Multicellular organisms are composed of a number of different specialised cells that all carry the same genetic material but are highly divergent in their functions and characteristics. This diversity is only allowed because sets of specific genes are expressed in one type of cells and silent in others. A precise control mechanism is required to fine-tune gene regulation and relies on chromatin structure and regulatory proteins. One of the largest families of DNA-binding factors that influence this in human and mouse is the KRAB zinc finger protein (KZFP) family. KZFPs are thought to have rapidly evolved alongside transposable elements and be mediators of transcriptional repression. The few KZFPs that have been characterised so far have been shown to be involved in a wide range of regulatory and biological processes; hence it is hard to make functional generalisations. During my PhD, I studied one member of the KZFP family in mouse, ZFP263, with the aim of understanding its mechanism of action in mouse embryonic stem cells (mESCs) and its role in mice. My work has shown that ZFP263 is an ancient protein highly conserved in mammals and under purifying selection. One of its two functional domains however is divergent from the consensus sequence found in most KZFPs and suggests that ZFP263 might have lost the ability to recruit repressive chromatin states. My research identified the targets of ZFP263 binding in mESCs and showed that it does not bind and silence transposable elements. Indeed it targets unique regions of the genome, mostly within transcribed regions of genes. These genes show a wide range of expression levels and are involved in several key biological processes. Surprisingly, binding sites are not associated with the canonical KZFP co-factor but mostly co-localize with active histone marks. My findings lead me to hypothesise that ZFP263 has evolved to target active epigenetic states to unique regions that are positive regulators of transcription, in contrast to the more canonical model of KZFP function. To test this hypothesis, I have generated targeted mutations at Zfp263 in mice using CRISPR-Cas9 and my preliminary results suggest that Zfp263 mutants have growth defects indicating a role for this protein in mouse development. My findings indicate that ZFP263 is a unique KZFP with non-canonical properties and provide novel insights into the evolution and functions of KZFPs in mammals.

Transcriptional control plays a major role in regulating hypoxic responses in plants. However, the transcriptional regulatory networks associated with hypoxia remain to be constructed. By transcriptomic analysis I show here that a novel systemic transcriptional reprogramming, which is mediated via the interplay of hormones, facilitates the survival of plants under flooding. A feasible strategy for identifying downstream targets of transcription factors (TFs) was developed. The downstream pathways of a hypoxia-responsive TF, WRKY22, were constructed. The results also show that AtERF73/HRE1 (Arabidopsis thaliana Ethylene Response Factor 73/Hypoxia Responsive ERF 1) modulate ethylene-dependent and -independent responses during hypoxia. Transcriptomic analysis of Arabidopsis in both root and shoot tissues during flooding of roots indicates the existence of a systemic communication through transcriptional reprogramming. By functional classification of affected genes, a comprehensive managing program of carbohydrate metabolism was observed. Through transcriptional profiling in ethylene and abscisic acid (ABA) signaling mutants, ein2-5 and abi4-1, an alteration of long-distance hypoxic regulation was uncovered in ein2-5 and abi4-1. Moreover, genes involved in ABA biosynthesis were also found to be differentially regulated between shoots and roots. Many members of the WRKY TF family were highly induced by hypoxia. One of the early-induced WRKYs, WRKY22, which has the highest induced level, was chosen for identifying its downstream targets. Anoxic tolerance was affected in WRKY22 overexpressing (WRKY22-OX) and knock-out (wrky22-ko) lines. Comparison of differential gene expression profiles between the wild-type and WRKY22-OX and between the wild-type and wrky22-ko lines by microarray analysis identified novel hypoxia-responsive genes as WRKY22 targets. Chromatin immunoprecipitation (ChIP) followed by microarray hybridization (ChIP-chip) and ChIP followed by quantitative PCR (ChIP-qPCR) were utilized to analyze in vivo interactions. To study the role of ethylene during hypoxia, I characterized an AP2/ERF (APETALA2/ethylene response factor) AtERF73/HRE1 that is specifically induced during hypoxia. I showed that the expression of AtERF73/HRE1 can be induced by exogenous 1-aminocyclopropane-1-carboxylic acid (ACC), a precursor of ethylene. Its hypoxic induction was reduced but not completely abolished in ethylene-insensitive mutants and in the presence of inhibitors of ethylene biosynthesis and responses. Increased ethylene sensitivity and exaggerated triple responses were observed in HRE1-RNAi knock-down lines. By comparing expression differences between the wild-type and HRE1-RNAi lines, I found that hypoxic induction of glycolytic and fermentative genes was reduced by the HRE1-RNAi knock-down mutations, whereas induction of a number of peroxidase and cytochrome P450 genes was increased. Collectively, these results show that AtERF73/HRE1 is involved in modulating ethylene responses under both normoxia and hypoxia.

The fluffy gene of Neurospora crassa is required for asexual sporulation. It encodes an 88 kDa polypeptide containing a typical fungal Zn2Cys6 DNA binding motif. To identify the target genes on which FL may act, I sought to identify target sequences to which the FL protein binds. Several strategies were attempted to obtain purified FL protein. Purification was achieved by expressing the DNA binding domain of FL in Escherichia coli as a fusion with glutathione S-transferase followed by affinity purification using glutathione sepharose chromatography. DNA binding sites were selected by in vitro binding assays. Comparison of the sequences of selected clones suggested that FL binds to the motif 5??-CGG(N)9CCG-3??. A potential binding site was found in the promoter region of the eas (ccg-2) gene, which encodes a fungal hydrophobin. In vitro competitive binding assays revealed a preferred binding site for FL in the eas promoter, 5??-CGGAAGTTTCCTCCG-3??, which is located 1498 bp upstream of the eas translation initiation codon. In vivo experiments using a foreign DNA sequence tag confirmed that this sequence is a target site for FL regulation. Using Saccharomyces cerevisiae as an experimental system, I demonstrated that the C-terminal portion of FL functions in transcriptional activation. Microarray analysis was performed to study the role of fl in gene regulation on a large scale. mRNA levels in a fl mutant were compared to those in a strain overexpressing the fl gene. Experiments with cDNA microarray containing 13% of the total number of predicted N. crassa genes revealed 122 genes differentially expressed in response to overexpression of fl. Among these, eas displayed the greatest level of response. The cDNA microarray approach also revealed a number of genes that may be indirectly regulated by fl but may be involved in development. This information provides a foundation for further analysis of the role of fl in conidial development.

Embryonic stem cells (ESCs) represent a promising source of cells for cell replacement therapy in the context of muscle diseases; however, before ESC-based cell therapy can be translated to the clinic, we must learn to modulate cell-fate decisions in order to maximize the yield of myocytes from this systems. In order to gain a better understanding of the myogenic cell fate, we sought to define the molecular mechanisms underlying the specification and differentiation of ESCs into cardiac and skeletal muscle. More specifically, the central hypothesis of the thesis is that myogenic signalling cascades modulate cell fate via regulation of transcription factors. Retinoic acid (RA) is known to promote skeletal myogenesis, however the molecular basis for this remains unknown. We showed that RA expands the premyogenic progenitor population in mouse stem cells by directly activating pro-myogenic transcription factors such as Pax3 and Meox1. RA also acts indirectly by activating the pro-myogenic Wnt signalling cascade while simultaneously inhibiting the anti-myogenic influence of BMP4. This ultimately resulted in a significant enhancement of skeletal myogenesis. Furthermore, we showed that this effect was conserved in human embryonic stem cells, with implications for directed differentiation and cell therapy. The regulation of cardiomyogenesis by the Wnt pathway was also investigated. We identified a novel interaction between the cardiomyogenic transcription factor Nkx2.5 and the myosin phosphatase (MP) enzyme complex. Interaction with MP resulted in exclusion of Nkx2.5 from the nucleus and inhibition of its transcriptional activity. Finally, we showed that this interaction was modulated by phosphorylation of the Mypt1 subunit of MP by ROCK, downstream of Wnt3a. Treatment of differentiating mouse ESCs with Wnt3a resulted in exclusion of Nkx2.5 from the nucleus and a subsequent failure to undergo terminal differentiation into cardiomyocytes. This likely represents part of the molecular basis for Wnt-mediated inhibition of terminal differentiation of cardiomyocytes. Taken together, our results provide novel insight into the relationship between myogenic signalling cascades and downstream transcription factors and into how they function together to orchestrate the myogenic cell fate in stem cells.

Versican, a chondroitin sulfate proteoglycan, is one of the main components of the extracellular matrix and hence plays a central role in tissue morphogenesis and a number of pathologic processes. My main goal has been to investigate the mechanisms of versican gene regulation, focusing on the signal transduction pathways, promoter regions, cis-acting elements,and trans- factors. This thesis puts forth new knowledge regarding transcriptional regulation of the human versican gene. In chapter III, I present the cloning of a 752-bp fragment of the human versican promoter (- 634/+118 bp) and nine stepwise 5' deletion fragments in the PGL3-luciferase reporter plasmid. Furthermore, I identify three potential enhancer and two repressor regions in this promoter. I also demonstrate that both cAMP and C/EBPf3 enhanced and repressed versican transcription in HeLa cells and rat aortic smooth muscle cells (SMC),respectively, suggesting that versican transcription is differentially regulated by the respective mediator and transcription factor in epithelial cells and SMC. In chapter IV, I reveal the role ofPI3K/PKB/GSK-30 signaling pathway in regulating versican promoter activity and transcription. Furthermore, I identify that the 0-catenin/TCF-4 transcription factor complex, one of the downstream targets of GSK-3[3, mediates versican promoter activity and transcription. In chapter V, I identify that variations in C-terminal regions of TCF family members determine the irrepressor or enhancer properties on Wnt target genes. Furthermore, I show that curcumin is a strong inhibitor of the P-catenin/TCF-p300 mediated gene expression. In chapter VI, I demonstrate that the androgen receptor trans-activates versican transcription in prostate cancer cells. Furthermore, I show cross-talk between the androgen receptor and 13-catenin in regulating versican transcription in prostate stromal fibroblasts. Overall, this study charts previously uncharacterized promoter elements, transcription factors, and signal transduction pathways involved in regulation of the versican gene.

Thymocyte development is a complex process that requires precise regulation of differentiation and proliferation. Basic helix-loop-helix (bHLH) transcription factors have been shown to be crucial for proper T cell development. HEB and E2A are structurally and functionally related E proteins of the bHLH family. These proteins directly regulate the expression of a number of genes essential for lymphocyte development in a lineage- and stage-specific manner. Abrogation or compromise of their function results in the manifestation of B and T cell developmental defects. Genetic and biochemical studies have provided evidence of a significant degree of functional redundancy among E proteins. The existence of compensational abilities among different E proteins has hampered the investigation and elucidation of E protein function. As such, single gene knockouts demonstrate only limited defects in lymphocyte development. Double E2A-HEB knockouts that could eliminate E protein redundancy are embryonic lethal. In addition, conventional gene knockouts are not well-suited for discerning between intrinsic and extrinsic defects caused by E protein disruption. To eliminate functional compensation and to test the T cell intrinsic roles of E proteins during thymocyte development, we developed a conditional HEB-E2A double knockout. Specifically, we employed a loxP/Lck-Cre recombinase system to drive E protein deletion during early thymocyte development. Using this approach, we were able to reveal overlapping roles for HEB and E2A in thymocyte development that had been obscured in previous single gene knockout studies. We find that simultaneous deletion of HEB and E2A results in a severe block in thymocyte development at the DN to DP stage transition. This developmental block is accompanied by a dramatic decrease in total thymic cellularity, an increase in apoptosis, and a reduction of pTα expression. These developmentally arrested thymocytes exhibit increased proliferation in vivo and dramatic expansion ex vivo in response to IL-7 signaling. Our findings suggest that E2A and HEB are not only critical for the regulation of T cell differentiation but are also necessary to retain developing thymocytes in cell cycle arrest prior to pre-TCR expression. Together, these results imply that E proteins are required to coordinate thymocyte differentiation and proliferation. / Dissertation

During embryogenesis, the architecture of the plant and the food reserves for seed germination are established. Abscisic acid (ABA) regulates seed development and dormancy. It controls genes involved in stress responses. ABA-responsive basic leucine zipper (bZIP) transcription factors are identified by interaction with ABA responsive cis-regulatory elements. The transcription factor ABI5 is one of these. It regulates gene expression during embryogenesis and in response to ABA. An ABA-insensitive mutant, abi5-6, exhibits no gross morphological defects other than the effect on seed germination in the presence of ABA. Thus, microarray analysis was employed to search for molecular phenotypes. We used cDNA microarrays to analyze ABA regulated gene expression and the role of ABI5 in seedlings. 310 genes were identified as ABI5/ABA regulated genes. 161 of these genes were regulated by ABI5, and 134 of ABI5-regulated genes were co-regulated by ABA. Only a small number of genes altered expression in both Pro35S:ABI5 and abi5-6 genetic backgrounds indicating the preferential binding of the bZIP protein dimers to specific promoter sequences. To determine the optimal platform for identifying ABI5-regulated genes in seeds, a cDNA microarray, the Agilent Arabidopsis Oligo microarray, and the Affymetrix ATH1 arrays were tested. Cross platform comparisons utilized 4,518 genes present on all three platforms. The best correlation was between the Agilent and the Affymetrix results. Furthermore, the Affymetrix results correlated best with qRT-PCR validation data for selected genes. A small number of genes including AtCOR413 pm-1 showed a consistent expression pattern across the three platforms. A robust ABRE cis-regulatory element was identified in the promoter of AtCOR413 pm-1. Further studies showed binding of ABI5 to the promoter of AtCOR413 pm-1 by Electrophoretic Mobility Shift Assays (EMSA) and validated the expression of ABI5 and AtCOR413 pm-1 in abi5-6 seeds by qRT-PCR and RNA gel blot analysis. Transactivation assays using AtCOR413 pm-1 promoter:GUS fusions in Arabidopsis dry seed and seedlings revealed ABI5 acts as a negative regulator for AtCOR413 pm-1 in dry seeds, while other proteins may play major roles in regulating responses to ABA and low temperature (LT) in seedlings.

The fluffy gene of Neurospora crassa is required for asexual sporulation. It encodes an 88 kDa polypeptide containing a typical fungal Zn2Cys6 DNA binding motif. To identify the target genes on which FL may act, I sought to identify target sequences to which the FL protein binds. Several strategies were attempted to obtain purified FL protein. Purification was achieved by expressing the DNA binding domain of FL in Escherichia coli as a fusion with glutathione S-transferase followed by affinity purification using glutathione sepharose chromatography. DNA binding sites were selected by in vitro binding assays. Comparison of the sequences of selected clones suggested that FL binds to the motif 5??-CGG(N)9CCG-3??. A potential binding site was found in the promoter region of the eas (ccg-2) gene, which encodes a fungal hydrophobin. In vitro competitive binding assays revealed a preferred binding site for FL in the eas promoter, 5??-CGGAAGTTTCCTCCG-3??, which is located 1498 bp upstream of the eas translation initiation codon. In vivo experiments using a foreign DNA sequence tag confirmed that this sequence is a target site for FL regulation. Using Saccharomyces cerevisiae as an experimental system, I demonstrated that the C-terminal portion of FL functions in transcriptional activation. Microarray analysis was performed to study the role of fl in gene regulation on a large scale. mRNA levels in a fl mutant were compared to those in a strain overexpressing the fl gene. Experiments with cDNA microarray containing 13% of the total number of predicted N. crassa genes revealed 122 genes differentially expressed in response to overexpression of fl. Among these, eas displayed the greatest level of response. The cDNA microarray approach also revealed a number of genes that may be indirectly regulated by fl but may be involved in development. This information provides a foundation for further analysis of the role of fl in conidial development.

An understanding of plant pathogen-host interactions is essential to design efficient strategies to control disease in crops. Magnaporthe oryzae, an ascomyceteous fungus and causal agent of rice blast disease, is a model organism to study host-microbe interactions. The overall aim of this dissertation project was to identify genes involved in pathogenicity through the construction and characterization of a random insertional mutagenesis library. In order to saturate the genome with DNA inserts, a collection of >54,000 insertion lines of the M. oryzae strain 70-15 was generated via two transformation methods, PEG/CaCl2 (polyethylene glycol)-mediated protoplast transformation and Agrobacterium tumefaciens-mediated transformation. The first part of this dissertation describes the optimization of both transformation approaches, compares their efficiency and provides a description of the high-throughput processing and phenotypic analysis of the insertion lines. An in vitro appressorium assay of 12,000 T-DNA insertion strains allowed the identification of 135 lines that were classified as morphologically or functionally different than wild-type. Rice infection assays demonstrated that 112 of these strains exhibited defects in pathogenicity.The second part of this dissertation project analyzed the T-DNA integration patterns in a subset of pathogenicity mutants. This section aimed to identify the disrupted genes via recovery of M. oryzae sequences adjacent to the sites of T-DNA insertion. Genomic mapping of 61 T-DNA insertions in pathogenicity mutants via rescuing M. oryzae chromosomal T-DNA flanking sequences using inverse PCR resulted in the identification of 22 conserved hypothetical genes with predicted function, 11 predicted open reading frames without a GenBank significant match, two unannotated regions of the genome assembly and seven intergenic regions. The final part of this dissertation describes the characterization of a M. oryzae pathogenicity mutant that contains a T-DNA insertion in the upstream region of two divergently transcribed genes that encode the vacuolar type-ATPase subunit c`` and the general transcription factor TFIIA subunit γ. Genetic complementation demonstrated the insertion of the T-DNA in the promoter region of the general transcription factor TFIIA subunit γ is responsible for observed defects in conidiation, appressorium morphogenesis, and appressorium function. This is the first report relating the function of TFIIA subunit γ to pathogenicity.

The current study aimed to determine the selective advantage of lateral amygdala (LA) neurons overexpressing the transcription factor CREB that enables their preferential incorporation into the fear memory trace. I hypothesized that overexpression of CREB drives the formation of dendritic spines, potentially providing these neurons with greater connectivity to sensory inputs at the time of learning. Using viral-mediated gene transfer, CREB tagged with GFP, or GFP as a control, was overexpressed in the LA of wild-type mice. Spine number and morphology were compared in homecage mice at the time when mice are normally trained in fear conditioning. Spine density was increased in neurons with CREB vector compared to neurons with GFP vector whereas spine head diameter and length was not different. Therefore, LA neurons overexpressing CREB have increased spine number at the time of learning, potentially providing these neurons with a selective advantage for incorporation into the fear memory trace.