ABSTRACT OF THE DISSERTATION

The Role of P300-Dependent MYC Acetylation in MYC Functions

by

Marina Vorontchikhina

Doctor of Philosophy, Graduate Program in Cell, Molecular and Developmental Biology

University of California, Riverside, December 2011

Dr. Ernest Martinez, Chairperson

The MYC oncoprotein regulates transcription of a multitude of downstream target genes triggering various biological outcomes, such as the induction of cellular proliferation, apoptosis, and oncogenic transformation. While MYC protein levels and activity are tightly controlled in normal cells, MYC is deregulated in most human malignancies. Since cancer is one of the leading causes of death worldwide, it is vital to elucidate the molecular and biochemical mechanisms underlying the modification and regulation of the MYC protein, whose overexpression contributes to the development of most malignant tumors. Posttranslational modifications are implicated in the regulation of MYC stability and function. For instance, several co-activators/histone acetyltransferases (HATs) have been shown to bind and acetylate the MYC protein affecting its turnover by the proteasome. Co-activator/HAT p300 interacts with MYC increasing its stability and transactivation functions. However, once p300 acetylates the oncoprotein at seven lysine residues, MYC becomes more unstable due to induced degradation via the proteasome pathway. While MYC acetylation has been established, the role of acetylation in MYC biological functions has not been determined. Here, I report that by using site-directed mutagenesis I have identified lysine 158 in MYC as the major residue acetylated by the p300. I also demonstrate that acetylation of K158 reduces MYC-activated apoptosis which could be related to MYC-dependent regulation of certain pro-apoptotic genes associated with mitochondrial function, as it was shown by Real-time quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR) analysis. Furthermore, I demonstrate that MYC transcriptionally activates p300 in mammalian cells and, by using Luciferase reporter assays, I further show that MYC and co-activator p300 synergistically activate the promoter of the Cyclin D2 (CCND2), a well-known cell cycle regulatory gene. Moreover, by utilizing immunoprecipitation methods, I establish a link between MYC acetylation by p300 and its interaction with the co-activator/HAT, TIP60. Other cell and molecular biology procedures, such as immunofluorescence and RNA interference, were used in this study as well. These findings begin to uncover a role of co-activator/HAT p300 in MYC biological functions and are important because they suggest potential new targets for the treatment of MYC-dependent cancers.

The events leading to the development of cancer often involve a series of sequential molecular derailments. The complexities of these derailments are frequently specific not only to the type of cancer, but can also be specific to an individual tumor. However, despite these challenges, investigations characterizing the global or universal aberrations seen in human cancers have provided many effective therapeutic strategies to the combat human malignancies. The current dissertation focuses on two neglected areas of study in the cancer biology field, the study of lncRNA isoform-specific alterations and the post-translational modifications of the critically important MYC oncoprotein in cancer.

In the beginning chapters of this dissertation, we explore isoform-specific alterations in a subtype of renal cell carcinomas, known as clear cell renal cell carcinoma (ccRCC). ccRCC is one of the most prevalent cancers within the United States, and can be particularly difficult to treat with conventional therapies. As such, new therapeutics strategies are needed to treat ccRCC in its later stages of the disease. ccRCC has been shown to have severe aberrant RNA production and processing, lending itself as a prime candidate to explore isoform-specific alterations. Furthermore, using new computational methods, we identified previously uncharacterized events of differential transcript and usage in ccRCC implicating several novel genes in the pathology. Discovered within these transcriptomic analyses was a long non-coding RNA, referred to as HOXA Transcript Antisense RNA, Myeloid-Specific 1 (HOTAIRM1), which was found to be specifically downregulated in ccRCC and regulates key genes involved in the hypoxia pathway. In chapter 3, we investigate HOTAIRM1 further examining its function in ccRCC and its role in kidney cell differentiation and maintenance.

In the last two chapters, we discuss and evaluate the effects of two disparate levels of MYC regulation. First, we discuss the complex nature of the interplay between MYC expression and numerous lncRNAs, in what is referred to as the lncRNA-MYC network. In this review, we reveal the breadth of the complexities of how MYC is regulated by lncRNAs and how MYC regulates its oncogenic function through the use of lncRNAs. Finally, with the development of innovative transgenic cell lines, expressing different mutant forms of MYC, we demonstrate that different acetylation states of MYC can have gene-selective effects and consequently alter different molecular pathways.

MYC is a transcription factor that is overexpressed in a wide range of human cancers, and MYC is known to play a role in many important biological processes such as cell death and cell proliferation. Heterodimerizing with the obligatory partner protein, MAX, the MYC and MAX dimer binds a specific DNA sequence, called an E box (CACGTG), through its basic helix-loop-helix zipper domain. MYC N-terminal transcription activation domain (TAD) has been shown to interact with proteins that are potential regulators of MYC transactivation and transformation activities. Among them, the TRRAP protein has been shown to contribute to the transformation activity of MYC through associations with the MYC box (MB) II and MBI-containing sequences. Importantly, TRRAP is also a subunit of several multiprotein complexes that possess histone acetyltransferase (HAT) activity, such as the STAGA (SPT-TAF-GCN5 Acetylase) coactivator complex. GCN5, another subunit of STAGA, is a HAT that acetylates lysine residues in histone H3 and non-histone proteins. MYC has been also identified as a target of GCN5, although the functional consequences of MYC acetylation by GCN5 remain to be investigated. In this work, we identify MYC K323 as the major site of acetylation by GCN5 and then demonstrate that GCN5 stabilizes MYC in a K323- and HAT domain-dependent manner. Interestingly, we also find that GCN5 enhances MYC ubiquitination in a HAT domain/activity-independent manner. Subsequently, we investigate which subunits of the STAGA complex directly interact with MYC, and then analyze the role of the MYC-STAGA interactions in the biological functions of MYC. We identify GCN5 as another STAGA subunit that directly associates with MYC. The direct interaction of MYC with GCN5 facilitates the interaction of MYC with the STAGA complex, acetylation of MYC, binding of MYC to the human TERT promoter, and activation of human TERT transcription. Lastly, we unexpectedly find a physical interaction between MYC and the tumor suppressor p53 that has not been reported previously, despite decades of intense research investigation on their functions. Based on our preliminary results, we suggest a prospective model in which MYC and p53 influence each other`s transcription regulatory activities via a direct physical interaction. These results further deepen our understanding of MYC biological functions and its role in cancer, and could eventually lead to the development of new therapeutic targets against cancer.

Deregulation of the MYC oncoprotein is one of the most prevalent alterations leading to cancer in humans. MYC is known to bind and interact with various transcription cofactors, including histone acetyl-transferases (HATs), and components of the basal transcription machinery influencing transcription initiation and elongation. MYC was previously reported by our laboratory and others to be post-translationally modified via acetylation. However, the functions of MYC acetylation had remained largely unclear. Here, I analyzed the possible gene regulatory and oncogenic cell transforming functions of MYC acetylation at three main lysine (K) residues K149, K158 and K323. By overexpressing MYC in Rat1a fibroblasts, I demonstrated that acetylation-defective MYC lysine(K)-to-arginine(R) mutants are defective in cell transformation and prevent MYC induction of anchorage-independent cell proliferation, a hallmark of cancer cells. Upon genome-wide identification of genes deregulated by the three MYC R-mutants, I confirmed by RT-qPCR that substitution of the acetylated K residues alters the expression of only a small number of specific MYC target genes. These results underly the functional importance of MYC acetylation and its potential as targets for treatment of cancer. Histone acetyltransferase (HAT) complexes are involved in the activation and transcription of the human genome and can be used as a prognostic indicator of cancers. Increased presence of the proteins that compose these complexes are a sign of cancerous cells, one of which is the YEATS2 protein. The YEATS2 protein is a previously characterized scaffolding and histone reader subunit of the ADA2a-containing (ATAC) complex. However, the functions of YEATS2 are still poorly understood and most studies so far have concentrated in cancer cells. Here, I analyzed the functions of the Yaf9, ENL, AF9, Taf14, Sas5 (YEATS) domain and the C-terminal histone fold domain of YEATS2 on gene regulation and cell growth and differentiation of non-malignant (non-cancer) cells. I found that loss of function of either domain inhibits cell growth/proliferation. Knockdown of YEATS2 in HEK293 cells and genome-wide RNA-sequencing analyses identified the ribosomal protein genes as a gene set activated by YEATS2. I further found that activation of the ribosomal protein genes requires both the YEATS and C-terminal histone fold domains of YEATS2. Finally, I showed that YEATS2 protein expression is lost during retinoic acid induced differentiation of proliferating pluripotent embryonic stem cells to neuronal precursor cells in vitro. Together, these results suggest a key role of YEATS2 as a component of the ATAC complex in normal cell growth and proliferation in part via the regulation of the translation machinery. These results expand our knowledge of ATAC complex function in non-tumor cells and give new insights into the functional domains of YEATS2.

The role of non-coding RNAs within the cell is an emerging topic that is currently in full force and gains in this field have added increased depth to our understanding of the central dogma of molecular biology. Non-coding RNAs are able to perform functions much like that of and in conjunction with proteins; and their ability to bind to proteins and nucleic acids provides them with powerful epigenetic regulatory capabilities that either enhance or silence gene expression to determine cell fate. One such process which is highly precise and essential to life is the coordinated and cell-type specific expression of the Hox cluster during early embryonic stem cell differentiation. These Homeobox (Hox) genes are responsible for the formation of cell type boundaries, making them essential to multicellular life. Here, I will show a non-coding RNA, Hotairm1, is involved in the regulation of an essential component of early embryonic stem cell differentiation: the expression of the Hoxa cluster; and implicate it in other major pathways as well as deregulatory processes that lead to various cancers.

Deregulated MYC expression is associated with most types of cancer in humans. MYC deregulation may occur via alterations of the MYC gene itself or via aberrant activation of upstream signaling pathways that activate MYC gene expression and/or increase MYC protein stability via post-translational modifications. Here, I show that MYC is endogenously acetylated at lysine (K) residues K149, K158, and K323 through the characterization of three corresponding novel acetyl-c-MYC antibodies. Having established the existence of endogenously acetylated MYC in cancer cells, this strongly implies a link between such modifications and the regulation of MYC biological functions. Therefore, I test for the importance of MYC K residues acetylated by p300/CBP and GCN5 by analyzing the functional properties of three acetylation-defective MYC K to arginine (R) mutants (e.g. K149R, K158R and K323R). The assays I use highlight MYC’s regulation of cellular proliferation, metabolism, apoptosis and oncogenic transformation in a well-established Rat1a fibroblast system that constitutively overexpresses MYC wild type (WT) and the MYC mutants. With this system I show that MYC K149R and K158R mutants are defective in oncogenic transformation, as indicated by the reduction of anchorage-independent growth of cells in soft agar and reduced tumor formation in athymic nude mice. I also demonstrate that the MYC K158R mutant (a site that would be preferentially modified by p300) is less proliferative and forms fewer foci after cells reach confluence, maintains stronger cellular adhesions compared to MYC WT and the other MYC mutants, and is more apoptotic than MYC WT under conditions of serum starvation. In comparison, the MYC K323R mutant (a site that would be preferentially modified by GCN5) does not exhibit any of these defects, but does appear to affect cellular metabolism, as cells expressing this mutant hastily acidify the culture medium. Finally, by means of RNA sequencing, I show that different sets of MYC-regulated genes are affected (both up- or down-regulated) by the K to R MYC mutants. This data strongly implies that MYC is capable of differential regulation of discrete subsets of genes, via gene-selective mechanisms involving K residues that are acetylated by p300 or GCN5.

ABSTRACT OF THE DISSERTATION

Mechanisms of Core Promoter Sequence-dependent RNA Polymerase II Transcription

by

Muyu Xu

Doctor of Philosophy, Graduate Program in Biochemistry and Molecular Biology

University of California, Riverside, December 2012

Dr. Ernest Martinez, Chairperson

The TATA-box, Initiator (INR), Downstream Promoter Element (DPE), Motif Ten Element (MTE), TFIIB Recognition Element (BRE) and the other core promoter elements contribute to the diverse architecture of core promoters and are paramount for transcriptional activation. Diverse core promoters can communicate with enhancer-bound activators to contribute in the second level gene expression regulation. However, the mechanisms of transcription initiation catalyzed by different core promoters remain unknown and/or controversial. Because TFIID and TFIIB bind most of the core promoter elements, many scientists believe that different core promoters are regulated by the same set of general transcription factors. In contrast, other scientists including us insist that additional core promoter sequence-specific transcription factors besides the general transcription machinery are required to regulate transcription from different core promoters. Several lines of evidence support this hypothesis: a TAFs and Initiator dependent Cofactor 1 (TIC1) fraction requires for the TATA/INR synergy; a TIC2 fraction supports TATA-less core promoter directed transcription; CK2, PC4 and Mediator facilitate the Sp1-activated INR/DPE transcription in mammalian system and NC2 mediates the INR/DPE transcription in Drosophila system. Here, we further purify the TIC1 fraction and identified HMGA1 and Mediator as the effective components that support TATA/INR synergy in vitro. In addition, we also verify the TATA/INR specific role of HMGA1 in mammalian cells. Furthermore, we demonstrate HMGA1 interacts with TFIID and Mediator, and the acidic COOH-tail of HMGA1 is required but not sufficient for HMGA1 to interact with TFIID and Mediator. Accordingly, HMGA1 COOH-tail is also required to support the maximal transcriptional synergy between the TATA-box and the INR. Finally, analysis of activated transcription by Gal4-fusion activators and the beta-Actin gene (ACTB) promoter upstream activating sequences further demonstrates that preferential communication between activators and core promoters contributes to the gene expression regulation. Amazingly, a strict TATA-specificificity by the ACTB upstream activating sequences is revealed for the first time.

One of the most heavily researched oncoproteins is the MYC oncoprotein and this oncogenic transcription factor can regulate a large fraction of cellular genes. MYC overexpression becomes oncogenic and has been shown to be observed in many types of human cancers. GCN5, a histone acetyltransferase (HAT), can acetylate MYC via the STAGA complex at K323, but MYC can also be acetylated by different HATs. For example, p300, another HAT, can preferentially acetylate MYC at K149 and K158. However, p300-mediated acetylation increases MYC turnover, whereas GCN5-mediated acetylation stabilizes MYC. However, the significance of MYC acetylation and the functions of these three acetylated residues have yet to be elucidated. In vertebrates, there is a paralogous gene called PCAF, which codes for an enzyme much like GCN5. Due to their almost nearly identical sequences, it is thought that PCAF and GCN5 have largely redundant functions. PCAF has been reported to have ubiquitin E3 ligase activity within the N-terminal PCAF homology domain and this domain is highly conserved in GCN5. However, it is unknown whether GCN5 has any intrinsic E3 ligase activity or whether it recruits an E3 ligase for its surprising role in ubiquitination. Here we show that these three main lysine residues of MYC are reversibly acetylated in various immortalized non-tumorigenic and malignant cells. Furthermore, acetylation of MYC at different lysine residues differentially affects its stability in a cell type-dependent manner and all acetylated lysine (AcK) residues are required for anchorage independent growth in vitro. We show that each individual AcK site has gene-specific functions in controlling select MYC-regulated processes and are required for the oncogenic transformation of MYC. In this study, we have also reported that not only can GCN5 directly acetylate MYC at K323, but it can also stimulate MYC ubiquitination in a GCN5-dependent manner. Furthermore, the PCAF homology domain of GCN5, described earlier, does not contribute to GCN5-dependent MYC ubiquitination, but a direct interaction between MYC and GCN5 is required for the stimulation of MYC ubiquitination. Altogether, we have established a novel and unexpected function of GCN5, and we propose a new model for the regulation of MYC by HATs. MYC binds to HATs, brings them to promoter DNA and recruits components of the preinitiation complex, and then becomes ubiquitinated to allow release of RNAPII for transcription elongation and allow for a new round of transcription. This model suggests that HAT complexes and associated acetylation or E3 ligase pathways could eventually lead to the development of new therapeutic targets to inhibit MYC-dependent cancers.