Promoter hypermethylation mediated gene silencing is a frequent epigenetic finding in many cancers that affects genes known to have important roles in several aspects of cell biology. Hematological malignancies have been reported to harbor multiple genes aberrantly silenced by promoter hypermethylation and as a result, cytosine analogs known to inhibit the DNA methylation machinery are currently being evaluated in clinical trials. As such, the general goal of this thesis was to identify genes silenced by promoter hypermethylation in human acute myeloid leukemia (AML) and to study the mechanism of promoter hypermethylation mediated gene silencing. Interestingly, the cyclin dependent kinase inhibitor p15 was found to be methylated at a high frequency in AML patients and cell lines in association with a lack of detectable p15 mRNA. Treatment with the cytosine analog 5-Aza-2-deoxycytidine (5-Aza-dC) in vitro resulted in promoter demethylation and p15 mRNA re-expression, which was associated with a release of a transcriptionally repressive complex at the p15 promoter. Importantly, 5-Aza-dC treatment also reversed specific histone amino-terminal modifications at the p15 promoter which are normally associated with transcriptionally inactive chromatin regions, implicating chromatin remodeling in promoter hypermethylation mediated gene silencing. The recently discovered DNA methylation inhibitor, zebularine considered more stable than 5-Aza-dC was also able to reconstitute p15 mRNA in vitro in association with promoter demethylation, regional enrichment of histone acetylation, and growth inhibition. To identify novel genes silenced by promoter hypermethylation in AML, cDNA microarray analysis was employed following in vitro pharmacological inhibition of DNA methylation and histone deacetylation. Of note, four genes from the metallothionein family of cysteine rich small molecules were consistently upregulated following drug treatment and further evaluation identified the gene MT1H to be hypermethylated at a high frequency in AML patients and cell lines. Taken together, the data suggests that aberrant promoter hypermethylation mediated gene silencing occurs in multiple genes from different gene families during the molecular pathogenesis of human AML. Furthermore, the mechanism of promoter methylation mediated transcriptional silencing acts in concert with specific histone modifications which, importantly, can be reversed by treatment with pharmacological inhibitors of DNA methylation.

The SRC gene encodes pp60c-Src, a 60 kDa non-receptor tyrosine kinase that is frequently activated and/or overexpressed in many cancers including colon cancer. In a subset of colon cancer cell lines, it has been shown, that the overexpression of c-Src can be explained, in part, by the transcriptional activation of the SRC gene. As a result, the general goal of this thesis was to further characterize how SRC is transcriptionally regulated in human cancer cell lines. Two highly dissimilar promoters, the housekeeping-like SRC1A promoter, as well as the HIF-1Ñ regulated tissue-specific SRC1Ñ promoter, regulate SRC expression. hnRNP K and the Sp family of factors regulate the SRC1A promoter; however, the true impact of Sp3 on SRC1A activity was not understood. In this thesis, a comprehensive analysis of the effect of Sp3 on SRC1A activity was performed. Physiologically, Sp3 exists as four translational isoforms that, in part, dictate the activation potential of Sp3. In general, the longer forms of Sp3 were modest transcriptional activators of the SRC1A promoter whereas the shorter forms were unable to activate the SRC1A promoter. An analysis of all Sp3 isoforms identified that the shorter Sp3 isoforms could be converted into transcriptional activators of SRC1A if the SUMOylation of a critical lysine residue within the inhibitory domain was prevented. Conversely, SUMOylation of the same isoform had little effect on the activation potential of the longer Sp3 isoforms at the SRC1A promoter. These results suggest that transcriptional activation by Sp3 is promoter context-, isoform- and modification-dependent.<p>SRC is transcriptionally repressed by histone deacetylase inhibitors (HDIs) and despite unsuccessful studies attempting to identify HDI-responsive elements within the SRC promoter regions none could be identified. This finding also suggests that histone deacetylases (HDACs) may be required for SRC expression. Historically, it was believed that HDIs act at the histone level to alter chromatin dynamics through the inactivation of HDACs to result in histone hyperacetylation and increased transcriptional activation. As such, a systematic investigation of the changes in histone H3 and H4 acetylation status at the transcriptionally repressed SRC promoter regions and the transcriptionally activated p21WAF1 promoter region was performed. The p21WAF1 promoter was used as control in this study as p21WAF1 is a classical example of a gene transcriptionally activated by HDIs. Interestingly, similar changes in histone acetylation at the p21WAF1 promoter and both SRC promoter regions were observed. Upon closer examination of acetylation changes at discreet histone residues, it was observed that in the rare case that a particular residue was differentially acetylated upon treatment at the promoter regions analyzed, the SRC1Ñ and p21WAF1 promoter regions demonstrated more similar changes in acetylation as compared to SRC1A. Taken together, these results suggest that histone acetylation status is not an accurate indicator of transcriptional activity following HDI treatment. To further investigate HDI-mediated SRC repression, RNA Pol. II occupancy at the promoter and regions downstream of the promoter were assessed. Despite the continued occupancy of RNA Pol. II at the promoter regions, RNA Pol. II was lost from the 3¡¦ UTR upon treatment with HDIs. These findings suggest that RNA Pol. II . may be sequestered at the promoter regions upon treatment with HDIs possibly as a result of impeded transcription initiation and/or elongation. Further analysis of the phosphorylation status of RNA Pol. II identified that transcriptional initiation was indeed occurring despite HDI treatment; however, productive transcriptional elongation could not be confirmed thus suggesting a role for abrogated elongation in HDI mediated SRC repression. Complimentary analysis of the effects of HDACs on SRC expression suggested that while class I HDACs abrogated SRC expression, class II HDACs were required for the maintenance of SRC transcript levels in a promoter-independent fashion. Together, these results provide the basis for a model whereby HDIs repress SRC transcriptional expression through the inhibition of class II HDAC activity to eventually result in curtailed SRC transcriptional elongation.

The post-translational modifications (PTMs) of histones regulate many cellular processes including transcription, replication, DNA repair, recombination, and chromosome segregation. A large number of combinations of PTMs are possible, with methylation being one of the most complex, since it is found in three states and is recognized in a sequence specific context. Methylation of histones at key lysine residues has been shown to work in concert with other modifications to provide a Histone Code that may determine heritable transcriptional conditions in normal and disease states. On the most basic level it is pivotal to understand how and by which proteins the numerous PTMs are recognized, as well as mechanisms for downstream signal propagation. To address this need we developed a high-throughput method that allows analysis of up to 600 PTMs in a single experiment. This approach was utilized to characterize macromolecules interacting with the specific modifications on histone tails and to screen for the marks that bound to Malignant Brain Tumor (MBT) proteins, important chromatin regulators implicated in cancer. All MBTs recognized either mono- or dimethyllysine histone marks, and using structure-based mutants we identified a triad of residues that were responsible for this discrimination. These results provide the foundation for the rational design of highly selective MBT inhibitors. Additionally, this thesis describes combinatorial recognition of histone modifications, as proposed in the original Histone Code hypothesis. We demonstrate that Tudor domains of UHRF1, a protein involved in epigenetic maintenance of DNA methylation, is able to read a dual modification state of histone H3 in which it is trimethylated at lysine 9 and unmodified at lysine 4. This study provides an elegant example of the combinatorial readout of histone modification states by a single domain. Together, our findings offer mechanistic insights into the recognition of methylated histone tails by MBT domains and Royal Family in general.

H2A.Z is an essential histone variant that has multiple chromosomal functions. One such role is transcriptional regulation. However, its role in this process is complex since it has been reported to function both as a repressor and activator. Earlier work in our lab showed that H2A.Z can be post-translationally modified with monoubiquitin (H2A.Zub1) and this form of H2A.Z is linked to transcriptional silencing. We further predicted that changes in the H2A.Z ubiquitylation status directly modulated its function in transcription. Furthermore, H2A.Z-containing nucleosomes possess a unique set of post-translational modifications (PTMs), compared to H2A nucleosomes, many of which are linked to transcriptional activation. The central aim of this thesis was to characterize the role of PTMs on H2A.Z nucleosomes in transcriptional regulation. To this end, I have provided the first evidence linking H2A.Z deubiquitylation to transcriptional activation. I demonstrated that ubiquitin specific protease 10 (USP10) is a deubiquitylase that targets H2A.Z in vitro and in vivo. Moreover, I found that both H2A.Z and USP10 are required for activation of androgen-receptor (AR)-regulated genes, and that USP10 regulates the levels of H2A.Zub1 at these genes. To understand how H2A.Z engages downstream effector proteins, in the nucleosome context, we used a mass spectrometry approach to identify H2A.Z-nucleosome-interacting proteins. Many of the identified proteins contained conserved structural motifs that bind post-translationally modified histones. For example, we found that Brd2 contains tandem bromodomains that engage H2A.Z nucleosomes through acetylated H4 residues. To investigate the biological relevance of this interaction, I present evidence that Brd2 is recruited to AR-regulated genes in a manner dependent on H2A.Z and the bromodomains of Brd2. Consistent with this observation, chemical inhibition of Brd2 recruitment greatly inhibited AR-regulated gene expression. Collectively, these studies have defined how H2A.Z mediates transcriptional regulation through multiple mechanisms and pathways.

In Saccharomyces cerevisiae, Set1p-mediated deposition of trimethylation on lysine 4 of histone H3 is a histone modification often associated with active transcription. Recently, it was discovered that members of the Jumonji family of proteins have the enzymatic ability to remove methylation on histone lysine residues. Here, I describe the function of the yeast Jumonji protein Jhd2p, the only yeast Jumonji with known demethylase activity towards histone H3 lysine 4 methylation. I find that during the development program of yeast sporulation, Jhd2p is responsible for demethylating lysine 4 on a global scale. Further, ChIP analysis examining lysine 4 methylation levels reveals that genes whose expression is dependent on JHD2 during sporulation are subject to what appears to be Jhd2p-mediated demethylation. Additionally, synthetic dosage lethality screens performed to identify genetic interactors of Jhd2p revealed that Jhd2p is a likely component of mitochondrial retrograde signaling, working alongside the transcription factors Rtg1p/Rtg3p.

In Saccharomyces cerevisiae, Set1p-mediated deposition of trimethylation on lysine 4 of histone H3 is a histone modification often associated with active transcription. Recently, it was discovered that members of the Jumonji family of proteins have the enzymatic ability to remove methylation on histone lysine residues. Here, I describe the function of the yeast Jumonji protein Jhd2p, the only yeast Jumonji with known demethylase activity towards histone H3 lysine 4 methylation. I find that during the development program of yeast sporulation, Jhd2p is responsible for demethylating lysine 4 on a global scale. Further, ChIP analysis examining lysine 4 methylation levels reveals that genes whose expression is dependent on JHD2 during sporulation are subject to what appears to be Jhd2p-mediated demethylation. Additionally, synthetic dosage lethality screens performed to identify genetic interactors of Jhd2p revealed that Jhd2p is a likely component of mitochondrial retrograde signaling, working alongside the transcription factors Rtg1p/Rtg3p.

Histone acetyltransferases (HATs) are important enzymes in transcriptional control and potential targets for chemotherapeutic intervention in malignant diseases. Among different HAT members, the yeast Esa1 and human Tip60 (the HIV-1 Tat interactive protein, 60KDa) play multiple roles in normal cellular processes including transcription, cell cycle and checkpoint machinery, double strand DNA break repair, apoptosis, and cell cycle progression. Tip60 is also implicated in several human diseases such as prostate cancer, and gastric cancer. These studies suggest that Tip60 is a potential therapeutic target for new cancer treatment. So, we designed experimental work to synthesize and investigate organic inhibitors of Tip60 using different strategies, including substrate analogs, small molecule screening, and modification of the natural product anacardic acid. These studies provide important chemical agents for basic biology research of HAT function, and produce potential lead compounds for future pharmacologic intervention of HAT deregulation in cancer. Currently, of the methods used for the measurement of acetyltransferase activities, many comprise tedious separation procedures and involve enzyme-coupled steps or radioactive materials. These shortcomings have limited their applications in high-throughput screening (HTS) of HAT inhibitors. To circumvent these problems, a homogenous fluorescent HAT assay based on engineered H4 peptide was designed, synthesized, and evaluated. The data showed that these fluorescent reporters can be used to detect the acetyltransferase activities.

The post-translational modifications (PTMs) of histones regulate many cellular processes including transcription, replication, DNA repair, recombination, and chromosome segregation. A large number of combinations of PTMs are possible, with methylation being one of the most complex, since it is found in three states and is recognized in a sequence specific context. Methylation of histones at key lysine residues has been shown to work in concert with other modifications to provide a Histone Code that may determine heritable transcriptional conditions in normal and disease states. On the most basic level it is pivotal to understand how and by which proteins the numerous PTMs are recognized, as well as mechanisms for downstream signal propagation. To address this need we developed a high-throughput method that allows analysis of up to 600 PTMs in a single experiment. This approach was utilized to characterize macromolecules interacting with the specific modifications on histone tails and to screen for the marks that bound to Malignant Brain Tumor (MBT) proteins, important chromatin regulators implicated in cancer. All MBTs recognized either mono- or dimethyllysine histone marks, and using structure-based mutants we identified a triad of residues that were responsible for this discrimination. These results provide the foundation for the rational design of highly selective MBT inhibitors. Additionally, this thesis describes combinatorial recognition of histone modifications, as proposed in the original Histone Code hypothesis. We demonstrate that Tudor domains of UHRF1, a protein involved in epigenetic maintenance of DNA methylation, is able to read a dual modification state of histone H3 in which it is trimethylated at lysine 9 and unmodified at lysine 4. This study provides an elegant example of the combinatorial readout of histone modification states by a single domain. Together, our findings offer mechanistic insights into the recognition of methylated histone tails by MBT domains and Royal Family in general.

The mixed lineage leukemia (MLL) proteins regulate an array of developmental and differentiation processes. Similar to other members of the SET1 family, association of MLL1-4 with Ash2L, RbBP5 and WDR5, collectively termed the MLL core complex, is required for MLL mediated histone H3 Lys-4 di/tri-methylation. Each member of the core complex has a unique role in modulating the activity of MLL1. WDR5 is key in nucleating the formation of the core complex by acting as a structural scaffold, whereas Ash2L and RbBP5 are responsible for stimulating MLL methyltransferase activity. Currently, the structural and biochemical mechanisms utilized by the core complex to regulate MLL1 activity are unknown. Through structural and biochemical dissection of the core complex we have assigned specific functions to core complex subunits and have identified the minimal structural requirements for methyltransferase activity. Furthermore, through structure based drug design, we have identified a peptidomimetic inhibitor of MLL1 methyltransferase activity.

Protein arginine methyltransferase 1 (PRMT1) is a key posttranslational modification enzyme that catalyzes the methylation of specific arginine residues in histone and nonhistone protein substrates, regulating diverse cellular processes such as transcriptional initiation, RNA splicing, DNA repair, and signal transduction. Recently the essential roles of PRMT1 in cancer and cardiovascular complications have intrigued much attention. Developing effective PRMT inhibitors therefore is of significant therapeutic value. The research on PRMT inhibitor development however is greatly hindered by poor understanding of the biochemical basis of protein arginine methylation and lack of effective assays for PRMT1 inhibitor screening. Herein, we report our effort in the kinetic mechanism study as well as the fluorescent probe and inhibitor development for PRMT1. New fluorescent reporters were designed and applied to perform single-step analysis of substrate binding and methylation of PRMT1. Using these reporters, we performed transient-state fluorescence measurements to dissect the rate constants along the PRMT1 catalytic coordinate. The data give evidence that the chemistry of methyl transfer is the major rate-limiting step, and that binding of the cofactor SAM or SAH affects the association and dissociation of H4 with PRMT1. Importantly, we identified a critical kinetic step suggesting a precatalytic conformational transition induced by substrate binding. On the other hand, we discovered a type of naphthyl-sulfo (NS) compounds that block PRMT1- mediated arginine methylation at micromolar potency through a unique mechanism: they directly target the substrates but not PRMT enzymes for the observed inhibition. We also found that suramin, an anti-parasite and anti-cancer drug bearing similar functional groups, effectively inhibited PRMT1 mediated methylation. These findings about novel PRMT inhibitors and their unique inhibition mechanism provide a new way for chemical regulation of protein arginine methylation. Addionally, to dissect the interplaying relationship between different histone modification marks, we investigated how individual lysine acetylations and their different combinations at the H4 tail affect Arg-3 methylation in cis. Our data reveal that the effect of lysine acetylation on arginine methylation depends on the site of acetylation and the type of methylation. While certain acetylations present a repressive impact on PRMT-1 mediated methylation (type I methylation), lysine acetylation generally is correlated with enhanced methylation by PRMT5 (type II dimethylation). In particular, Lys-5 acetylation decreases activity of PRMT1 but increases that of PRMT5. Furthermore, hyperacetylation increases the content of ordered secondary structures of H4 tail. These findings provide new insights into the regulatory mechanism of Arg-3 methylation by H4 acetylation, and unravel that complex intercommunications exist between different posttranslational marks in cis.