Unusual DNA secondary structures, especially G-quadruplexes and i-motifs, play important roles in gene transcriptional regulation and have been identified as novel drug targets. In this dissertation, I explored their formation in the human VEGF and RET promoters and their roles in gene transcriptional regulation. VEGF is a key regulator of angiogenesis and is up-regulated in many types of tumors. A poly-guanine/poly-cytosine (polyG/polyC) tract in its proximal promoter (-85 to -50 base pairs relative to the transcription starting site) is essential for both basal and inducible VEGF expression. I demonstrated that the guanine-rich (G-rich) and cytosine-rich (C-rich) strands in the VEGF proximal promoter are able to form G-quadruplex and i-motif structures, respectively. The major G-quadruplex formed by the VEGF G-rich sequence is an intramolecular parallel G-quadruplex containing three G-tetrads and a 1:4:1 arrangement of three double-chain-reversal loops (two single-base loops and one loop with four bases). The complementary C-rich sequence in the same region forms an intramolecular i-motif containing six semiprotonated cytosine-cytosine⁺ base pairs and a 2:3:2 loop configuration (two double-base loops and one loop with three bases). The Gquadruplexes formed by the native VEGF G-rich and its derivative sequences were also confirmed by NMR. In addition, various transcription factors including Sp1, hnRNP K, CNBP and nucleolin, which recognize different DNA structural elements including single-stranded, double-stranded or G-quadruplex/i-motif DNA in the VEGF proximal promoter, have been confirmed by EMSA, siRNA and chromatin immunoprecipitation (ChIP) assay, suggesting that the DNA in the VEGF proximal promoter region is capable of undergoing transitions between those three structures. Based on my studies, I have proposed a model to describe how various transcription factors recognize different DNA structures in the VEGF proximal promoter to regulate transcription. In the proximal promoter of another important oncogene RET, I demonstrated that the guanine-rich strand forms an intramolecular parallel G-quadruplex containing three G-tetrads and a 1:3:1 arrangement of three double-chain-reversal loops. The complementary cytosine-rich strand forms an i-motif structure containing six semiprotonated cytosine-cytosine⁺ base pairs and a 2:3:2 loop configuration. Moreover, G-quadruplex-interactive compounds TMPyP4 and telomestatin were shown to further stabilize the RET G-quadruplex structure.

Induction of vertebrate heart development requires inhibition of canonical/<i>β</i>-catenin dependent Wnt signalling, activation of non-canonical/<i>β</i>-catenin independent Wnt signalling and transcription factor activation. Wnt/<i>β</i>-catenin signalling may also have a later regulatory role in cardiogenesis. The recent discovery of Wnt6 expression next to and within the developing heart during the relevant stages of cardiomyogenesis, combined with knockdown and over-expression data suggests that Wnt6 may have a role in the regulation of this process. Inhibition of canonical signalling leads to increased expression of cardiac associated transcription factors such as members of the Nkx2 and GATA family. These families are expressed in overlapping regions which specify the early heart field prior to the expression of the later cardiomyocyte-specific genes. This study demonstrates the ability of <i>β</i>-catenin to inhibit cardiogenesis during later developmental stages, before the cardiac mesoderm begins to differentiate into myocardium (heart muscle) and that the newly discovered Wnt6 exerts inhibition of cardiogenesis in a <i>β-</i>catenin<i> </i>dependent manner. This inhibition of cardiogenesis by <i>β-</i>catenin can occur in a cell-autonomous manner and is a result of direct inhibition of cardiac transcription factors of the GATA family. Over-expression of these pro-cardiogenic transcription factors GATA4 and GATA6 can restore the cardiomyogenic differentiation programme in embryos where it has previously been inhibited by <i>β</i>-catenin. In conclusion GATA factors are the relevant targets of Wnt/<i>β-</i>catenin signalling in the inhibition of normal cardiac development. The subsequent loss of cardiac gene expression observed is therefore a result of insufficient GATA expression and function.

Cell-type specific dendritic morphologies emerge via complex growth mechanisms modulated by intrinsic and extrinsic signaling coupled with activity-dependent regulation. Combined, these processes converge on cytoskeletal effectors to direct dendritic arbor development, stabilize mature architecture, and facilitate structural plasticity. Transcription factors (TFs) function as essential cell intrinsic regulators of dendritogenesis involving both combinatorial and cell-type specific effects, however the molecular mechanisms via which these TFs govern arbor development and dynamics remain poorly understood. Studies in Drosophila dendritic arborization (da) sensory neurons have revealed combinatorial roles of the TFs Cut and Knot in modulating dendritic morphology, however putative convergent nodal points of Cut/Knot cytoskeletal regulation remain elusive. Here we use a combined neurogenomic, bioinformatic, and genetic approach to identify and molecularly characterize downstream effectors of these TFs. From these analyses, we identified Formin3 (Form3) as a convergent transcriptional target of both Cut and Knot. We demonstrate that Form3 functions cell-autonomously in class IV (CIV) da neurons to stabilize distal higher order branching along the proximal-distal axis of dendritic arbors. Furthermore, live confocal imaging of multi-fluor cytoskeletal reporters and IHC analyses reveal that form3 mutants exhibit a specific collapse of the dendritic microtubule (MT) cytoskeleton, the functional consequences of which include defective dendritic trafficking of mitochondria and satellite Golgi. Biochemical analyses reveal Form3 directly interacts with MTs via the FH1/FH2 domains. Form3 is predicted to interact with two alpha-tubulin N-acetyltransferases (ATAT1) suggesting it may promote MT stabilization via acetylation. Analyses of acetylated dendritic MTs supports this hypothesis as defects in form3 lead to reductions, whereas overexpression promotes increases in MT acetylation. Neurologically, mutations in Inverted Formin 2 (INF2; the human ortholog of form3) have been causally linked to dominant intermediate Charcot-Marie-Tooth (CMT) disease E. CMT sensory neuropathies lead to distal sensory loss resulting in a reduced ability to sense heat, cold, and pain. Intriguingly, disruption of form3 function in CIV nociceptive neurons results in a severe impairment in nocifensive behavior in response to noxious heat, which can be rescued by expression of INF2 revealing shared primordial functions in regulating nociception and providing novel mechanistic insights into the potential etiological bases of CMT sensory neuropathies.

Student Number : 9503239E - PhD thesis - Faculty of Science / Malaria is one of the most debilitating pathogenic infections known to man, responsible for approximately three million deaths annually, primarily children in sub-Saharan Africa. The parasite has evaded multiple attempts at eradication, predominantly through the complexity of its life cycle, the ability to elude host immune response, and gametocyte formation to ensure dissemination. The recent completion of the genome sequence has opened up a multitude of avenues for exploration and identification of novel drug and vaccine targets, as well as providing a glimpse into the complex mechanisms that have contributed to the success of this pathogen. The mechanisms of gene regulation, especially those governing gametocytogenesis, have, however, not yet been elucidated. In this research, differential display has been used to identify some of the genes that are differentially expressed between the asexual parasite and gametocyte stages of P. falciparum. Numerous genes involved in diverse aspects of metabolism, protein synthesis and immune evasion were identified. A combination of BLASTN and BLASTX similarity searches was used to categorize and increase the confidence with which a transcript could be identified. Expression data for confidently identified genes were confirmed using reverse slot blot and available microarray data. PfMyb2, a novel transcription factor which may regulate genes involved in gametocytogenesis, was characterized. The DNA binding domains of the protein were cloned and expressed as a histidine fusion protein. Mobility shift assays were used to assess the in vitro binding capability of the recombinant 6xHis-PfMyb2, which bound to oligonucleotides containing the consensus Myb regulatory element. Two of the oligonucleotides represent sequences located within promoters of P. falciparum genes (Pfcrk1 and Pfmap1) known to play a role in regulating the cell cycle, a function ascribed to many members of the vertebrate Myb family. The identification of PfMyb2 as a bona fide transcription factor is a first step into gaining some insight into the many regulatory processes that occur during the life cycle of this complex organism. A better understanding of the molecular mechanisms that govern its survival is essential for the ultimate eradication of this deadly parasite.

The development and function of the heart is governed by a conserved set of transcription factors (TFs) that regulate gene expression in a cell-type, time point and stimulus driven manner. Of these core cardiac TFs, the most ubiquitously expressed is the zinc finger protein GATA4. In cardiomyocytes, GATA4 is central to proliferation, differentiation, hypertrophy and induction of pro-survival pathways. In cardiac endothelial cells, it is required for valve and septal development, although the exact mechanisms remain unclear. To regulate such a wide array of functions in a spatially and temporally controlled manner, GATA4 interacts with specific protein partners, the majority of whom have been identified in cardiomyocytes. However, a complete understanding of the protein interactome of GATA4, particularly in cardiac endothelial cells, has not yet been achieved. Using a mass spectrometry-based approach, we have identified a series of novel GATA4 interacting partners in cardiac endothelial cells. 3xFlag GATA4 was stably overexpressed via retroviral transduction in the TC13 cardiac endothelial precursor cell line, immunoprecipitated from nuclear protein extracts and sent for HPLC-ESI-MS/MS. Several novel GATA4 interacting partners were identified including the chaperone protein Heat Shock Protein 70 (HSP70), the inducible orphan nuclear receptor Nerve Growth Factor 1β (NGFIβ, NUR77) and the Drosophila-Binding/Human Splicing protein family members Non-POU Domain Containing Octamer Binding Protein (NONO) and Paraspeckle 1 (PSPC1). Chapter 1 discusses the interaction between GATA4 and HSP70 and its role in cardiomyocyte survival upon exposure to chemotherapeutic agent Doxorubicin (DOX). HSP70 binds directly to GATA4, preventing DOX-mediated cleavage and degradation by Caspase-1, cardiomyocyte cell death and heart failure. Chapter 2 focuses on the cooperative interaction between GATA4 and NUR77 in cardiac microvascular endothelial cells and its central role in myocardial angiogenesis in response to pressure overload. The GATA4-NUR77 complex transactivates the promoter of Angiopoietin-Like 7 (ANGPTL7), a secreted pro-angiogenic chemotactic factor, triggering endothelial cell proliferation and tube formation in cultured cardiac endothelial cells and increasing myocardial capillary density in vivo. Chapter 3 discusses the interaction between GATA4 and the DBHS proteins NONO and PSPC1 in the regulation of cardiac development. These proteins play opposing roles when bound to GATA4 as PSPC1 enhances GATA4 activation of critical cardiac promoter targets and NONO acts as a rheostat to repress GATA4 activity. In vivo, loss of NONO results in left ventricular non-compaction consistent with humans with loss-of-function mutations. However, simultaneous Gata4 haploinsufficiency partially rescues this phenotype. Together, this data identifies multiple novel cell type and time point specific GATA4 protein partners and sheds light on GATA4 regulatory mechanisms in cardiac development and disease.

Numerous drugs in the current pharmacopoeia originate from plant sources. Plant cell culture represents an alternative source for producing high-value secondary metabolites including paclitaxel. Paclitaxel is mainly derived from the plant genus Taxus and has been widely used in cancer chemotherapy. However, plant cell culture is often not commercially viable because of difficulties associated with culturing dedifferentiated plant cells (DDCs) on an industrial scale. Therefore, we isolated and cultured innately undifferentiated cambial meristematic cells (CMCs) from Taxus cuspidata, which possess superior growth properties relative to DDCs. These CMCs have been demonstrated to be a cost effective platform for the sustainable production of paclitaxel. Using 454 sequencing, we determined the transcriptome of T. cuspidata CMCs. Utilizing this transcriptome as a reference, we then employed Solexa digital gene expression profiling to identify transcriptional regulators that were induced by methyl jasmonate, an activator of paclitaxel biosynthesis. This lead to the discovery of 19 putative transcription factors (TFs) belonged to 5 TF families which were further confirmed by associated molecular methods. We aimed to identify which of these 19 regulatory proteins drive the expression of 5 paclitaxel biosynthetic genes by employing yeast one-hybrid analysis and electrophoretic mobility shift assays. Further, the cis-regulatory elements associated with these TFs were identified in the promoter regions of the two early, taxadiene synthase (TASY) and taxadiene 5α hydroxylase (T5αH), and three late, 10-deacetylbaccatin III-10-O-acetyltransferese (DBAT), phenylalanine aminomutase (PAM) and 3'-N-debenoyl-2-N-benzoyltransferase (DBTNBT), paclitaxel biosynthetic genes to facilitate the TF-DNA binding studies. Finally, understanding the TF regulatory network underlying paclitaxel biosynthesis can guide the engineering of CMCs to elevate the production of this key pharmaceutical.

Phytochromes are the photoreceptors mainly responsible for the detection of red and far-red (FR) light and the following responses. HFR1 is a basic helix-loop-helix type putative transcription factor involved in Phytochrome A signaling pathway. First we look at the early phenotype of mutant seedlings lacking a functional HFR1 gene and we show that auxin is involved in the increased hypocotyl phenotype of these seedlings. Northern blots and RT-PCRs showed that ATHB-2, a gene involved in shade avoidance is regulated by HFR1 under FR light. Microarray experiments were performed to find the genes that are early targets of regulation by HFR1.

Despite the identification of MYCN amplification as an adverse prognostic marker in neuroblastoma, no drugs that target MYCN have yet been developed. Here, by combining a whole genome shRNA library screen and Master Regulator Inference Algorithm (MARINa) analysis, we identified Transcription Factor Activating Protein 4 (TFAP4) as a novel synthetic lethal interactor with MYCN amplification in neuroblastoma. Silencing TFAP4 selectively inhibits MYCN amplified neuroblastoma growth both in vitro and in xenograft mice models. TFAP4 expression is inversely correlated with patient survival in MYCN-high neuroblastoma. Mechanistically, silencing TFAP4 induces neuroblastoma differentiation, as seen by increased neurite outgrowth, and up-regulation of neuronal markers. TFAP4 regulates a downstream signature similar to the signature of the oncogene anaplastic lymphoma kinase (ALK). Taken together, our results validate TFAP4 as an important master regulator in MYCN amplified neuroblastoma and a novel synthetic interactor with MYCN amplification. Thus, TFAP4 may be a novel drug target for neuroblastoma treatment.

At the core of gene regulatory networks are transcription factors (TFs) that recognize specific DNA sequences and target distinct gene sets. Characterizing the DNA binding specificity of all TFs is a prerequisite for understanding global gene regulatory logic, which in recent years has resulted in the development of high-throughput methods that probe TF specificity in vitro and are now routinely used to inform or interpret in vivo studies. Despite the broad success of such methods, several challenges remain, two of which are addressed in this thesis. Genomic DNA can harbor different epigenetic marks that have the potential to alter TF binding, the most prominent being CpG methylation. Given the vast number of modified CpGs in the human genome and an increasing body of literature suggesting a link between epigenetic changes and genome instability, or the onset of disease such as cancer, methods that can characterize the sensitivity of TFs to DNA methylation are needed to mechanistically interpret its impact on gene expression. We developed a high-throughput in vitro method (EpiSELEX-seq) that probes TF binding to unmodified and modified DNA sequences in competition, resulting in high-resolution maps of TF binding preferences. We found that methylation sensitivity can vary between TFs of the the same structural family and is dependent on the position of the 5mCpG within the TF binding site. The importance of our in vitro profiling of methylation sensitivity is demonstrated by the preference of human p53 tetramers for 5mCpGs within its binding site core. This previously unknown, stabilizing effect is also detectable in p53 ChIP-seq data when comparing methylated and unmethylated sites genome-wide. A second impediment to predicting TF binding is our limited understanding of i) how cooperative participation of a TF in different complexes can alter their binding preference, and ii) how the detailed shape of DNA aids in creating a substrate for adaptive multi-TF binding. To address these questions in detail, we studied the in vitro binding preferences of three D. melanogaster homeodomain TFs: Homothorax (Hth), Extradenticle(Exd) and one of the eight Hox proteins. In vivo, Hth occurs in two splice forms: with (HthFL) and without (HthHM) the DNA binding domain (DBD). HthHM-Exd itself is a Hox cofactor that has been shown to induce latent sequence specificity upon complex formation with Hox proteins. There are three possible complexes that can be formed, all potentially having specific target genes: HthHM-Exd-Hox, HthFL-Exd-Hox, and HthFL-Exd. We characterized the in vitro binding preferences of each of these by developing new computational approaches to analyze high-throughput SELEX-seq data. We found distinct orientation and spacing preference for HthFL-Exd-Hox, alternative recognition modes that depend on the affinity class a sequence falls into, and a strong preference for a narrow DNA minor grove near Exd's N-terminal DBD. Strikingly, this shape readout is crucial to stabilize the HthHM-Exd-Hox complex in the absence of a Hth DBD and can thus be used to distinguish HthHM from HthFL isoform binding. Mutating the amino acids responsible for the shape readout by Exd and reinserting the engineered protein into the fly genome allowed us to classify in vivo binding sites based on ChIP-seq signal comparison between “shape-mutant” and wild-type Exd. In summary, the research presented here has investigated TF binding preferences beyond sequence context by combining novel high-throughput experimental and computational methods. This interdisciplinary approach has enabled us to study binding preferences of TF complexes with respect to the epigenetic landscape of their cognate binding sites. Our novel mechanistic insights into DNA shape readout have provided a new avenue of exploiting guided protein engineering to probe how specific TFs interact with their co-factors in a cellular context, and how flanking genomic sequence helps determine which multi-TF complexes will form and which binding mode a complex adopts.

The progressive loss of developmental potential is a hallmark of all differentiating cells in multicellular organisms. At the chromatin level, this restriction in cell-fate plasticity is established through the silencing of active and poised lineage-specific genes that are incompatible with the terminal fate of the maturing cell type. The effective and stable inhibition of gene expression relies on the coordinated action of transcriptional repressors. These repressors are often transiently expressed only at the time of cell-fate specification and direct lineage decisions by suppressing alternative developmental programs. However, compared to the numerous studies examining the mechanisms by which cell-type specific transcriptional activators program cellular identity, little is currently known regarding how transient repressors execute permanent silencing of gene regulatory networks. To address this question, I have examined the mechanisms through which the transiently expressed transcription factor (TF) Nkx2.2 represses the acquisition of motor neuron (MN) identity in V3 neuronal progenitors. While it is well-established that Nkx2.2 functions as a transcriptional repressor through its interactions with the Groucho (Grg) family of co-repressors, how these interactions manifest in gene silencing has remained unknown. Moreover, the effects of Nkx2.2 occupancy on chromatin modifications have not been determined. In this dissertation, I demonstrate that surprisingly, Nkx2.2 decommissions enhancers of the MN developmental program not through the recruitment of additional co-repressor proteins but rather through the eviction of co-activator complexes. While this displacement is dependent upon an intact Grg-interacting domain, Nkx2.2 binding does not increase Grg enrichment. In addition, extensive profiling of Nkx2.2 genome-wide binding events in neural precursors unexpectedly revealed that Nkx2.2 occupies not only enhancers of MN progenitor genes acutely repressed by Nkx2.2 but also enhancers of genes expressed exclusively in postmitotic MNs, long after Nkx2.2 expression has been down- regulated. In vivo lineage tracing experiments and in vitro genomic analyses demonstrated that Nkx2.2 also functions in a repressive capacity at these poised regulatory regions. Here, Nkx2.2 binding prevents the activation of postmitotic genetic networks through a preferential enlistment of histone deacetylase complex 2 (HDAC2) proteins. However, this binding is not accompanied by the deposition of repressive chromatin modifications, and removal of Nkx2.2 in differentiating V3 neurons leads to the ectopic expression of the postmitotic MN TFs Isl1 and Hb9. Collectively, these studies indicate that transiently expressed repressors may establish gene suppression by counteracting the activities of transcriptional activators, rather than by directly establishing repressive chromatin signatures. As transcriptional reprogramming of differentiated cell linages often fails to adequately silence the expression programs of the starting population, these results may help to inform new methodologies for instructing cell conversions.