Hey1 and Hey2 binding sites in HEK293 cells. 1453 and 4288 target genes were identified for Hey1 and Hey2, respectively, of which 1147 genes were targets of both Hey1 and Hey2. Although the binding sites were located upstream and downstream of coding sequences, or even in exons and introns, 55 % and 49 % of all binding sites for Hey1 and Hey2, respectively, were located in proximal promoter regions between -0.5 kb and the first exon. An in silico binding motif analysis suggests a repetitive GC-rich sequence for Hey binding which is probably located in CpG islands, indicating a direct regulation of the transcriptional machinery by the Hey proteins. A comparison of the target genes from microarray analysis and ChIP-PET sequencing data demonstrates a large number of downregulated genes with binding sites that are directly bound by the Hey proteins. While 60 % of Hey2 downregulated genes contain binding sites, only 20 % of upregulated genes have binding sites for Hey2, invoking a repressor function for Hey proteins. To investigate, whether the Hey proteins regulate different target genes in a cell type specific manner, embryonic stem cells (ES cells) were generated which also overexpress doxycycline-inducible Hey1 or Hey2. These ES cells could be differentiated efficiently into cardiomyocytes and thus could also be used for gene expression analysis after induction of Hey overexpression. Microarray analysis of ES cells and cardiomyocytes resulted in more up- than downregulated genes in comparison to HEK293 cells. The overlap of common regulated genes in HEK293, ES cells and cardiomyocytes was very low. Only two Hey2 target genes were regulated more than 2-fold in both HEK293 and ES cells (Hes1, Zic2). This disparity indicates a narrow cell-type specific gene regulation by Hey proteins. Gene ontology analysis of all target genes demonstrated interactions of Hey proteins with different signaling pathways (e.g. TGFβ, Id or Wnt signaling), which are all indispensable for early developmental processes. These results show that Hey proteins influence and modulate gene expression levels in different signaling pathways in a cell-type specific manner. These data provide new possibilities for future research efforts to elucidate the role of Hey proteins in early organ development.

Notch Proteine gehören zu einer Familie hochkonservierter Transmembranrezeptoren, die bei Entwicklungsprozessen, beispielsweise im hämatopoetischen System, eine bedeutende Rolle spielen. So konnte eine Beteiligung von Notch und seinen Liganden bei der Differenzierung lymphatischer Zellen im Knochenmark, Thymus sowie in peripheren Organen gezeigt werden. Insbesondere unterdrückt Notch1 nach Ligandenbindung die Bildung von B-Zellen im Thymus und fördert stattdessen die Entwicklung von T-Zellen. Durch Aktivierung der γ-Sekretase wird die intrazelluläre Domäne von Notch1 freigesetzt und transloziert in den Kern. Dort wirkt Notch1 zusammen mit Proteinen der CSL Familie als Transkriptionsfaktor und reguliert so die Expression von Zielgenen wie beispielsweise HES1. Weiterhin kann die Signaltransduktion von Notch1 im Zytosol durch Wechselwirkung mit Proteinen wie Numb oder Deltex1 moduliert werden. Angesichts der Fähigkeit, die Proliferation unreifer Zellen aufrecht zu erhalten, kann eine aberrante Expression von Notch1 zur Entstehung von Neoplasien beitragen. Numb wird eine wichtige Rolle in der Regulation von Notch1 zugeschrieben, doch seine genaue Funktion in der T-Zellentwicklung ist unklar. Die Klonierung aller vier in der Ratte exprimierten Isoformen erlaubte erstmals deren detaillierte Charakterisierung. Die differentielle Expression und die preferentielle Interaktion von Numb i/o mit Notch1 deuten darauf hin, dass den verschiedenen Splice-Varianten nicht nur in der T-Zellentwicklung sondern auch in der Kontrolle von Notch1 unterschiedliche Funktionen zukommen. Diese Schlussfolgerung wird durch eine spezifische Expression der Numb Isoformen in humanen T-ALL Zelllinien weiter unterstützt. Transgene Ratten welche die intrazelluläre Domäne von Notch1 (N1IC) unter der Kontrolle des proximalen lck-Promoters überexprimieren (NICA), sind ein geeignetes Werkzeug, um die Rolle von Notch1 für die Genexpression und die Lymphomgenese zu untersuchen. Junge NICA-Ratten exprimieren N1IC spezifisch in Thymozyten, nicht jedoch in peripheren T-Zellen. In Folge dessen kommt es zu einer starken Expression bekannter Notch1 Zielgene, unter anderem dem präTα Rezeptor. Dies wiederum führt zur Substitution klassischer T-Zellrezeptor-Komplexe durch den präT-Zellrezeptor, was in adulten Tieren zur Entwicklung von thymischen Lymphomen beiträgt. Die Lymphomzellen in NICA-Ratten exprimieren das N1IC-Transgen sowohl im primären Tumor als auch nach Infiltration peripherer Organe. Dabei könnte die beobachtete Hochregulation von Notch3 ein möglicher Mechanismus der Tumorgenese in NICA-Ratten darstellen. Um die Ursache der Lymphomgenese besser zu verstehen, wurde eine genomweite Expressionsanalyse der Tumore mittels SAGE und Affymetrix DNA-Chips durchgeführt. Dabei wurden zahlreiche potentiell bedeutende Gene identifiziert, welche bei der Notch1-vermittelten Lymphomgenese eine Rolle spielen könnten. Die Beobachtung, dass ein Teil dieser Gene auch in humanen T-ALL-Zelllinien verändert ist, unterstreicht deren Bedeutung auch für die Pathogenese des Menschen. Eines der in NICA-Lymphomen am stärksten veränderten Gene wurde bislang weder in der Literatur noch in Sequenzdatenbanken beschrieben. Dieses von uns als Nip1 bezeichnete Protein ähnelt einem Enzym aus dem Isoprenoid-Stoffwechsel und ist vor allem im Thymus exprimiert. Ein anderes interessantes Kandidatengen ist CD30, ein Transmembran-Rezeptor welcher bei Hodgkin Lymphomen beschrieben wurde. Neben T-ALL wurde auch bei Hodgkin Lymphomen eine wichtige Rolle von Notch1 berichtet. Dabei könnte dessen hohe Expression im Zusammenhang mit der Transformation und dem Verlust der B-Zell-Identität dieser Tumore stehen. Um dies näher zu untersuchen, wurde Notch1 mittels RNAi bzw. durch Überexpression des negativen Regulators Deltex1 in Hodgkin-Zelllinien inaktiviert. Die hierbei erhaltenen Ergebnisse unterstützen die These, dass nach Repression der Notch1 Aktivität die Expression B-Zellspezifischer Marker wieder gewonnen werden kann. Zusammengefasst deuten die Ergebnisse dieser Arbeit auf eine bedeutende Funktion von Notch1 bei der T-Zellentwicklung und der Entstehung von Lymphomen hin. Dabei scheint insbesondere die Interaktion mit Proteinen wie Numb und Deltex1 sowie die Regulation der Genexpression wichtig zu sein. Die gewonnenen Daten könnten in Zukunft einen neuen Ansatzpunkt zur Entwicklung verbesserter Strategien zur Therapie von Krebserkrankungen des hämatopoetischen System bilden. / Notch proteins belong to a family of highly conserved transmembrane receptors, which play an important role in development processes such as in the hematopoietic system. Participation of Notch and its ligands could be demonstrated during differentiation of lymphatic cells in the bone marrow, thymus as well as in peripheral organs. After ligand binding, Notch1 particularly suppresses the formation of B-cells in the thymus while promoting the development of T-cells. The intracellular domain of Notch1 is released by activated γ-secretase and translocates into the nucleus. There, Notch1 and proteins of the CSL family act as a transcription factor and regulate the expression of target genes, for example HES1. Furthermore, signal transduction by Notch1 is modulated within the cytosol by interaction with proteins such as Numb or Deltex1. In view of the ability to maintain the proliferation of immature cells, aberrant Notch1 expression can also contribute to the formation of neoplasias. Numb is attributed an important role in the regulation of Notch1, but its exact function in T-cell development remains unclear. Cloning of all four Numb iso-forms in the rat permitted for the first time their detailed characterisation. The differential expression and the preferential interaction of Numb i/o with Notch1 indicate that the different splice variants not only play distinct roles in T-cell development but also in the control of Notch1 function. The observation that the Numb iso-forms in human T-ALL cell lines are also differentially expressed further supports this conclusion. Transgenic rats (NICA) which overexpress the intracellular domain of Notch1 under the control of the proximal lck promoter (N1IC), are a suitable tool to study the role that Notch1 has for gene expression and lymphomagenesis. Young NICA rats specifically express N1IC in thymocytes but not in peripheral T-cells. This leads to a strong expression of Notch1 target genes, e.g. the preTα chain. Consequently, classical TCR complexes are substituted by the preT cell receptor, which presumably contributes to the development of thymic lymphomas in adult animals. The lymphoma cells in NICA rats express the N1IC transgene both in the primary tumor and after infiltration of peripheral organs. In this context, the observed upregulation of Notch3 could represent a possible mechanism of tumorigenesis in NICA rats. In order to understand the cause of the lymphomagenesis, a genome-wide expression analysis of the tumors was accomplished by means of SAGE and Affymetrix DNA chips. Numerous potentially meaningfull genes were identified, which could play a role for Notch1-dependent lymphomagenesis. The observation that part of these genes is also altered in human T-ALL cell lines, underlines their meaning also for human pathogenesis. One of the genes that was most strongly upregulated in NICA lymphomas has so far neither been described in literature nor in sequence data bases. This protein, designated Nip1, resembles an enzyme from the isoprenoid metabolism and is particularly expressed in the thymus. Another interesting candidate gene is CD30, a transmembrane receptor which was described in Hodgkin lymphoma. Besides T-ALL, an important role of Notch1 was also reported in the context of Hodgkin lymphomas. Its high expression is potentially related to the transformation and loss of the B-cell identity of these tumor. In order to examine this in more detail, Notch1 was inactivated in Hodgkin lymphoma cell lines by means of RNAi and by overexpression of the negative modulator Deltex1. The results that have hereby been obtained are in support of the the notion that expression of B-cell specific markers can be regained after repression of Notch1 activity. In summary, the results of this work point towards an important function of Notch1 during T-cell development and the formation of lymphomas. In particular, interaction with proteins such as Numb and Deltex1 as well as regulation of gene expression appear to be important for these Notch1 functions. In the future, the obtained data could form a basis for the development of improved strategies to treat malignancies of the hematopoietic system.

Thesis advisor: Marc A.T. Muskavitch / The Notch pathway is an evolutionarily conserved mechanism of intercellular signaling that plays a central role in the development of metazoans. Here I summarize two genetic screens that utilize a rough eye phenotype created by Delta overexpression in the Drosophila eye to identify modulators of Notch pathway signaling activity. Among the many "hits" obtained from both screens, I have mapped to the Not1 gene a single complementation group that exhibits strong genetic interactions with Notch pathway mutants. NOT1 is a component of the CCR4-NOT complex, a global regulator of gene expression that exerts its effects through a variety of mechanisms, including mRNA deadenylation and direct transcriptional repression. I have conducted a series of genetic and molecular experiments in an effort to obtain more insight into the relationship between the CCR4-NOT complex and the Notch pathway. Both Not1 EMS mutations and RNAi-mediated knockdown of NOT1 expression produce phenotypes that mimic those of Notch loss-of-function pathway mutants. Knockdown of NOT1 in the developing bristle organ disrupts Notch-mediated inhibition of neuronal specification, resulting in supernumerary neurons and aberrant sheath cell specification. Knockdown of NOT1 within the developing wing margin disrupts expression of the Notch target genes Cut and Wingless, as well as the Notch ligand Delta. Phenotypic rescue experiments imply that Not1 functions downstream of Notch signal activation and acts directly on Notch target gene expression. These results suggest that NOT1 is required for Notch signal transmission in certain developmental contexts and implicate the CCR4-NOT complex as a positive regulator of the Notch pathway. / Thesis (PhD) — Boston College, 2009. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Biology.

Notch signaling plays an important role in developmental and pathological angiogenesis. Notch ligands, Dll4 and Jag1, have been implicated in tumor angiogenesis. Inhibition of Dll4-mediated Notch signaling results in hypersprouting of non-functional vasculature in tumors. We have constructed and analyzed pan-Notch ligand inhibitors, Notch1 decoys 1-24 and 1-36, which are based on the extracellular EGF-like repeats of Notch1. Both Notch1 decoys block angiogenesis in in vitro endothelial cell-based assays and in the mouse retina. We also show that they similarly inhibit Dll4- and Jag1-induced Notch signaling in vitro and result in a significant decrease in tumor growth and tumor vasculature in mouse and human tumor xenograft models. Interestingly, truncated Notch1 decoy variants, Notch1 decoys 1-13 and 10-24, act as ligand-specific Notch inhibitors. Notch1 decoy 1-13 is Dll4-specific whereas Notch1 decoy 10-24 is Jag1-specific. Ligand-specific Notch1 decoys effectively reduce tumor growth in tumor xenograft models in the mouse, including Mm5MT-FGF4, KP1-VEGF, LLC, and B16-F10. Notch1 decoy 1-13 has been demonstrated to increase tumor vasculature by increasing endothelial sprouting and number of tip cells. However, similar to the previously reported effects of Dll4 blockade, the tumor vessels are poorly perfused and hardly functional. On the other hand, Jag1-specific Notch1 decoy 10-24 significantly reduces tumor vessel density and disrupting endothelial-pericyte interactions, causing the impaired vascular structure and attenuated vascular perfusion. In addition, Notch1 decoys 1-13, 10-24, and 1-24 show an anti-metastatic potential in causing a delay of lung metastasis in the B16-F10 tumor model. Unlike gamma-secretase inhibitors and Dll4-blocking agents, Notch1 decoys do not cause GI-associated toxicity or vascular neoplasms. Therefore, our Notch1 decoys may represent a novel alternative and may hold future promise for Notch-targeted cancer therapy.

Non-alcoholic steatohepatitis (NASH) is a chronic liver disease associated with the worldwide spread of obesity. NASH predisposes development of fibrosis and hepatocellular carcinoma (HCC), but has no approved therapy due to incomplete understanding of the pathogenesis. Notch signaling normally specifies cell fate during development, but here we investigate how this pathway becomes dysregulated in NASH and contributes to fibrosis and cancer. In the first study, we show that hepatocyte Notch activity tracks with disease severity and treatment response in NASH patients, and is similarly increased in a mouse model of diet-induced NASH and liver fibrosis. Different genetic models demonstrate causatively that hepatocyte Notch induces liver fibrosis via secretion of the fibrogenic factor Osteopontin that activates hepatic stellate cells (HSCs), while pharmacologic inhibition of hepatocyte Notch could ameliorate NASH-associated fibrosis. In the second study, we research how hepatocyte Notch activation leads to HCC in mice on NASH diet. Transcriptomic analysis reveals nerve growth factor (NGF) as a Notch target gene in hepatocytes, and the abundance of hepatocyte NGF precursor protein (proNGF) is uniquely associated with HCC. We provide evidence that proNGF may facilitate HCC growth and expansion in a non-cell autonomous manner by inducing HSC deactivation and fibrosis remodeling. In summary, hepatocyte Notch maladaptively contributes to fibrogenesis and possibly HCC expansion by directly signaling to HSCs at different stages of NASH progression, and could be an accessible target for treatment of NASH-associated liver pathologies.

The Notch signalling pathway plays an essential role in cell-fate decisions and morphogenesis, and is frequently ectopically activated in human cancers. The signal is initiated through DSL ligand-dependent Notch proteolysis, which releases its intracellular domain. However, over-expression of the E3 ubiquitin ligase Deltex can bypass the requirement for Notch ligands and ectopically activate Notch by directing it into the endocytic pathway. It has been shown that Deltex induced endocytic-Notch signalling, but not DSL-induced signalling, requires both HOPS and AP-3 complex components that mediate trafficking to lysosomes and related organelles. I showed through a combination of the analysis of the dxsm mutant allele and by expression of mutant forms of Deltex in cell culture, that the C-terminal region of Deltex is important for Notch signalling, but is not required for Notch endocytosis. This suggests that the C-terminal region contains an interaction site that may direct endocytosed Notch to the correct endocytic compartment. A key question however was whether this endocytic pathway is utilised for full Notch signaling in normal development alongside the canonical activation mechanism. I therefore investigated the expression of Drosophila midline single-minded (sim), which is a Notch signal reporter gene, and embryonic neurogenesis which is repressed by Notch signalling. I found that deltex, HOPS and AP-3 mutants displayed gaps in sim expression and also a neurogenic phenotype similar to Notch loss-of-function, and consistent with a role for ligand-independent Notch activation in normal development. I found that the penetrance of these phenotypes increased when flies were cultured at higher temperature. These results suggest that Dx/HOPS/AP-3-dependent Notch activation provides a developmental robustness to the Notch signalling network. It was also found that AP-3 and HOPS components have an additional role in regulating cell survival, which is partially separable in time using a temperature shift assay. I also revealed a requirement of maternal Notch for cell survival in early embryonic development, which may be related to the HOPS-dependent function. Comparison of deltex, AP-3 and HOPS mutant phenotypes suggested that there may be functional redundancy of deltex with components that regulate Notch endocytosis, and of AP-3 with proteins that mediate subsequent trafficking to the late endosome/lysosome. Finally I characterised the molecular lesions of an allele of the HOPS component light and the AP-3 component ruby and identified lesions which were consistent with the loss-of-function of these genes.