UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Resistance to BET inhibitors in lung adenocarcinoma is mediated through casein kinase 2 phosphorylation… Calder, Jack 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_may_calder_jack.pdf [ 7.7MB ]
Metadata
JSON: 24-1.0365542.json
JSON-LD: 24-1.0365542-ld.json
RDF/XML (Pretty): 24-1.0365542-rdf.xml
RDF/JSON: 24-1.0365542-rdf.json
Turtle: 24-1.0365542-turtle.txt
N-Triples: 24-1.0365542-rdf-ntriples.txt
Original Record: 24-1.0365542-source.json
Full Text
24-1.0365542-fulltext.txt
Citation
24-1.0365542.ris

Full Text

RESISTANCE TO BET INHIBITORS IN LUNG ADENOCARCINOMA IS MEDIATED BY CASEIN KINASE 2 PHOSPHORYLATION OF BRD4    by  Jack Calder  B.Sc., The University of Victoria, Dec 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pathology and Laboratory Medicine) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2018  ©Jack Calder, 2018 ii		Abstract  Lung cancer is the leading cause of cancer related death in both men and women worldwide, mainly due to the lack of effective therapies. The development of specific chemical compounds that target epigenetic post-translational modifications has recently emerged as an excellent approach for validating new treatment strategies for diseases that have complex underlying mechanisms. JQ1 is a small-molecule inhibitor of the bromodomain and extraterminal (BET) family proteins, which function as important reader molecules of acetylated histones and recruit transcriptional activators to specific promoter sites. In many cancer lines the down-regulation of MYC, a known oncogenic transcription factor and contributor to the pathogenesis in certain cancer types, has been linked to BET inhibitor (BETi) treatment. In addition, resistance to BETis has only been examined in MYC-dependent cancers, with all forms of resistance involving re-expression of MYC, through several mechanisms. Previously, our lab has shown that lung adenocarcinoma (LAC) cells are inhibited by JQ1 through a mechanism independent of MYC down-regulation, identifying FOSL1 as a mediator of response. This suggests that the epigenetic landscape of cells from different origins and differentiation states influences response to JQ1. Therefore, I aim to investigate how LAC cells, independent of MYC down-regulation, acquire resistance to BET inhibition, to elucidate mechanisms of primary resistance and potential treatment strategies for LAC. Here, I establish resistance in two JQ1 sensitive LAC cell lines and demonstrate that MYC levels were not significantly altered, nor was FOSL1 expression reactivated in resistant lines, indicating a novel mechanism of resistance. Interestingly, resistant lines were still dependent on the BET protein BRD4, as demonstrated by siRNA knockdown, iii		suggesting that BRD4 may drive resistance through regulating gene transcription independent of its acetyl-binding domain. Both resistant lines showed increased levels of phosphorylated BRD4, and also up-regulation of casein kinase 2 (CK2), a kinase previously shown to phosphorylate BRD4. Furthermore, combining JQ1 with a CK2 inhibitor showed synergistic effects in both resistant lines, with treatment leading to decreased levels of pBRD4. Overall, we have determined that LAC cells develop JQ1 resistance through mechanisms independent of MYC, identifying CK2 phosphorylation of BRD4 as a likely mechanism of resistance.                               iv		Lay Summary  Lung cancer is the leading cause of cancer related deaths worldwide due to the advanced stage of diagnosis where there are limited effective therapies. BET inhibitors are a class of small-molecule inhibitors that target a family of epigenetic reader proteins and inhibit their function. Epigenetic readers are proteins that can regulate expression of genes by binding to “marks” on histones, which are proteins that package DNA in a cell. Currently there are several BET inhibitors in clinical trials; however, resistance to inhibitors is quite common and can limit their effectiveness. Here I identify a unique mechanism of resistance in lung cancer and present a potential strategy of combination-based drug therapy to circumvent acquired and primary resistance to BET inhibitors. This work will help to further understand epigenetic readers and mechanisms of resistance to BET inhibitors and potentially lead to better treatment strategies for advanced stages of lung cancer.    v		Preface  The experimental design was developed with guidance and assistance from Dr. William W. Lockwood. Cell line STR profiling was completed by Genetica DNA Laboratories Lab (North Carolina, USA) and RNA expression data was generated by TCAG (The Center for Applied Genomics (Toronto, Canada). Amy Nagelburg initially started culturing resistant cell lines with help from Jenny Wu. All other experiments performed in this study; including RNA data analysis and Compusyn combinatorial analysis, were performed by myself. The writing of this thesis was completed by me with guidance and editing by Dr. William W. Lockwood.  I was partially funded by a Canadian Graduate Scholarship-Master’s (CGS M) through the Canadian Institute of Health Research (CIHR)    vi		Table of Contents Abstract..............................................................................................................................ii Lay Summary....................................................................................................................iv Preface.................................................................................................................................v Table of Contents..............................................................................................................vi List of Tables.....................................................................................................................ix List of Figures.....................................................................................................................x List of Symbols and Abbreviations................................................................................xii Acknowledgements..........................................................................................................xv Dedication........................................................................................................................xvi Chapter 1: Introduction....................................................................................................1 1.1 Lung Cancer...........................................................................................................1  1.1.1 Classification of LC.......................................................................................2 1.1.2 Current treatment strategies in LC.................................................................4 1.2 Epigenetics.............................................................................................................7 1.2.1 Chromatin structure and histone modifications.............................................8 1.2.2 Epigenetic writers, erasers and readers........................................................12 1.3 Bromodomain and extraterminal family proteins................................................17 1.3.1 BRD4...........................................................................................................20 1.4 Bromodomain and extraterminal inhibitors.........................................................23 1.4.1 BETi treatment in cancer.............................................................................25 1.5 Hypothesis and aims............................................................................................28  vii		Chapter 2: Material and Methods..................................................................................29 2.1 Cell lines and media.............................................................................................29 2.2 Establish resistant lines........................................................................................29 2.3 Dose response experiments and long-term growth assays...................................29 2.4 RNA interference (RNAi) experiments using siRNAs........................................30 2.5 CX-4945 and JQ1 combination treatment...........................................................31 2.6 Quantitative analysis of combinational treatment................................................32 2.7 Gene expression profiling and data analysis........................................................33 2.8 Protein extraction and western blot analysis........................................................34 2.9 Antibodies, inhibitors and other reagents............................................................35 Chapter 3: Results............................................................................................................36 3.1 Establishment of isogenic BET inhibitor sensitive and resistant  LAC cell lines......................................................................................................36 3.2 LAC lines acquire resistance to BET inhibitor in a MYC and FOSL1 independent manner.............................................................................................41 3.3 Epithelial-mesenchymal transition (EMT) does not induce resistance to JQ1 in LAC......................................................................................................................43 3.4 Resistant lines are still dependent on BRD4........................................................48 3.5 Increased phosphorylation of BRD4 by casein kinase 2 (CK2) may induce resistance to JQ1..................................................................................................51 3.6 Combinational treatment with a CK2 inhibitor and JQ1 shows synergistic effects in resistant lines....................................................................................................55 3.7 Primary JQ1 resistant LAC line is sensitized through inhibition of CK2............57 viii		3.8 AXL and SPOCK1 identified as possible targets of pBRD4 and mediators of JQ1 resistance......................................................................................................50 Chapter 4: Discussion......................................................................................................64 4.1 LAC presents a unique opportunity to study mechanisms of resistance to BETis.. ............................................................................................................64 4.2 Established resistant lines are resistant to BET inhibition and not off-target or multi-drug effects.................................................................................................66 4.3 Acquired resistance to BETis in two LAC lines is independent of MYC and FOSL1..................................................................................................................68 4.4 EMT does not confer resistance to BETis in LAC..............................................69 4.5 JQ1 resistant cell lines are still dependent on BRD4...........................................71 4.6 Phosphorylation of BRD4 by CK2 is a possible mechanism of acquired resistance in LAC lines........................................................................................72 4.7 CK2 inhibition synergizes with JQ1 treatment in JQ1 sensitive and acquired resistant lines........................................................................................................75 4.8 Synergistic effects are also present in a primary JQ1 resistant LAC line............77 4.9 RNA expression analysis identifies AXL and SPOCK1 as possible targets of pBRD4 in H1975 and H23 resistant lines............................................................78 Chapter 5: Conclusions...................................................................................................81 5.1 Summary of research...........................................................................................81 5.2 Future directions..................................................................................................82 Chapter 6: Extended Figures and Tables......................................................................86 References.........................................................................................................................95 ix		List of Tables Extended Table 6.1  Gene set list of analogously regulated genes..................................93  Extended Table 6.2  Gene set list of conversely regulated genes....................................94                              x		List of Figures Figure 1.1  Incidence of driver oncogenes in non-small cell lung cancer…………..........6 Figure 1.2  Structure of chromatin....................................................................................10 Figure 1.3  Bromodomain tertiary structure......................................................................14 Figure 1.4  Phylogenic-tree of the human bromodomain containing families..................16 Figure 1.5  Domain structure of the BET family proteins................................................18 Figure 1.6  Regulation of transcription by BRD4 is mediated at three different  levels..................................................................................................................................21 Figure 1.7  Chemical structures of the two BET inhibitors, JQ1 and I-BET762..............24 Figure 1.8  Mechanisms of resistance to BETis show dependence on re-expression of MYC..................................................................................................................................27 Figure 3.1  JQ1 resistance established in two JQ1 sensitive lung adenocarcinoma  lines....................................................................................................................................38 Figure 3.2  MYC and FOSL1 show insignificant changes in both resistant lines............42 Figure 3.3  Induction of EMT is not sufficient to induce resistance to JQ1 ....................46 Figure 3.4  JQ1 resistant lines dependencies on BRD4 knockdown mimic that of control lines....................................................................................................................................50 Figure 3.5  pBRD4 and CK2 levels are elevated in both JQ1 resistant lines and may be targeted synergistically to induce death.............................................................................52 Figure 3.6  Synergistic trends are also seen in a primary JQ1 resistant line.....................58 Figure 3.7  RNA expression analysis identifies AXL and SPOCK1 as highly  up-regulated genes in H1975 and H23 resistant lines, respectively.................................62 Extended Figure 6.1  JQ1 resistant lines are permanently changed and are confirmed to be isogenic offspring of parental lines...............................................................................86 Extended Figure 6.2  TGFβ1 treatment does not induce increased resistance to JQ1 treatment in H23 line.........................................................................................................87 Extended Figure 6.3  Both resistant and control lines show synergistic values upon JQ1 and CX-4945 combination treatments...............................................................................88 Extended Figure 6.4  10-day combination growth assays show synergy for H23 control and H1975 resistant and control lines................................................................................89 Extended Figure 6.5  H2030 confirmed to be resistant to BETis....................................90 xi		Figure 6.6  Dose response experiments with Trametinib, Afatinib, Crizotinib and  BEZ-235.............................................................................................................................91 Figure 6.7  Heat-maps of the top 20 deregulated genes in only the resistant lines...........92                                 xii		List of Symbols and Abbreviations  aa  amino acid ABCA3 ATP-binding cassette sub-family A member 3 ALL  Acute lymphoblastic leukemia ALK  ALK receptor tyrosine kinase AKT  Serine/ threonine-specific kinase (also known as Protein kinase B) AML  Acute myeloid leukemia ANOVA Analysis of variance ATCC  American Type Culture Collection BD1/2  Bromodomain  BET  Bromodomain and extra terminal BETi  Bromodomain and extra terminal inhibitor BH  Benjamini-Hochburg BL  Burkitt lymphoma BRD4  Bromodomain-containing protein 4 BRD4-LF Long isoform of BRD4 BRD4-SF Short isoform of BRD4 BSA  Bovine serum albumin ChIP  Chromatin immunoprecipitation CI  Combination index CK2  Casein kinase 2 CK2i  CK2 inhibitor (refers to CX-4945 in this thesis) CK2α  Casein kinase 2 alpha CK2α’  Casein kinase 2 alpha prime Co-IP  Protein complex immunoprecipitation CT  Computed tomography CTD  C-terminal domain CTM  C-terminal motif CV  Crystal violet DMSO  Dimethyl sulfoxide DNA  Deoxyribonucleic acid xiii		EGFR  Epidermal growth factor receptor ET  Extraterminal domain FACS  Fluorescence-activated cell sorting FBS  Fetal bovine serum FOSL1 Fos-related antigen 1 (also known as Fra1) HAT  Histone acetyl-transferase HDAC  Histone deacetylase HRP  horseradish peroxidise IP  Immunoprecipitation KIF11  Kinesin family member 11 KRAS  Part of Ras (GTPase proto-oncogene) subfamily  LAC  Lung adenocarcinoma LC  Lung cancer MLL  Mixed-lineage leukemia MM  Multiple myeloma MYC  Also known as c-MYC NonT  Non-target NRAS  Part of Ras (GTPase proto-oncogene) subfamily  NSCLC Non-small cell lung cancer PDAC  Pancreatic ductal adenocarcinoma P-TEFb Positive transcription elongation factor b PTM  Post-translational modification RIPA  Radioimmunoprecipitation assay RNA  Ribonucleic acid RNAi  RNA interference RNA pol II RNA polymerase II SCLC  Small cell lung cancer siRNA  Small/ short interfering RNA STR  Short-tandem repeat TBS-T  Tris-buffered saline-Tween 20 TCAG  The Center for Applied Genomics xiv		TGFβ1  Transforming growth factor beta 1 TKI  Tyrosine kinase inhibitor TNBC  Triple-negative breast cancer xv		Acknowledgements  I would like to extend my sincerest gratitude to my supervisor Dr. Will Lockwood for his patience, insight, and guidance for my growth as a researcher and as an individual. I am very fortunate to have received the opportunity to work in lung cancer research and with such a dedicated and enthusiastic supervisor, and for this opportunity and mentorship I am eternally grateful. To the members of the Lockwood lab, thank-you for being my safety net to bounce ideas and frustrations off of through this three-year journey of mine and for making late nights in the lab some of the most enjoyable times. In particular I would like to thank Bryant Harbourne and Min Oh for their assistance in various experiments and troubleshooting, and for helping to edit this thesis. I would also like to thank my supervisory committee (Dr. Sam Aparicio, Dr. Gregg Morin and Dr. Phillipp Lange) for their supportive and insightful feedback throughout this and other projects.   Lastly, I would like to thank my family and friends for their continued support and understanding of my sometimes unusual schedule, with particular mention to Dave Rowbotham who was major part of my journey both professionally and personally.   xvi		Dedication  I dedicate this thesis to my Nona. Someone who was taken too soon due to lung cancer having never smoked, and has been in my thoughts many times during my work at BCCRC in tough times and pushed me on.1		Chapter 1: Introduction 1.1 Lung cancer  Lung cancer (LC) is defined as any malignant neoplasm occurring in the trachea, bronchus or lung tissue, and is the leading cause of cancer related deaths in both men and women worldwide [1]. Globally, LC accounts for 19% of all cancer related deaths, more than breast, prostate, and colorectal cancer combined [2]. The median age of diagnosis for LC is 70 years old [3]. This can be attributed to the long latency period (~30 years) between the exposure to tobacco smoke, which is the primary cause of LC, and the development of LC [1]. As such, the incidence of LC often parallels the trends in smoking habits and prevalence of certain populations. For instance, cigarette smoking in Chinese men went from an average of one pack a day in 1952 to an average of 10 packs a day in 1992. Unsurprisingly, LC incidence in Chinese men has steadily increased since 1988 [1].   In 2012 there were estimated to be 1.82 million new cases of LC worldwide, an increase from 1.35 million in 2002 [1]. While the incidence of LC has started to decrease in many first world countries, many second and third world countries show increases in LC incidence, with almost half of all LC cases (49%) occurring in these countries [1]. This is mainly due to the lack of public awareness of the negative health impacts caused by smoking in second and third world countries. However, smoking is not the only cause of LC, as evident in 2000 and 2012, where approximately half of all female lung cancers were not associated with primary consumption of cigarette smoke [1]. Though second-hand smoke likely plays a role in many of these cases, even in people who have never smoked (termed “never smokers” [4]), there is still a large number of LC cases not 2		attributed to smoking at all. Furthermore, certain subtypes of LC are associated with smokers and smoking related cancers and others with never smokers, as will be discussed below.   Although there has been a steady increase in the 5-year survival of cancer patients in many different types of cancer (prostate, breast, colorectal), the 5-year survival rate of LC patients remains low at 18% [2]. This, coupled with a very high incidence rate in both men and women, is the main reason that LC is the leading cause of cancer related deaths worldwide. The main reason for the low survival rate in LC is that patients do not present symptoms until the late or metastatic stages of the disease (stage III and IV), to which there are limited treatment options. If the disease is detected early, the 5-year survival rate is significantly higher at  55% when localized and 27% when regionally contained [3]. However, once the disease reaches an advanced stage and metastasis has occurred the 5-year survival rate drops to only 4% [2].  Unfortunately, the majority of LCs are diagnosed at an advanced stage (57%) with only 16% of cases being diagnosed at the local stage [3].   1.1.1 Classifications of lung cancer LC is histologically classified into two main groups; non-small-cell lung cancer (NSCLC) and small-cell lung cancer (SCLC). The majority of LCs are diagnosed as NSCLC (~85%,) with only 15% being diagnosed as SCLC [5]. As the name suggests, a major difference between these two classifications is the size of the cell, with SCLC comprised of smaller cells as compared to the bigger NSCLC cells. SCLCs are usually located in the central bronchi and are almost exclusively caused by smoking. SCLC cells 3		have a faster doubling time and are more aggressive than NSCLC as reflected through the 5-year survival rates of 7% and 21%, respectively [3].  NSCLC can be sub-classified into squamous cell, adenocarcinoma, and large-cell carcinoma of the lung, based on histology.  Squamous cell carcinoma, which makes up about 30% of LC [6], arise predominantly from epithelial cells of the conducting airways in the lung and form flattened cells with high levels of keratin, similar to the squamous cells of the skin. Adenocarcinomas, the most diagnosed subtype of NSCLC (~40% of LCs) [6], are located more peripherally in the lung and originate from alveolar type II and/or Club cells, which secrete surfactant similar to glandular cells as the name “adeno” implies. Lastly, large cell carcinomas make up about 3% of LCs and are primarily found in the central bronchi. Cancer cells in this subgroup lack any distinguishing features to align them with either adeno or squamous carcinomas. Of these three subtypes, squamous cell carcinomas are the most strongly linked to smoking, while adenocarcinomas are the most commonly linked to LCs of never smokers [5]. This may be due to the location of these cancers as both SCLC and squamous (and large cell) are more centrally located in the major bronchi and are therefore in closer proximity to cigarette smoke entering the lung. This concept has further evidence by the fact that an increase of adenocarcinoma in smokers was found to correlate with design changes of cigarette in the 1950s that caused deeper inhalation for the individual [1].     4		1.1.2 Current treatment strategies for LC In terms of LC incidence and deaths, the single most important factor to reduce these numbers is the control and prevention of smoking. However, this should not undercut the importance of research as this disease still has one of the lowest 5-year survival rates of any major cancers. As previously alluded to, the low 5-year survival rate can be primarily attributed to the fact that the majority of cases are diagnosed at an advanced stage. This has led to two main areas of research for LC, namely; to be able to detect and diagnose the disease earlier through screening and biomarker identification and to investigate new therapies and strategies to treat advanced stages of the cancer. Like many cancers the most effective treatment strategy currently for LC is surgical resection. Though this strategy is often used with adjuvant chemotherapy and/or radiation treatment, it is often only effective for cancers diagnosed at an early stage when the tumor is localized and metastasis has not occurred [1]. As most LCs are diagnosed at an advanced stage, one of the main areas of research is to investigate new strategies to allow detection of LC earlier, where the 5-year survival rate ranges from 27-55%, depending on localization [3]. Currently, the majority of early-detection research has focused on identification of biomarkers, such as microRNA-based screening in sputum and plasma samples, though these methods still require clinical trials for further validation [1]. Another strategy being closely looked at is using low-dose CT (computed tomography) screening of high-risk groups. Initial results have demonstrated a significant reduction in mortality rates in patients at high risk of developing LC [7]. However, further work is needed before such screening can be applied in a population based setting, such as possible harmful effects of multiple low-dose CT scans on the body. 5		Chemotherapy, with or without radiation, is currently the most common form of treatment for advanced stage LC [1].  Though advances in chemoradiation techniques have increased the short-term survival of advanced stage LC patients, its long-term efficacy has shown to be limited. For instance, the increase in the 1-year survival rate for LC patients in 1977 (34%) compared to 2011 (45%) was primarily due to improvements in surgical and chemoradiation techniques [3]. However, the 5-year survival of LC patients did not follow a similar increase demonstrating the limitations in long-term effectiveness of chemoradiation for LC patients. Therefore, there is a need for better strategies to treat advanced stages of LC, with most research and treatments focusing on NSCLC. Such strategies as angiogenesis inhibitors and immunotherapeutic drugs are being used to treat some types of advanced NSCLC with promising early results [1]. In addition, many mutations that cause tumorigenesis of NSCLC, termed “driver mutations [8],” have been identified (Figure 1.1). Driver mutations, unlike “passenger” mutations, are positively selected for during the cells progression to cancer due to their ability to impart some growth advantage for the cell, and are often required for the cancer cells continued survival [8]. Often these driver mutations causing a phenomenon termed “oncogene addiction [9]” where the cancer becomes dependent on the specific oncogene, both the mutant protein itself and the pathway it is involved in, to survive. Therefore, driver mutations often provide possible targetable therapeutic opportunities to specifically target cancer cells through small-molecule inhibitors [10]. For instance, EGFR, KRAS and ALK are the three most common genes mutated in NSCLC (Figure 1.1), which has led to EGFR and ALK inhibitors being implemented for use in certain NSCLC patients in just the last decade [1]. However, resistance to these targeted 6		therapies is common and more research is needed to identify new driver genes, as there are still a large number of cancers with no known driver mutations including in LC (Figure 1.1). Understanding the mechanisms and pathways critical to LC survival will allow new targets and therapeutic strategies to be developed to treat advanced stages of the disease.         Figure 1.1: Incidence of driver oncogenes in non-small cell lung cancer. Pie chart outlines the approximate frequencies of known driver oncogenes in NSCLC. Other known driver oncogenes include 10 genes in total (6 listed plus NRAS, AKT1, MEK1, and RET). Reprinted from: Mancini, M. and Y. Yarden, Mutational and network level mechanisms underlying resistance to anti-cancer kinase inhibitors. Semin Cell Dev Biol, 2016. 50: p. 164-76 [10], with permission from Elsevier.   7		1.2 Epigenetics The concept that an environmental stimulus could lead to inherited traits not linked to genetic differences was first introduced by Waddington in 1939 [11] and set the basis for an ever evolving field that we know today as “epigenetics.” Since then our understanding of the proteins and mechanisms involved in epigenetics has increased substantially and the concept can now be defined as heritable changes in gene expression that are not due to alterations in the DNA [12].  Epigenetics therefore provides a once missing link between how cells with the exact same genetic code can have vastly different phenotypes, such as the different cells in our body. Additionally, epigenetic changes have become increasingly recognized to occur during tumorigenesis and have led to potential therapeutic targets in many cancer types [12]. There are several molecular mechanisms that contribute to epigenetic gene regulation including; DNA methylation, regulation of expression by non-coding RNAs, chromatin remodelling and exchange of histone variants through ATP-dependent processes, and covalent post-translational modifications (PTMs) to histones and histone tail regions [13]. Of these, DNA methylation and PTMs of histones make up the majority of modifications across the entire human epigenome, with a number of therapies targeting these two modifications being already introduced in clinical practice [14].  DNA methylation is the covalent modification of the DNA nucleotide cytosine by the addition of a methyl group, forming 5-methylcytosine [15]. This modification does not constitute a change in the genetic code (mutation) as 5-methylcytosine still pairs with guanine and addition of the methyl group is reversible through several mechanisms. However, the attachment of a methyl group can interfere with proteins, such as a 8		transcription factor, from binding to the DNA and thus represses transcription of downstream genes. Methylation of a single cytosine nucleotide is generally insufficient to permit such repression. However, there are certain regions of DNA with a high percentage of cytosine and guanine nucleotides termed “CpG islands”. These CpG islands are typically found in or around promoter regions and therefore hyper-methylation of this region can result in repressing of downstream genes [15]. This mechanism can play a major role in certain cancers as the hyper-methylated promoter region can cause silencing of tumour suppressor genes, which has led to the identification and use of DNA methylation inhibitors (5-azacytidine, decitabine) in certain hematological malignancies [16].  1.2.1 Chromatin structure and histone modifications To understand how covalent modifications of histones can affect gene regulation it is important to understand the structure and formation of chromatin. The formation of chromatin is integral for the cell as it provides DNA compaction, increases resiliency to DNA damage, and governs gene expression [15, 17]. The structure of chromatin is dynamic and influences genetic expression so greatly that epigenetics is often used more broadly to describe the study of chromatin and chromatin associated proteins [12]. Briefly, DNA (~147 base pairs) is wrapped around histone octamers, made up of two copies of each core histone (H2A, H2B, H3 and H4) [17], forming the basic unit of chromatin known as “nucleosomes”. As DNA is continuous, there are sections of DNA that are wrapped around the histones and other sections that link one nucleosome to 9		another (termed “linker DNA”). A section of nucleosomes can therefore be thought of as resembling “beads-on-a-string”, with the beads being the nucleosomes and the string being linker DNA [17]. These nucleosomes are then further packaged to form chromatin (Figure 1.2); however, the exact mechanism of this packaging is still not well understood. Nevertheless, it is apparent that chromatin can be in one of two states; heterochromatin, which is tightly packed chromatin; or euchromatin, which is less condensed [12]. How tightly a section of DNA is packaged can influence if genes in that section are expressed or repressed. For instance, as euchromatin contains less-tightly packed nucleosomes DNA is more accessible to proteins such as transcription factors and RNA polymerases, resulting in the majority of active transcription to be in euchromatin sections [12]. In addition, chromatin compaction can be influenced by covalent PTMs of histones. In this way PTMs of histones can regulate sections between euchromatin and heterochromatic states, thereby influencing expression of genes in these areas [13].      10		 Figure 1.2: Structure of chromatin. Schematic representation of the formation of chromatin. Initially DNA (~147bp) is wrapped around a histone octamer forming a nucleosome, the primary unit of chromatin. The tails of histones protrude out from the nucleosome and are subject to post-translational modifications. A section of unwound chromatin resembles “beads-on-a-string,” with the nucleosomes as the beads and linker DNA as the string. The nucleosomes then further condense to form chromatin, the main structure of DNA in the cell. During mitosis chromatin can be clearly seen as chromosomes. Adapted with permission from Annual Reviews:	Hansen, J.C., Conformational dynamics of the chromatin fiber in solution: determinants, mechanisms, and functions. Annu Rev Biophys Biomol Struct, 2002. 31: p. 361-92. [18],   11		As previously mentioned, the second major mechanism of epigenetic gene regulation in the cell is the covalent modification of histones. These reversible, post-translational modifications have emerged as playing an integral role in the regulation of gene expression [19]. As a protein made up of a chain of amino acids, histones are able to receive PTMs throughout their entire sequence. However, the majority of these modifications are found on the N-terminal end of histones, referred to as “histone tails,” which protrude out from the globular core of the nucleosome (Figure 1.2). These modifications include; acetylation, methylation, ubiquitination, phosphorylation, and SUMOylation, as well as a few other less common types [20].  These covalent modifications can cause very different effects depending on the specific modification (acetyl, methyl, etc.), the specific residue of the histone, the number of groups added, and whether they are in close proximity to other functional groups. Acetylation (the addition of an acetyl group, CH3CO) is the most studied of histone PTMs, with acetylation of lysine residues being the most common [13]. The addition of an acetyl group to a lysine negates the lysine’s normally positive charge. This results in reduced affinity for negatively charged DNA causing the chromatin to adapt a more relaxed/open structure, although certain acetylation marks have also been associated with chromatin compaction [20]. In addition, acetylation of specific residues can also be recognized and bound by proteins containing bromodomains, which usually increases transcription of genes in that area, making acetylation a fundamental mark of active gene transcription [21]. As these covalent modifications are reversible, it is unsurprising that there are many proteins that regulate addition or removal of these functional groups.  12		1.2.2 Epigenetic writers, erasers and readers Over the past decade the knowledge of post-translational modifications of histones and the proteins involved in these modifications has increased substantially [13]. The epigenetic proteins involved with these covalent modifications can be categorized into three main classes; epigenetic writers, erasers, and readers. In general, epigenetic writers add functional groups to histone tails, erasers remove these groups, and readers bind to the functional groups. For instance, histone acetyl-transferase (HAT) proteins are a class of epigenetic writers that enzymatically add an acetyl group to the ε-amino group of certain lysine side chains of histone tails [13]. Epigenetic writers also include proteins that catalyze modifications to DNA such as DNA methyltransferases [15]. Therefore, different epigenetic writers can play opposing roles in gene expression, as acetylation of histones is usually linked to a more open structure and increase in gene expression, while DNA methylation is usually associated with repression of genes, as previously described. Counteracting the writers are the epigenetic erasers, a class of proteins that are able to remove functional groups. These include proteins such as DNA demethylases and histone deacetylases (HDACs), proteins that remove methyl and acetyl groups from DNA and histones, respectively [15, 19]. The regulation of these proteins is critical to the cells survival and they are frequently deregulated in many diseases including cancer. What seems apparent is that mutation of epigenetic writers or erasers can provided a mechanism for global rewiring of a cell, being linked not only as a mechanism of adaption to environmental stressors and maintenance of transformed cells, but also as driving factors in the progression of cancer cells [12]. For instance, abnormal lysine acetylation in cancer is often found to cause activation of pro-survival and proliferation 13		pathways and has even been linked to inactivation of tumor suppressor functions [22]. The realization in the importance of epigenetic writes and erasers in the progression and homeostasis of cancer has led to a wide array of exciting new therapies in the last few years. Clinically, the most advanced of these therapies have been inhibitors of DNA methyltransferases [12] and HDAC inhibitors, including Vorinostat, the first HDAC inhibitor to be approved for relapsed T-cell lymphoma [21]. The last class of epigenetic proteins are the readers, proteins that are able to bind to specific functional groups covalently attached to DNA or histones [22]. Often these proteins influence gene expression by acting as scaffolding proteins that function through recruitment of other transcriptional activators or repressors to the area of interest [23]. This class consists of structurally diverse proteins that contain different effector modules that allow them to specifically bind to certain covalent chemical modifications. One of the most important modules that has garnered increasing interest due to its role in various cancers, and having been successfully targeted by small molecule inhibitors, are proteins containing bromodomains [20]. The bromodomain is a conserved ~110 residue sequence that folds into a structure consisting of four alpha helices (αZ, αA, αB, αC) linked by two loop regions (ZA and BC loops) (Figure 1.3) [24]. ZA and BC loops can both vary in length and contribute to the substrate specificity for each bromodomain. Once folded, bromodomains create a deep hydrophobic pocket that recognizes and interacts with acetyl-lysine, giving these domains their function [20].    14		            Figure 1.3: Bromodomain tertiary structure. Bromodomains fold into a conserved structure of four alpha-helices (αZ, αA, αB, αC) linked by two variable loop regions (ZA and BC). This creates a hydrophobic pocket allowing the domain to bind to acetylated-lysine residues of histone tail regions. Shown is the first bromodomain (BD1) of BRD4 with the hydrophobic pocket indicated in green and the acetyl lysine in red. Adapted with permission from International Journal of Molecular Sciences: Taniguchi, Y., The Bromodomain and Extra-Terminal Domain (BET) Family: Functional Anatomy of BET Paralogous Proteins. Int J Mol Sci, 2016. 17(11) [24]. 15		In total, the human proteome encodes 61 different bromodomains within 46 proteins, categorized into 8 bromodomain-containing protein families (Figure 1.4) [21, 25]. Bromodomain containing proteins are not confined to only epigenetic readers and include; HATs and HAT-associated proteins, histone methyltransferases, helicases, chromatin remodelers, and nuclear scaffolding proteins. Still, the most studied of the bromodomain containing proteins, and the most important to this thesis, are the bromodomain and extraterminal (BET) family of proteins (Figure 1.4) [25].             16		            Figure 1.4: Phylogenic-tree of the human bromodomain containing families. In total, 61 distinct bromodomain modules are present in 46 unique proteins that are categorized into eight families (roman numerals). Family II represents the BET family of epigenetic readers (red arrow). Adapted by permission from Macmillan Publishers Ltd: Filippakopoulos, P., et al., Selective inhibition of BET bromodomains. Nature, 2010. 468(7327): p. 1067-73 [25].   17		1.3 Bromodomain and extraterminal family proteins The BET proteins are an important family of epigenetic readers that are able to recognize and bind to acetylated lysines of histone tails [23]. The BET family consists of the proteins BRD2, BRD3, BRD4, and BRDT, with BRDT being the only member not ubiquitously expressed but instead confined to germ cells [21]. BET proteins are characterized by two tandem bromodomains (BD1 and BD2), located N-terminally, an extraterminal (ET) domain, as well as a variable length C-terminal end (Figure 1.5) [26]. The tandem bromodomains are quite variable from each other, exhibiting only ~44% sequence identity, whereas BD1 and BD2 across BET family member’s exhibit ~75% sequence identity [27]. However, the specificity and affinity of BET proteins to acetylated lysines depends on the orientation of the hydrophobic pockets themselves, which are unique to each different BET protein bromodomain [28]. The C-terminal end is also variable in its length between different family members, with BRD4 and BRDT being the only BET family members to contain the C-terminal interacting motif (CTM) at their C-terminal ends (Figure 1.5) [20].          18		             Figure 1.5: Domain structure of the BET family proteins.  Schematic representation of the BET family members and their respective domains; bromodomains (BD1 and BD2), ET (extraterminal domain), and CTM (C-terminal motif). In addition, the three isoforms of BRD4 are shown, which are identical at the N-terminal end except for the final three amino acids, and variable at the C-terminal end [29]. BRD4 A represents the long isoform, BRD4 C represents the short-isoform, the two main forms in the cell, while BRD4 B represents a lesser known isoform only shown to be expressed in an osteosarcoma line [30]. Reprinted from: Shi, J. and C.R. Vakoc, The mechanisms behind the therapeutic activity of BET bromodomain inhibition. Mol Cell, 2014. 54(5): p. 728-36 [31], with permission from Elsevier.     19		The main function of the CTM region is its interaction with positive transcription elongation factor b (P-TEFb), a main player in the regulation of transcription [30]. Through binding, BRD4/T enables P-TEFb to dissociate from its inactive complex with HEXIM1 and 7s snRNA and be recruited to sites of lysine acetylation [21]. Here P-TEFb, a cyclin dependent kinase, can then phosphorylate the C-terminal domain (CTD) of RNA polymerase II (RNA pol II), an essential step in the progression from initiation to elongation of the mRNA transcript [29]. This is important as, for most genes, RNA pol II becomes trapped after initiation on the promoter sequence and is unable to proceed with transcribing the mRNA. Through phosphorylation, P-TEFb is able to alleviate this entrapment and allows RNA pol II to fully transcribe the mRNA. Without CTD phosphorylation RNA pol II is often forced to release from the DNA, resulting in truncated mRNA transcripts that are quickly degraded. Through this function BRD4, the only ubiquitously expressed BET member to contain the CTM, has been implicated to be involved in many different cancers through regulation of c-MYC (MYC) [32-36] and other oncogenic transcription factors [37, 38].  BRD4 was also shown to form a translocation fusion product with NUT t(15;19), with the resulting BRD4-NUT oncoprotein responsible for an aggressive midline carcinoma [39]. Furthermore, in the majority of cases inhibitors of BET proteins (BETis, discussed later) have shown to impart their therapeutic effect on cancers through targeting BRD4, even though these inhibitors ubiquitously inhibit all BET proteins [25, 31]. However, BRD2 and BRD3 have been shown to have their own functions in the cell through associating with sections of hyper-acetylated chromatin and being involved in transcription and histone chaperone activity [40, 41]. Still, for the majority of cancers the BET family of proteins is discussed 20		in regards to BRD4, as in many of these cancers knockdown of BRD4 recreates the effects of BETis best, which our lab has previously shown is the case in LC [37].  For these reasons BRD4 was the main BET proteins studied in this thesis in terms of BETi treatment and resistance in LAC.  1.3.1 BRD4 BRD4, as with the other BET family members, contains two bromodomains that confer binding to acetylated lysines of histone tails. BRD4 was initially linked to epigenetic memory and cell-cycle control due to its continued association with chromatin during mitosis [30].  Since then its role as an epigenetic reader has grown with BRD4 now being linked to three major steps in the transcriptional process (Figure 1.6). All these functions involve its ability to bind to acetylated lysines of histone tails (H3 and H4) and recruit transcriptional proteins to these sites. First, BRD4 has been shown to interact with transcription factors such as Gdown1, MED26 and others, resulting in their localization to certain genes and “committing” the gene to transcription [42]. Secondly, BRD4 has been shown to interact with the mediator complex, a critical component in the pre-initiation complex that links transcription factors to RNA pol II, and therefore initiates transcription of a gene. Lastly, as described previously, BRD4 has been shown to interact with P-TEFb, an integral part of RNA pol II progressing to the elongation stage of transcription. In addition, BRD4 often occurs at regions of the DNA that contain clusters of enhancer elements, termed “super enhancer” regions [26, 43]. These super enhancer regions are essential for maintaining cellular identity and frequently play key roles in regulating oncogene expression during cancer development and progression [23, 26]. 21		Therefore, there are many genes whose expression is dependent on BRD4 regulation, including oncogenic driver genes in many cancers.             Figure 1.6: Regulation of transcription by BRD4 is mediated at three different levels. By binding to acetylated lysine residues of histone H3 and H4 and through interaction with other proteins BRD4 is able to regulate transcription of genes at 3 steps in the transcription process. Namely; transcription commitment through binding transcription factors (left), initiation of transcription through interactions with the mediator complex (middle), and elongation progression through its recruitment of P-TEFb (left). Figure reprinted with permission from Faculty of 1000: Chiang, C.M., Brd4 engagement from chromatin targeting to transcriptional regulation: selective contact with acetylated histone H3 and H4. F1000 Biol Rep, 2009. 1: p. 98 [44].     22		BRD4 occurs as two main isoforms; the long isoform (BRD4-LF) (aa 1–1362) and the alternatively spliced short isoform (BRD4-SF) (aa 1–722) [27, 45]. There is also a third isoform, which is truncated at 722aa like the short isoform but contains an additional 76aa C-terminal peptide (Brd4 B, Figure 1.5). However, this isoform is poorly characterized and understood and has currently only been shown to be expressed in an osteosarcoma line (U2-OS) [30].  BRD4-LF represents the full-length form of the protein (Brd4 A, Figure 1.5) and is the only form that contains the CTM. This form is often the only isoform expressed in a cell and therefore is often referred to as BRD4 [30], and unless indicated will be the isoform in reference to BRD4 in this thesis.  Unlike BRD4-LF little is known about the function of the alternatively spliced BRD4-SF, although it is clear that it does not contain the C-terminal motif and therefore does not interact with P-TEFb (Brd4 C, Figure 1.5). Initial evidence supports BRD-SFs role as being opposed to BRD4-LF; such as mediating increase chromatin compaction, and through competitive interaction with BRD4-LF binding partners SIPA1 and RRP1B [29]. It is also unclear if BRD4-SF would compete with BRD-LF for binding to acetylated lysine residues as they contain identical bromodomains (Figure 1.5). However, one paper recently found that BRD4-SF remains associated with the nuclear-membrane while BRD4-LF is associated within the nuclear matrix, suggesting that competitive interactions of BRD4-SF may be infrequent [29]. Lastly, it is worth noting that there is some discrepancy between the predicted molecular weight of BRD4-LF during western blot analysis, with the protein being alternatively described as having a molecular weight of 152kDa or 200kDa with little understanding of this cause of discrepancy. Though it is unclear why two different molecular weights can be essentially detected from the same transcript, most literature 23		has focused on the 200kDa version, which most antibodies specifically detect, and which was detected in this thesis.    1.4 Bromodomain and extraterminal inhibitors   The successful development and use of inhibitors to this family of proteins (BETis) is the first example for successful targeting of epigenetic readers [21]. Though initial fragment-like bromodomain inhibitors had been described in 2005 [22] it was not until 2010 that Filippakopoulos et al. identified the first BET bromodomain inhibitor, JQ1 (Figure 1.7, left) [25]. Since then other structurally similar BETis have been identified. Some have progressed further in clinical trials than JQ1, which was found to have a short half-life in vivo [46]. In total clinical trials for 10 BETis have been initiated; most notable are RVX-208 (Resver-logic), as the only BETi currently to reach phase III clinical trials (seven trials in total) for atherosclerosis and associated cardiovascular diseases, and I-BET762 (also known as GSK525762A) (Figure 1.7, right), which has reached Phase II clinical trials in NUT midline carcinomas, and is used in this thesis [22, 23]. Furthermore, additional trials are planned for combination based therapies with chemotherapy, immunotherapy, and other targeted therapies [23].      24		                   Figure 1.7: Chemical structures of the two BET inhibitors, JQ1 and I-BET762. Each inhibitor contains an acetyl-lysine binding motif (in pink) that displaces the ε-acetyl-lysine group of the histone and binds to the BET proteins Bromodomains. Adapted by permission from Macmillan Publishers Ltd: Belkina, A.C. and G.V. Denis, BET domain co-regulators in obesity, inflammation and cancer. Nat Rev Cancer, 2012. 12(7): p. 465-77 [47]. 	      25		1.4.1 BETi treatment in cancer  As BRD4 has been shown to regulate the transcription of oncogenes, the efficacy of BET treatment of several different cancers is being evaluated. Despite JQ1 having limited clinical effectiveness it is still the predominant BETi used in in vitro studies. This is predominantly due to the structural similarities between the different inhibitors (Figure 1.7), it being inexpensive and readily available, and its prior use in many different cancers. For these reasons JQ1 was used as the predominant BETi in previous work in our lab [37] and as the main BETi in this thesis. JQ1 inhibits BET proteins by binding to the hydrophobic pockets of the bromodomains, thereby inhibiting their ability to bind to acetylated lysines. By inhibiting this ability, BRD4 is unable to localize normally and genes that were under the influence of BRD4 for their expression are repressed. Several papers have shown that this inhibition is likely due to loss of BRD4 at super enhancer regions, resulting in selective transcriptional repression of certain oncogenes [22, 23]. For the majority of cancers, including both hematological malignancies and solid tumors, the main down-regulated gene due to BET inhibition is MYC, a multifunctional transcription factor and contributor to the pathogenesis of certain cancer types [42, 48]. These “MYC-dependent” cancers include those with MYC translocations such as acute myeloid leukemia (AML) [32], acute lymphoblastic leukemia (ALL) [49], Burkett’s lymphoma (BL) [42], mixed-lineage leukemia (MLL) [50] and multiple myeloma (MM) [36], as well as neuroblastoma [35], colorectal [43, 51], breast [52], and pancreatic [33, 34] cancers. However, previously our lab has shown that lung adenocarcinoma (LAC) lines are not inhibited by JQ1 through downstream down-regulation of MYC. Instead our lab identified FOSL1, an oncogenic transcription factor itself, as being down-regulated upon BET inhibition and its loss concurring with sensitivity in LAC lines [37]. Since this 26		study, only osteosarcoma has also shown MYC-independent sensitivity to BET inhibition, which also identified FOSL1 as the likely downstream regulatory gene [38].   Similar to many inhibitors, there are certain cancer lines within a given type of cancer that will be initially resistant to BETi treatment, referred to as “primary resistant” lines. Additionally, cancers cells are often genetically unstable and frequently develop mutations that can lead to resistance to inhibitor treatment, referred to as “acquired resistant” lines. Clinically, this can often limit the effectiveness of target therapies in cancer patients as patients may not respond to initial treatment (primary) or the patient will develop resistance to treatment (acquired) [33]. One way to increase the success and effectiveness of a targeted therapy in patients is through screening processes to identify genetic mutational profiles that respond to the therapy and therefore select patients that have a high likelihood of responding to treatment. However, unlike certain inhibitors (TKIs, PARP inhibitors) that target cancers with specific mutations, BETi treatment effectiveness is not linked to cancers driven by certain mutations as they act on a more global scale. Furthermore, the epigenetic landscape in a given cell can vary considerably between cancer types and even between patients with the same cancer. Therefore, it is currently difficult to identify common mechanisms of resistance to epigenetic inhibitors and thus would be difficult to identify which patients will respond to an inhibitor such as a BETi. Therefore, further research is needed to better understand the mechanisms and pathways involved in response to BETi treatment and why certain cells are sensitive and other cells resistant to treatment. This could eventually lead to better screening procedures and patient response to BETi treatment as well as identify novel genes/ pathways that resistant lines are dependent on and that can be used to target them through 27		some other fashion. One way to help elucidate mechanisms of resistance is to establish resistant cell lines from initially sensitive cell lines. This has been done in several cancer types, with all cases using cancers sensitive to JQ1 through MYC down-regulation. In these cases, resistance was conferred through MYC re-expression mediated by several different mechanisms such as activation of Wnt signaling [32, 53], co-regulation by GLI2 [33], and most recently through phosphorylation of BRD4 [52] (Figure 1.8).            Figure 1.8: Mechanisms of resistance to BETis show dependence on re-expression of MYC. BETis function to inhibit BRD4 from binding to acetylated lysines (K) and prevent downstream MYC transcription. However, mechanisms of resistance to BETis have shown to occur from rapid re-expression of MYC through several mechanisms. In AML, both Rathert et al. [32] and Fong et al. [53] show that resistance occurs through activation of the Wnt signaling pathway, with β-catenin being recruited and re-activating transcription of MYC. In triple negative breast cancer (TNBC), casein kinase 2 (CK2) was shown to phosphorylate BRD4 allowing it to bind to the transcriptional regulator protein MED1 and cause re-expression of MYC [52]. Reprinted by permission from Macmillan Publishers Ltd: Settleman, J., Cancer: Bet on drug resistance. Nature, 2016. 529(7586): p. 289-90 [54]. 28		1.5 Hypothesis and aims Resistance to BETis has never been studied in a cancer type that shows initial sensitivity to BETis through MYC-independent down-regulation, such as in LAC [37]. I therefore aim to investigate how LAC cells acquire resistance to BETis and hypothesize that lung adenocarcinoma lines will acquire resistance to BET inhibition through a novel mechanism, independent of MYC.   The Aims of this thesis are as follows: Aim 1: To establish and characterize isogenic counterparts resistant to BETis from two sensitive LAC lines  Aim 1a: To create JQ1 resistant lines from two initial JQ1 sensitive LAC lines through culturing in increasing concentrations of JQ1  Aim1b: To further validate and characterize resistant lines to BET inhibition through dose response experiments with JQ1 and I-BET762 Aim 2: To determine common mechanisms of resistance between the two resistant lines  Aim 2a: To identify and validate candidate genes through expression analysis of sensitive lung adenocarcinoma lines and their resistant isogenic counterparts  Aim 2b: To investigate if EMT or phosphorylation of BRD4 have roles in acquiring resistance to JQ1 in LAC lines Aim 3: To determine if there are common mechanisms of resistance between primary and acquired resistant LAC lines         29		Chapter 2: Materials and methods  2.1 Cell lines and media All three cell lines used here (H23, H1975, H2030) are lung adenocarcinoma lines initially purchased from the American Type Culture Collection (ATCC), and cultured in RPMI-1640 medium (Gibco, cat. #11875119) supplemented with 5% (vol/vol) FBS (Sigma), 1% penicillin-streptomycin solution (Gibco), and 1% Glutamax (Gibco) (standard growth media). Cultured cells were incubated in a humidified incubator at 37°C with 5% (vol/vol) CO2 and 95% (vol/vol) air.  2.2 Establishing resistant lines JQ1 resistant lines were generated by treating parental H23 and H1975 cells with increasing concentrations of JQ1 over a period of 6 months until cells were able to grow in 10µM JQ1. In parallel, cells were cultured in standard growth media plus 0.1% (vol/vol) DMSO as a control. Resistant lines were maintained in 10µM JQ1 media and the control lines in 0.1% DMSO media. Both resistant and control lines were authenticated by STR profiling through Genetica. 2.3 Dose response experiments and long-term growth assays Dose response experiments entailed 21-point doubling dilutions (in DMSO) starting at 10µM for JQ1, Trametinib, Crizotinib, and BEZ-235, 25µM for I-BET762, and 100µM for Afatinib. 10µl of each dilution was added in quadruplicate to wells of a 96-well plate (Falcon). 10µl of 100µM of Etoposide (Selleckchem, cat. # S1225) was also added in quadruplicate to four wells of the 96 well plates. Cells were then added (in 90ul complete growth media) at 3000-8000 cells per well, based on growth rate to ensure optimal final 30		densities, with final drug concentrations having a 0.1% DMSO concentration (vol/vol). Plates were incubated (at previously described conditions) for 72hrs at which time 10ul (10% of total volume) of Alamar Blue cell viability reagent (Invitrogen, cat. # Dal1100) was added and fluorescence measured using a BioTEk Cytation 3 imaging reader (Excitiation: 540nm and Emission 590nm). Results were obtained using BioTek Gen5 software version 2.06.10. IC50 values were generated using GraphPad Prism7 software and are presented as the mean ± SEM (Standard error of the measurement) of between 2-4 biological replicates as indicated. For long-term cell growth assays cells, 20,000 cells were seeded into 6-well plates (Falcon) for H1975 control and resistant lines; 25,000 for H23 control line; and 50,000 for H23 resistant line. Cells were seeded in complete growth media with either 1µM or 10µM JQ, or equivalent vehicle (0.1% DMSO) as a control, with fresh media being exchanged every 2-3 days. On day 10, media was aspirated and cells stained with Crystal violet (CV) staining solution (Sigma) (50% CV, 25% methanol, 25% PBS) for 30min before rinsing plates of excess CV solution and imaging the following day. Experiments were repeated at least twice.  2.4 RNA interference (RNAi) experiments using siRNAs Cells were seeded into antibiotic-free standard growth media in 6-well plates, with seeding densities ranging from 200,000 for H1975 control and resistant; 250,000 for H23 control; and 350,000 cells per well for H23 resistant. Wells were checked the following day (24hrs) for ~70-80% confluency before continuing. Cells were transfected with 10uL of 10µM ON-TARGETplus SMARTpool siRNAs (Dharmacon), specific to the gene of 31		interest or a non-targeting control (NonT), using DharmaFECT 1 transfection reagent (Dharmacon) according to the manufacturer’s instructions. 16-18hrs after transfection cells were trypsinized and counted, and then re-seeded in 96-well plates at 2000 (H23 control, H1975 control and H1975 resistant) and 4000 (H23 resistant) cells per well, in quadruplicate for each condition. Cells were also re-seeded into 6 well plates for an additional 48hrs and then lysed using 100uL radioimmunoprecipitation assay (RIPA) lysis buffer (Pierce Biotech) containing Halt Protease and Phosphatase Inhibitor Mixture (Thermo Scientific). Cell growth was analysed 96hrs after re-seeding into 96-well plates using Alamar blue as described above and fluorescence measured as previously described. Viability for each siRNA was averaged between the four replicates and then compared to the NonT control (set at 100% viability) to calculate relative viability. For each line, resistant and control values for each specific siRNA were compared using the two-way ANOVA (Analysis of variance) test in the Prism 7 software package (GraphPad), with significance being denoted as ns = not significant, * = P < 0.05, ** = P < 0.01, *** = P < 0.005, **** = P < 0.001. The two-way ANOVA test was also used to determine significance of siRNA to NonT within each control and resistant line. Each experiment was performed in triplicate. 2.5 CX-4945 and JQ1 combination treatment For combination growth assays, cells were seeded into 6-well plates (Falcon) at 100,000-150,000 (H23 control, H1975 control, and H1975 resistant) and 250,000 (H23 resistant) for 72hrs treatment plates and 10,000-15,000 (H2030, H23 control, H1975 control, and H1975 resistant) and 40,000-50,000 (H23 resistant) for 10-day treatment plates. For 10-day plates, media was changed with fresh inhibitor media every 2-3 days. For both 72hrs 32		and 10-day cells were seeded in regular culture media and then the media changed the following day with media containing inhibitors, which was designated the 0 time-point/day. After 72hrs or 10-day, cell viability was measured by addition of 200ul Alamar blue to each well (2ml of media, 10%) and then quadruplicate 100ul samples extracted and put into a 96 well plate and fluorescence measured as previously described. Therefore, for each combination of inhibitors 4 replicates were extracted and measured. This was repeated twice for 72hrs plates and in triplicate for 10-day plates. After extracting the 100ul samples the rest of the media was aspirated and cells stained with CV staining solution (Sigma) (50% CV, 25% methanol, 25% PBS) as above. 2.6 Quantitative analysis of combinational treatment For each experiment, the four replicates were averaged and then compared to the control well (0 JQ1, 0 CX-4945), which was set as 100% viability. These values for each combination were then averaged from the two or three replicate experiments to give the percent growth inhibition for each combination of inhibitors. Drug synergism was analyzed from these average values using CompuSyn software (www.compusyn.com), which is based on the Median-Effect Principle (Chou and the Combination Index-Isobologram Theorem (Chou-Talalay) [55]. Following the instructions of the software, drug combinations at non-constant ratios were used, with any values >0.99 being changed to 0.99, to allow the software to calculate Combination Index (CI) values, where CI<0.75 indicates synergism; CI between 0.75-1.25 indicates additive effects, and CI>1.25 indicates antagonism.  33		2.7 Gene expression profiling and data analysis In triplicate, RNA was extracted using the RNeasy Mini Kit (Qiagen) from H23 and H1975 resistant lines (growing in 10µM JQ1 media), control lines (growing in 0.1% DMSO media), and control lines treated with 10µM JQ1 for 6hrs, all in 6-well plates. Samples were sent to The Center for Applied Genomics (TCAG, Toronto, Canada) where sample quality, sample labeling, array hybridization, and data acquisition were performed, with the Affymetrix Human PrimeView Array being used as previously described [37]. Data obtained from TCAG was already normalized by Robust Multiarray Analysis [56] as previously described [57]. Therefore, for each probe, triplicate expression values (in log2 scale) for resistant, control, and 6hrs treated cells were obtained. Resistant and 6hrs expression was then compared (separately) to the control line, and genes with ≥2-fold change in expression (≥1 or ≤1 of log2 values) and a Benjamini-Hochberg correct p-value <0.005 were kept. To identify differentially regulated genes in the resistant lines, the resistant and 6hrs gene lists were compared using GeneVenn program (genevenn.sourceforge.net) for both H23 and H1975, with genes only expressed in the resistant lines being kept. The two resistant only gene sets (H23 and H1975) were then compared using GeneVenn again, to determine common genes likely to be important for resistance mechanisms in LAC to JQ1.    34		2.8 Protein extraction and western blot analysis Cells were grown in either 6-well or 6cm plates and were lysed with 50-200ul of RIPA lysis buffer (Pierce Biotech) containing Halt Protease and Phosphatase Inhibitor Mixture (Thermo Scientific), after washing with ice cold phosphate-buffer saline (PBS). Cells were harvested by scrapping and placed in -80 at least overnight. For CK2 inhibitor (CK2i) treatment (Figure 3.4C) cells were treated with 10µM, 20µM or 40µM of CX-4945 in 6-well plates for 2hrs before lysing as above. Cell lysates were sonicated, cleared of cell debris through centrifugation at 18,000 X g for 15 min, and then protein concentration determined using the Pierce BCA Protein Assay kit (Thermo Scientific) according to the manufacturer’s instructions, with absorbance being measured at 562nm using a plate reader as previously described. 25ug of protein was used for each sample and denatured by boiling at 100°C for 10min in 4x Laemmli sample buffer (BioRad) with 1:10 addition of 2-Mercaptoiethanol (Thermo Fisher). Samples were loaded on Novex 4-12% Bis Tris Gels (NuPage) and electrophoresed at 150V for 2 hrs, then transferred to polyvinylidene fluoride (PVDF) membranes (Millipore) at 110 V for 1 hr 10 min. Membranes were blocked at room temperature for 1-2hrs with either 5% (wt/vol) non-fat dry milk or 5% (wt/vol) bovine serum albumin (BSA) (Sigma), made up in TBS-T (1× Tris-buffered saline, 0.1% (vol/vol) Tween-20) (TBS BioRad), followed by immunoblotting with primary antibodies (described below) overnight at 4°C with shaking. Membranes were incubated with the appropriate horseradish peroxidase (HRP) conjugated secondary antibodies (donkey anti-rabbit (sc-2313) and donkey anti-mouse (sc-2314) from Santa Cruz) at 1:2,000-1:10,000 dilutions for 1.5hrs at room temperature and subsequent detection using SuperSignal West Pico Chemiluminescent Substrate 35		(34087) or SuperSignal West Femto Chemiluminescent Substrate (PI34095) (Pierce Biotech).  2.9 Antibodies, inhibitors and other reagents All antibodies used for western blot analysis were 1:1000 dilution except for CK2α (1:500) and GAP (1:2000) and are as follows: From Cell Signaling Technology (CST); BRD4 (13440), AXL (8661S), E-cadherin (3195), N-cadherin (13116), β-catenin (8480), Vimentin (5741), EGFR (2232S), Ras (8955S), EGFR (L858R mutant specific) (3197), MYC (5605), and FOSL1 (5281). From abcam; CK2α (ab137788), and from Santa Cruz; GAP (sc-47724). Lastly the pBRD4 antibody was a gift from C. M. Chiang with specificities as previously described [28].  (+)-JQ1 (active enantiomer) (cat. # 4499) was purchased from TOCRIS. I-BET762 (cat. # 10676) was purchased from Caymen Chemical. Trametinib (cat. # S2673), Crizotinib (cat. # S1068) and CX-4945 (also known as Silmitasertib) (cat# S2248) were purchased from Selleck Chemical, and BEZ235 (cat. #CT-BEZ) and Afatinib (also known as BIBW-2992) (cat. # CT-BW2992) was purchased from ChemieTEK. All drugs were suspended in DMSO. Recombinant human TGF-beta 1 (TGFβ1) protein was purchased from R & D Systems (cat. # 240-B-002) and resuspended in 4mM HCl containing 1 mg/mL BSA to a final concentration of 10ug/ml. This was then added to regular cultured media at 1:1000 to give a final concentration of 10ng/ml.   36		Chapter 3: Results  3.1 Establishment of isogenic BET inhibitor sensitive and resistant LAC cell lines  To elucidate mechanisms of JQ1 resistance in LAC, two JQ1 sensitive LAC lines, H23 and H1975 - a KRASG12C mutant and an EGFRL858R/T790M dual mutant line, respectively - were cultured with increasing concentrations of JQ1 over time until resistance was achieved. In total, cells were passaged over a six-month period until the lines were able to grow in 10µM JQ1, at which point they were considered resistant (Figure 3.1A). In parallel, H23 and H1975 cells were also cultured in the same manner in vehicle only media (media containing 0.1% DMSO), as to control for effects associated with cell passaging (56 times for both lines). Therefore, for each line, both a control and resistant population were established.  To confirm JQ1 resistance, long-term growth assays and JQ1 dose response experiments were performed in duplicate and triplicate, respectively. Both JQ1 resistant lines showed growth after 10-days of 1µM JQ1 treatment while both controls at this concentration showed no cells (Figure 3.1B). JQ1 resistance was further confirmed through dose response experiments with both resistant lines having an IC50 over 10µM while the H23 and H1975 control lines were still sensitive to JQ1 treatment (IC50s of 0.3µM ± 0.1 and 1.1µM ± 0.4, respectively) (Figure 3.1C and D). However, it was observed that both resistant lines grew at a slower rate upon JQ1 treatment, as compared to DMSO control in the long-term growth assays, and through observation of a descending curve in the dose response experiments for both resistant lines. This suggested that JQ1 may still lead to growth inhibitory effects in the resistant lines and lead me to assess whether resistance was only temporary and test if cells would revert 37		back to JQ1 sensitivity if cultured in media lacking JQ1. Therefore, both resistant lines were cultured for approximately 2½ months in 0.1% DMSO media and then resistance was tested through long-term growth assays. Both lines still showed resistance to JQ1 as evident by growth in the 10µM JQ1 wells, confirming that resistance acquisition is a permanent, stable phenotype (Extended Figure 6.1B). In addition, H1975, which showed substantial change in morphology (Figure 3.2A), did not show a reversion back to the original morphology upon culturing in JQ1 lacking media (Extended Figure 6.1A), further exemplifying that resistance lines are permanently altered.  To confirm that the cell lines were isogenic counterparts, cell authentication by short tandem repeat (STR) profiling was conducted on control and resistant lines, as well as the parental line that they were originally started from (“in lab” line). These STR profiles were then compared to the reference American Type Culture Collection (ATCC) STR database, and percent match assessed with ≥80% being considered authenticated [58]. In addition, as the ATCC reference STR profile only contains nine core reference loci, the resistant and control lines were compared to their parental counterparts using all 16 STRs analyzed, with all comparisons exceeding the 80% cut-off, confirming authenticity (Extended Figure 6.1D) (raw STR profiles shown in Extended Figure 6.1E and F). Therefore, both resistant and control lines were confirmed to be isogenic lines of the H23 and H1975 parental lines. This was also seen through western blot analysis in H1975 as both the control and resistant line still expressed the L858R EGFR mutant protein (Extended Figure 6.1B).   38		024681012April May June July Aug SeptJQ1 (uM)H23H1975JQ1 I-BET762    													    0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0J Q 1J Q 1  (u M )Relative Fluorescence to controlH 2 3  C o n tro lH 2 3  R e s is ta n tH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n t0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0I-B E TI-B E T  7 6 2  (u M )Relative Fluorescence to controlH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n tH 2 3  C o n tro lH 2 3  R e s is ta n tCell line IC50 (µM) H23 Control 0.3 ± 0.1 H23 Resistant >10 H1975 Control 1.1 ± 0.4 H1975 Resistant >10 Cell line IC50 (µM) H23 Control 0.3 ± 0.1 H23 Resistant >20 H1975 Control 3 ± 1.8 H1975 Resistant >20 A B C D H23 E FD G 39		Figure 3.1: JQ1 resistance established in two JQ1 sensitive lung adenocarcinoma lines. Two JQ1 sensitive LAC lines, H23 and H1975, were cultured in increasing concentrations of JQ1 media, or 0.1% DMSO media as a control, for approximately 6 months until cells were able to sufficiently grow in 10µM JQ1 media (A). Resistance was confirmed in resistant lines through long-term growth assays (B) and JQ1 dose response experiments (C). Graph is representative of one experiment, 4 replicate values for each dose, while accompanying IC50 values (D) are averaged from four experiments ± SEM. Representative dose response graph for I-BET762 (E) and IC50 values (F) from three experiments ± SEM. Western blot analysis used to assess apoptosis through cleaved-PARP of H23 resistant and control lines (G). Cells were treated with indicated concentrations of JQ1 for 48hrs and then assessed for protein levels with GAPDH serving as a loading control.                     40		As inhibitors can potentially have off-target effects in cells it could be postulated that either line may have been sensitive to JQ1 due to some mechanism not related to inhibition of the BET family proteins, and that resistance had been acquired to this mechanism. Therefore, the resistant lines were assessed on their resistance to BET inhibition through dose response experiments with another widely used BETi, I-BET762 (Figure 3.1E) [22]. Both resistant lines had an IC50 over 20µM, the highest I-BET762 dose used, while H23 and H1975 control lines had an IC50 of 0.3µM ± 0.1 and 3µM ± 1.8 (Figure 3.1F), respectively, indicating that cells are resistant to BET inhibition and not off-target effects. Similar to JQ1 treatment both resistant lines showed decreased dose response slopes, indicating that viability is decreasing upon higher JQ1 treatment. This suggests that resistant cells are still undergoing death or are growing at a slower rate. To test if resistant cells were undergoing apoptosis due to BET inhibition western blot analysis was used after treatment with different concentrations of JQ1, as previously demonstrated in JQ1 sensitive lines [37]. For this, H23 control and resistant lines were treated with increasing concentrations of JQ1 for 48hrs and apoptosis levels were assessed through detection of cleaved-PARP, which detects the large fragment of PARP produced by caspase cleavage, an indicator of apoptosis [59]. As Figure 3.1G shows, cleaved-PARP levels increased in the control line at 2µM and 10µM JQ1 treatment, but did not increase in the resistant line with the same doses of JQ1, indicating that there is no induction of apoptosis in the resistant lines upon JQ1 treatment. Together, this confirms that while both resistant lines show a decrease in viability through Alamar Blue metabolism at higher JQ1 treatment levels, this is not due to cells undergoing apoptosis and is likely due to the resistant lines having a slower growth rate in the presence of JQ1. 41		3.2 LAC lines acquire resistance to BET inhibitors in a MYC and FOSL1 independent manner. MYC and FOSL1 have previously been shown to be the two main proteins that are down-regulated due to BETi treatment in various cancers, with MYC re-expression being linked to multiple resistance mechanisms. Therefore, it was hypothesized that re-expression of either of these proteins could confer resistance in LAC. To assess this, MYC and FOSL1 levels were compared between resistant, control, and 6hrs JQ1 treated control lines through RNA expression and western blot analysis (Figure 3.2). MYC expression showed an increase of 1.66 and 2.08-fold in H23 and H1975 resistant lines, respectively, compared to their respective control lines. This indicates that an increase in MYC expression could be a cause of resistance. However, MYC expression also increased 2.58 and 1.56-fold for H23 and H1975, respectively, upon 6 hour JQ1 treatment, indicating that the change in MYC expression is likely due to the presence of JQ1 and not a mechanism of resistance. This association was further illuminated in western blot analysis as both resistant lines have higher levels of MYC as compared to control, but have less compared to the JQ1 treated control lines (6hr), as shown through densitometry calculations (Figure 3.2B).  FOSL1 expression, on the other hand, showed decreased levels upon JQ1 treatment and even lower levels in both resistant lines (Figure 3.2A). This was unexpected as previous papers [32, 33, 52, 53] had all shown cells acquiring resistance through re-expression of the initial down-regulated gene, which for LAC had previously been shown to be FOSL1 [37]. However, this decrease in FOSL1 production was confirmed at the protein level (Figure 3.2B) with both resistant lines having lower levels 42		of FOSL1 than their respective control lines as calculated through densitometry (0 and 0.31 for H23 and H1975 resistant lines, respectively). Therefore, BET inhibitor resistance in H23 and H1975 is independent of MYC and FOSL1.      						.		   Figure 3.2: MYC and FOSL1 show insignificant changes in both resistant lines. RNA expression (A) and western blot analysis (B) both show that MYC levels are higher due to JQ1 treatment (6hr), minimizing the fact that they are up in both resistant lines. FOSL1 levels both show initial decrease upon JQ1 treatment and further decrease in both resistant lines. Resistant lines and 6hr 10µM JQ1 treated lines were lysed while being cultured in 10µM JQ1 while control lines were lysed while being culture in regular media.       β-actin A	 B	43		3.3 Epithelial–mesenchymal transition (EMT) does not induce resistance to JQ1 in LAC During the process of creating resistant lines it was observed that the H1975 line had a dramatic change in its morphology early in dose-escalation at 1µM JQ1. This change in morphology constituted a shift from an epithelial phenotype (as still seen in the control line) to a more mesenchymal phenotype for the resistant line (Figure 3.3A), indicating the line could be going through a process known as EMT (epithelial-to-mesenchymal transition). As this morphological change was shown to be permanent (Figure 6.1A) and EMT has previously been shown to be a cause of resistance to other inhibitors [60, 61], it was worth investigating if EMT could be playing a role in acquiring BETi resistance in either LAC line.  To establish if the resistant lines were going through EMT western blot analysis was performed, with mesenchymal and epithelial phenotypes being assessed through analysis of known EMT marker proteins [62]. As expected, the H1975 resistant line showed indications of a shift to a more mesenchymal phenotype, characterized by decreased levels of E-cadherin and β-catenin, and increased levels of vimentin (Figure 3.3B). The H23 resistant line also showed indications of transitioning to a more mesenchymal state, as β-catenin levels decreased and vimentin levels slightly increased in the resistant line (Figure 3.3B). However, the H23 line is already mesenchymal-like in both its morphology (Figure 3.3A) and in expression of mesenchymal markers N-cadherin and vimentin, and lack of expression of the epithelial marker E-cadherin. The H23 resistant line also showed decreased expression of N-cadherin (Figure 3.3B), indicating a shift to a more epithelial phenotype. Therefore, it is unclear if the H23 resistant line is transitioning towards a more mesenchymal or epithelial phenotype. 44		Nonetheless, as the H23 resistant line shows expression of some mesenchymal markers and the H1975 resistant line shows indications of shifting towards a more mesenchymal phenotype, it was worth investigating if induction of EMT could confer resistant to JQ1 in LAC. To test if the transition of an epithelial-to-mesenchymal phenotype could induce resistance to JQ1, H23 and H1975 parental lines were treated for 72hrs (data not shown) and two weeks with 10ng/ml TGFβ1, a cytokine known to induce EMT of epithelial lines [63]. Resistance to JQ1 was once again assessed by dose response experiments and induction of EMT was assessed morphologically and by western blot analysis, with the parental and resistant lines acting as benchmarks. For H1975, treatment with TGFβ1 showed induction of EMT both morphologically (Figure 3.3E) and in protein expression, paralleling the H1975 JQ1 resistant changes with decreased levels of E-cadherin and β-catenin and increased levels of vimentin (Figure 3.3F). However, TGFβ1 was unable to increase H1975 resistance to JQ1 treatment (Figure 3.3C), resulting in an IC50 value of 0.9µM ± 0.2 as compared to the H1975 parental having an IC50 of 1.2uM ± 0.3, and the resistant line having an IC50 >10µM (average of two replicates) (Figure 3.3D). TGFβ1 treatment of H23, similar to the H23 resistant line, showed no morphological differences between them and the H23 parental line (Extended Figure 6.2A), with the only indication of induction of EMT being decreased levels of β-catenin (Extended Figure 6.2B).  Again, TGFβ1 treatment did not lead to increased JQ1 resistance (Extended Figure 6.2C), with the TGFβ1 treated line having an IC50 value of 0.9µM ± 0.2 as compared to the parental value of 0.8µM ± 0.4, and the resistant line having a IC50 >10µM (average of two 45		replicates) (Extended Figure 6.2D). Therefore, these results suggest that EMT is not a driving force in acquiring resistance to BET inhibitors in LAC.                     46			 			 						 	      0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0H 1 9 7 5J Q 1  (u M )Relative Fluorescence to controlH 1 9 75H 1 9 7 5  T G F B 1H 1 9 7 5  R e s is ta n tCell line IC50 (µM) H1975 1.2 ± 0.3 H1975 TGFβ1 0.9  ± 0.2 H1975 Resistant >10 Resistant	Resistant	H1975	H23		Control	Control	H1975	 H1975	TGFβ1	treated	A	C	 D	B	E	 F	47		Figure 3.3: Induction of EMT is not sufficient to induce resistance to JQ1. Phase-contrast microscopy of H1975 shows a more mesenchymal morphology for the resistant line as compared to the epithelial morphology of the control line. These morphological differences are not seen in the H23 resistant line (scale bars = 200µm) (A). Western blot analysis of EMT marker proteins parallels the morphological differences as the H1975 resistant line shows expression and repression of proteins consistent with a shift towards a more mesenchymal phenotype, while H23 shows no clear shift (GAPDH acting as a loading control) (B). H1975 and H23 (Figure 6.3) parental lines were then treated with 10ng/ml of TGFβ1 for 2 weeks to induce EMT. Resistance was determined by 72hr dose response experiments (C) with IC50 values representing average of two experiments ± SEM. Induction of EMT was assessed morphologically using phase-contrast microscopy (scale bars = 400µm) (E) and western blot analysis of EMT marker proteins, with GAPDH serving as a loading control (F).                 48		3.4 Resistant lines are still dependent on BRD4 RNAi experiments were used to help verify that BRD4 was the main BET protein being targeted in the JQ1 sensitive lines in LAC [37], and that resistance was acquired to this inhibition. Through transfection of siRNAs, expression of a selected protein can be “knocked down” and its effects on cell viability measured. Confirmation of knockdown of the selected protein would then be verified by western blot analysis. As a positive control, siRNA towards Kinesin Family Member 11 (KIF11) – which is essential for mitosis [64]- was used for all lines, as well as KRAS and EGFR knockdown in H23 and H1975 lines, respectively, which target their respective driver oncogenes. It was important to use KIF11 as an additional positive control as it was unclear if the resistant lines would still be dependent on KRAS/EGFR. As expected both KIF11 and EGFR/KRAS knockdown significantly (p-value<0.001, two-way ANOVA, n=3) reduced the viability of both control lines. These results were also seen in both resistant lines except for EGFR knockdown in the H1975 resistant line, which showed no significant change compared to the NonT control. This indicates that the H1975 resistant line is much less dependent on EGFR signalling and suggests that this gene is no longer a driving force in the line. Western blot analysis confirmed successful knockdown of both KRAS and EGFR in H23 and H1975 control and resistant lines, respectively (Figure 3.4B and D). It should be noted that the H23 line has high expression of NRAS, a protein with a similar molecular weight to KRAS, and that the antibody used detects both NRAS and KRAS. However, KRAS (which has a higher molecular weight of 21.7kDa compared to NRAS at 21.2kDA) is still shown to have successfully knockdown through densitometry isolation of the top section of the KRAS/NRAS blot (Figure 3.4B).   49		  As expected BRD4 knockdown also significantly decreased viability of both control lines (H23 p-value <0.001, H1975 p-value <0.01, two-way ANOVA, n=3). Interestingly, BRD4 knockdown had similar effects on both resistant lines, with no significant increase in viability in either resistant line as compared to their controls (Figure 3.4A and C). In addition, there was still a significant difference in viability upon JQ1 treatment between control and resistant lines for both H23 and H1975, with no significant difference when grown in the control (0.1% DMSO) media (Figure 3.4A and C). This, along with verification of successful knockdown of targeted proteins (Figure 3.4B and D), suggests that LAC lines that have acquired resistant to JQ1 treatment may still be dependent on BRD4 for their survival. As JQ1 functions by binding to the two bromodomains of BRD4, therefore inhibiting its ability to bind to acetylated lysine residues, BRD4 must have a bromodomain independent role in both resistant lines necessary for their viability.     50			 		 	 Figure 3.4: JQ1 resistant lines dependencies on BRD4 knockdown mimic that of control lines. Relative viability following siRNA knockdown in H23 (A) and H1975 (C) resistant and control lines with western blot analysis to verify protein depletion (B)(D). Cell lines were transfected with the indicated siRNAs and viability was assessed 96hrs after transfection. Values are relative to the Non-target (NonT) control and represent the mean of triplicate experiments (error bars indicate SEM). KIF and KRAS/ EGFR act as positive controls. Asterisks denote level of statistical significance between the control and resistant line (*P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.001) using two-way ANOVA. Western blot lysates were collected 72hrs after transfection and analysed, with GAPDH serving as a loading control. N on T K IFK RA SB RD 4D MS O JQ105 01 0 0Relative ViabilityC on tro lR es is tan t* * * *n sn sn sn sN on T K IFE GF RB RD 4D MS O JQ105 01 0 0Relative ViabilityC on tro lR es is tan t* * ** * * *n sn sn sH23	H1975	H23	H1975	A	C	 D	B	siRNA: siRNA: 51		3.5 Increased phosphorylation of BRD4 by casein kinase 2 may induce resistance to JQ1  Phosphorylation of BRD4 (pBRD4) has previously been shown to link BETi resistance with cell dependency of BRD4 [52]. As phosphorylation of BRD4 cannot be measured at the RNA level this association could only be looked at through western blot analysis. Strikingly, both resistant lines showed considerable increased levels of pBRD4, with 28- and 11-fold more pBRD4 than controls for H23 and H1975, respectively (Figure 3.5A). Total levels of BRD4 were also assessed at the protein level as well as at the RNA level, through RNA expression analysis. BRD4 was only found to be up-regulated as compared to the control lines at the RNA level by probes that were specific for the long isoform of the protein (red arrows, Figure 3.5B). This BRD4 up-regulation was also seen at the protein level, using an antibody that specifically detects the 200kDa band, which represents the long isoform of the protein (2.4 and 1.3-fold more in H23 and H1975 resistant lines, respectively) (Figure 3.5A). Therefore, these results suggest that the resistant lines could be selectively increasing transcription of the long isoform of the protein, which may be preferentially phosphorylated.          52		H23	Resistant		Growth	inhibition	(%)	     -   10 100 500 JQ1 (nM)   10 100 500 CX4945 (nM)       -           10         100       500      1000     5000    Combination	Index	Scores				Growth	inhibition	(%)				CX4945 (nM)       10          100        500       1000      5000    JQ1 (nM) H23 Resistant   CX4945 (nM)       -          10         100        500      1000      5000      0  10 100 500  					 																																																										                      JQ1 (nM) A	C	 D	B	E	 F	53		Figure 3.5: pBRD4 and CK2 levels are elevated in both JQ1 resistant lines and may be targeted synergistically to induce death. Western blot analysis of pBRD4 and BRD4 levels, specific to the long isoform, as well as CK2α levels in JQ1 resistant and their control counterparts (A). Densitometry data (values below) are representative of protein levels compared to the specific lines control GAPDH, used as the loading control. Red values indicate pBRD4 levels compared to BRD4 levels for each line. RNA expression analysis shows increased expression in BRD4 and CK2α levels for resistant lines only, but no change in CK2α’ levels (B). Values represent fold changes as compared to the control line for each specific cell line. Red arrows indicate probes specific for the long isoform of BRD4. Western blot analysis of H23 resistant cells treated for 2hrs with the CK2 inhibitor CX-4945. pBRD4 long represents 150kDa bands as 200kDa bands were unable to be obtained, while pBRD4 short represents bands at 84kDa (C). Combination growth assays for H23 resistant line treated with a combination of JQ1 and CX-4945 for 72hrs. Image is one of two replicates (D). Cell viability was measured at the endpoint the percent inhibition calculated and plotted in plated format as compared to the control well (0, 0), with the mean value of two replicates being shown (E). CompuSyn was used to calculate combination index scores as a way of calculating synergy for two given doses, also shown in plate layout form (F). Legends for percent growth inhibition and combination index values shown at the bottom.	           E	54		Previously, the protein casein kinase 2 (CK2) has been implicated as being a kinase of BRD4 [52].  To determine the role of CK2 in phosphorylating BRD4 in resistant cell lines the two main genes that comprise the catalytic activity of the CK2 tetrameric kinase; casein kinase 2 alpha 1 (CK2α) and casein kinase 2 alpha prime (CK2α’) [65], were compared at the RNA and protein level. CK2α’ showed inconsistent results of being up or down-regulated at the RNA level in either resistant line. However, CK2α levels demonstrated an increase in both resistant lines at both the protein (Figure 3.5A) and gene expression level (Figure 3.5B). Therefore, though CK2α does not show equivalent increase as pBRD4 in resistant lines, its parallel increased levels suggest that CK2 may be associated with BRD4 phosphorylation.  To examine whether CK2 plays a role in phosphorylating BRD4 the inhibitor CX-4945, known to inhibit CK2 activity and currently in clinical stages [66], was used to treat H23 resistant cells and pBRD4 levels measured through western blot analysis (Figure 3.4C). As expected, pBRD4 levels decreased upon treatment with CX-4945. However, the most substantial decrease of pBRD4 was seen at the 84kDa band (short isoform), and only a slight decrease at the band around 150kDa (long isoform) with no clear bands being observed at 200kDa, which was previously detected as the main long isoform band (Figure 3.5A). Still, it is clear the long isoform of BRD4 and pBRD4 levels pertaining to 200kDa are increased in both resistant lines and that CK2 plays a role in phosphorylating BRD4, though further results are needed to confirm this exact link. Still, these results suggest that pBRD4 may be a common mechanism of resistance and provides a unifying explanation of how lines that are resistant to JQ1 treatment are still dependent on BRD4. 55		3.6 Combinational treatment with a CK2 inhibitor and JQ1 shows synergistic effects in resistant lines. As I now hypothesized that resistance to JQ1 in LAC could be a result of increased levels of pBRD4 due to CK2 phosphorylation it was of interest to test whether these resistant lines could be re-sensitized to JQ1 through inhibition of this phosphorylation. To test this, different concentrations of JQ1 and CX-4945 we used in combination in both resistant and control lines in 72hr and 10-day combination growth assays. 72hr assays were performed in duplicate (Figure 3.5 and Extended Figure 6.3) while 10-day assays were performed in triplicate (Figure 3.6 and Extended Figure 6.4). Cell viability was measured on the last day (72hr or 10th day) before plates were stained using crystal violet for a visual representation (Figure 3.5D). Percent growth inhibition relative to the control wells were then calculated and the average between replicates shown in plate layout format (Figure 3.5E). To calculate the effects of inhibitor combinations CompuSyn was used, a program based on the Chou-Talafay principle [55], which calculates a combination index (CI) score for each different inhibitor combination. The lower the CI value the more synergistic the effect of the two inhibitors are, with values between 0 and 0.75 indicating synergy, 0.75 and 1.25 indicating additive effects, and values greater than 1.25 signifying antagonistic effects of the two inhibitors [55]. To better visualize trends CI scores were plotted in plate format (Figure 3.5F).  A synergistic effect can be seen in the 72hr combination assays for the H23 resistant line as CI values <0.75 become more abundant as the concentration of CX-4945 and JQ1 increase (Figure 3.5F). This kind of synergistic trend was also seen in the 10-day combinational treatment assays for the H23 resistant line, with all combinational wells resulting in CI values less than 0.75, indicating synergy (Figure 3.6A). Together, these 56		results indicate that CX-4945 and JQ1 act synergistically on the H23 resistant line, providing further evidence that pBRD4 may be a driving force of resistance.  For the H1975 resistant line 72hr combination growth assays (Extended Figure 6.3A) showed more variation between wells with only a few demonstrating CI values indicative of synergy and more indicating additive effects. However, 10-day assays on the H1975 resistant line revealed a more synergistic outcome, with the majority of treatment combinations showing synergy (Extended Figure 6.4A). Similar observations were made for the H1975 control line for the 72hr assays in that the two drugs did not show a synergistic trend (Extended Figure 6.3B), though there was some synergy in the 10-day assays (Extended Figure 6.4B) However, CI values were unable to be calculated for this line as the 1000nM CX-4945 control well showed less growth inhibition than the 500nM well, resulting in a positive slope in CompuSyn. Still, from the 100nM JQ1 wells it can clearly be seen that both 500nM and 1000nM CX-4945 treatment causes considerable increase in growth inhibition as compared to the control wells (Extended Figure 6.4B). Overall, while greater variations exist across the combinations for H1975 than seen in the H23 resistant and control lines, data from the 10-day assays show synergistic trends for both lines. Lastly, synergy was also observed between the two inhibitors in the H23 control line, with CX-4945 and JQ1 having a synergistic effect in both the 72hr (Extended Figure 6.3C) and 10-day (Extended Figure 6.4C) combination growth assays. The 72hr plates showed that as both concentrations of inhibitors increased the CI values increased in synergy as well. Overall, these results suggest that CX-4945 and JQ1 have a synergistic effect in both acquired resistant and JQ1 sensitive lines and indicate that pBRD4 may be important beyond a role in JQ1 resistance. 57		3.7 Primary JQ1 resistant LAC line is sensitized through inhibition of CK2 As there are many LAC lines resistant to JQ1 treatment (primary resistant lines) it was hypothesized that identification of dependencies of acquired JQ1 resistance could help elucidate mechanisms of resistant in these primary resistant lines and identify possible ways to sensitize these lines to JQ1 treatment. Therefore, it was of interest to see if inhibition of CK2 would sensitize primary resistant lines to JQ1 treatment. To test this, the primary resistant line H2030 was selected, which is a JQ1 resistant LAC line with a KRASG12C mutation. To verify H2030 as being resistant to BET inhibition, 72hr does response experiments were used with both JQ1 and I-BET762 (Extended Figure 6.5A and B), with both results showing IC50 values >10µM and >20µM, respectively, similar to the values observed in the acquired resistance lines. To assess if H2030 could become sensitized to JQ1 through CK2 inhibition, 10-day combination growth assays were completed. Although there was some variation across the three biological replicates, the trends of all three experiments remained the same and showed clear synergistic effects (Figure 3.6B). This observation of synergy was confirmed through Compusyn CI values with every well showing synergy between CX-4945 and JQ1. Therefore, it seems clear that a primary resistant line can become sensitized to JQ1 treatment through inhibition of CK2, possibly due to inhibition of phosphorylation of BRD4.       58		H23 Resistant  H2030 Combination	Index	Scores				Growth	inhibition	(%)				      -   100   500 2500  -            500         1000           CX4945 (nM)         -            500         1000                 -   100   500 2500  CX4945 (nM)      -           500       1000          -         500       1000                 -   100   500 2500        -   100   500 2500  CX4945 (nM)       500          1000                    100   500 2500           100   500 2500  500          1000            					 											          Figure 3.6: Synergistic trends are also seen in a primary JQ1 resistant line.  10-day combination growth assays for the primary resistant line H2030 (B) and the acquired resistant line of H23 (A). Plate image is one of three replicates (top). Percent growth inhibition was calculated as the mean of replicates as compared to the control well (middle). CI values, as calculated using CompuSyn, show only synergistic values for both lines (bottom). Legends for percent growth inhibition and combination index values shown at the bottom. JQ1 (nM) A	 B	59		3.8 AXL and SPOCK1 identified as possible targets of pBRD4 and mediators of JQ1 resistance While pBRD4 appears to play a role in resistance to JQ1, the genes regulated by this transcriptional regulator in JQ1 resistant lines are unknown.  As has been previously shown, both acquired resistant lines are independent of MYC or re-expression of FOSL1; therefore, the genes linked to resistance remain elusive. To help illuminate what genes may be regulating JQ1 resistance genome-wide RNA expression profiles were generated and analyzed, with the goal of identifying common genes in both resistant lines. For each line, resistant and 6hr JQ1 treated groups were compared to control lines and genes were identified as candidates if they demonstrated at least two-fold up or down-regulation and a BH corrected p-value<0.005 (unpaired t-test). For the H1975 line 4358 and 168 genes were identified for the resistant and 6hr treatment, respectively, that passed these cut-offs. For the H23 line 429 and 1249 genes were identified for the resistant and 6hr treatment, respectively. Within each line, the resistant and 6hr treatment gene sets were then compared to each other. This allowed selection for deregulated genes only found in the resistant lines and not due to exposure to JQ1. This left 4277 deregulated genes in the H1975 resistant line and 332 deregulated genes in the H23 resistant line. Heat-maps were created for the top 20 up and down-regulated genes from each of these lists (Extended Figure 6.7), in hopes of identifying common deregulated genes between the two lines. However, no common genes were identified. Instead, the two resistant gene sets were compared to each other and from this comparison 101 common genes between the two resistant lines were identified (Figure 3.7A). From this list 70 genes were identified as being analogously regulated, genes which are up or down-regulated in both resistant lines, and 31 conversely regulated, genes which are up in one line and down in the other 60		or vice versa (Full gene lists shown in Extended Tables 6.1 and 6.2, respectively). To try and identify significant genes that the resistant lines may be dependent on the top 10 up and down-regulated genes from the 101 common genes were plotted in a heat-map for each resistant line (Figure 3.7B). Strikingly, the top up-regulated gene for each line (AXL and SPOCK1 for H1975 and H23, respectively) was four and five-fold greater than the second up-regulated gene in that line. These genes therefore presented possible dependencies in their respective resistant lines, as it was hypothesized that the resistant lines could be up-regulating these genes to signal through different pathways as an adaptive mechanism to JQ1. To test this hypothesis 72hr does response experiments were used to test the sensitivity of the resistant line as compared to its control line to inhibitors that targeted the up-regulated genes pathway or predicted pathway. For AXL (a receptor tyrosine kinase), the inhibitor Crizotinib was used (Figure 3.7D) as it shows specificity for this kinase as well as ALK, which is commonly translocated in subsets of LAC [67]. For SPOCK1, the inhibitor BEZ-235 was used (Figure 3.7F) as it targets the Pohsphoinositide-3 Kinase (PI3K) pathway, one of the main pathways SPOCK1 has been shown to regulate [68, 69]. In both cases the inhibitors had a much greater effect on the resistant line as compared to the control (Figure 3.7D & F). For Crizotinib, the H1975 resistant line had an IC50 of 0.35µM ± 0.05 while the control line had an IC50 of 2.3µM ± 0.7 (Figure 3.7E). For BEZ-235, the H23 resistant line had an IC50 of 19nM ± 0.6 while the control line had an IC50 of 70nM ± 30 (Figure 3.7G). In addition, AXL showed decreased protein levels upon BRD4 knockdown, with 0.55-fold less than the NonT control (Figure 3.4D), suggesting that pBRD4 may be regulating its expression. However, AXL levels also decreased (0.51-fold) upon EGFR knockdown. Unfortunately, 61		SPOCK1 bands were unable to be obtained during western blot analysis of knockdown lysate and are something that is currently in progress. Overall these results suggest that AXL and SPOCK1 are up-regulated genes that the H1975 and H23 resistant line are dependent on, respectively, and that pBRD4 may be regulating the expression of these genes in the resistant lines.                               62		Crizotinib BEZ-235 															 	 	 	 	 	 																								      Cell line IC50 (µM) H1975 Control 2.3 ± 0.7 H1975 Resistant 0.35 ± 0.05 	 		 	Cell line IC50 (µM) H23 Control 0.07 ± 0.02 H23 Resistant 0.019 ± 0.0006 H1975	H1975	H23	 H23	H1975	H23	A	 B	C	 D	E	F	 G	Top 10 up and down-regulated genes for each line from 101  101	common	resistant	differentially	regulated	genes	63		Figure 3.7: RNA expression analysis identifies AXL and SPOCK1 as highly up-regulated genes in H1975 and H23 resistant lines, respectively. Schematic representation of RNA expression analysis used to identify differentially regulated genes common to the two resistant lines, 101 which were identified. Briefly, for H23 and H1975 resistant and 6hr JQ1 treated lines were compared with control line and only genes that were differentially regulated (≥2 fold with a BH corrected p-value<0.005 (unpaired t-test)) were retained. Comparison of these two gene sets for each line yielded 4277 and 332 resistant only differentially regulated genes for H1975 and H23, respectively, which were further compared to identify common differentially regulated genes in JQ1 resistance. From this, 101 genes were identified (70 analogously regulated and 31 conversely regulated) and the top 10 up and down-regulated genes from this 101 list are shown in (B). Interestingly, the top genes for each line are both conversely regulated between the two resistant lines (C). 72hr does response assays were used to assess inhibition of identified gene specific to H1975 (D) and H23 (F) resistant lines.  IC50 values are given for H1975 (E) and H23 (G) and represent means from two replicate experiments (± SEM). Graphs are representative of one of the two experiments.                       64		Chapter 4: Discussion  4.1 LAC presents a unique opportunity to study mechanisms of BETis The function of the BET proteins and their inhibitors (BETis) has recently been shown to be important in the context of many different types of cancer. Of these BRD4 often plays the most important role and inhibition of the BET family of proteins is often solely discussed in terms of BRD4 inhibition. Mechanistically, BETis inhibit the ability of BRD4 to bind to acetylated lysine residues of histones, and through this, limit its recruitment of P-TEFb. Therefore, genes dependent on P-TEFb recruitment for phosphorylation of RNA pol II to advance it from initiation to elongation during transcription would have their expression inhibited by BETis activity. For many cancers, the main inhibitory effect of BETis was found to be down-regulation of MYC, an oncogenic transcription factor and driving force in some cancers, especially hematological malignancies where translocations activating these genes are common [32, 36, 42, 49, 50]. However, as research has accumulated it has become evident that the role of BET proteins, specifically BRD4, is quite varied in different cell types, suggesting that the epigenetic landscape of a cell could impart the role of these inhibitors. This variation in different cell types was first shown in previous work of our lab, where LAC lines were sensitive to BETis independent of MYC down-regulation [37]. Further work identified FOSL1, itself an oncogenic transcription factor [70], as being down-regulated due to BETi treatment and subsequent knockdown of the FOSL1 protein mimicking BETi effects in sensitive and resistant lines. However, over-expression of FOSL1 was not sufficient to “rescue” sensitive lines from treatment, indicating that it may not be the only important down-stream target of BETis in LAC.  This mechanism of action was 65		duplicated in osteosarcoma a few years later [38], with cells displaying sensitivity to BETis independent of MYC and finding that FOSL1 down-regulation could partially be a cause of sensitivity in these lines.  However, there are many lines in these cancer types initially resistant to JQ1 treatment, termed “primary resistant” lines. Unfortunately, resistance or sensitivity is not linked with any certain driver mutations or mutational profiles and therefore further work is needed to identifying common resistance mechanisms to JQ1 that may be used to better select and treat patients in the future.  One way to study and better understand effects of inhibitors in a cell is through identifying mechanisms of resistance [52]. This has been done several times already for BETis in cancers such as breast [52], AML [32] and PDAC [33]. However, one thing that all these cancers have in common is that they are inhibited by BETis through MYC down-regulation. Interestingly, all these cancers were found to acquire resistance through re-expression of MYC, though not through the exact same mechanism. For instance, in AML Fong et al. showed that human and mouse leukemia cells demonstrated resistance to BETis through increased Wnt pathway activation, which leads to β-catenin binding to DNA and re-activation of MYC expression [53]. In addition, these resistant cells were able to be re-sensitized to BETis upon inhibition of the Wnt pathway. In PDAC, Kumar et al. observed a different mechanism of resistance by which the hedgehog signaling pathway protein GLI2 co-regulated re-expression of MYC, and loss of GLI2 re-sensitized cells to BETis [33]. However, resistance to BETis has never been studied in the context of a MYC independent cancer. This presented an opportunity to better elucidate BETis function in different cell contexts and identify how the epigenetic landscape of a cell can influence sensitivity or resistance to an epigenetic inhibitor. 66		4.2 Established resistant lines are resistant to BET inhibition and not off-target or multi-drug effects The establishment of resistance to BET inhibitors in two initially sensitive lines entailed a 6-month process of culturing cells in increasing concentrations of the BET inhibitor JQ1 (Figure 3.1A). JQ1 was the main BETi used in this thesis as it had been the primary BETi used in our labs previous work [37] as it most frequently used in other in vitro studies including studies on BETi resistance [22]. In addition, this mechanism of acquiring resistance is quite common and has been used in multiple other studies [33, 52, 65]. Once cells were able to grow in 10µM JQ1 media they were considered resistant, as this had been the cut-off in previous work by our lab [37].  The two sensitive LAC lines, H23 and H1975, were chosen as they contained a KRAS and EGFR driver mutation, respectively, the two most common types of mutations in NSCLC (Figure 1.1) and specifically in LAC.  Resistant lines were confirmed to be isogenic resistant counterparts through STR profiling (Extended Figure 6.1D) and resistance was confirmed to be due to BET protein inhibition and not off-target effects of JQ1 through dose response experiments using another BET inhibitor, I-BET762 (Figure 3.1E).   Looking at the resistant lines it may be noticed that there appears to be a decrease in cell viability when treated with high concentrations of JQ1 and I-BET762. This is evident by fewer cells in the 1µM JQ1 wells as compared to the DMSO control wells of the long-term growth assays (Figure 3.1B). Furthermore, both resistant lines have decreasing slopes in both the JQ1 (Figure 3.1C) and I-BET762 (Figure 3.1E) dose response experiments. However, it was observed that this decrease in cells was not due to increased apoptosis upon JQ1 treatment in the H23 resistant line, unlike the H23 control 67		line which showed increased apoptosis upon 2µM and 10µM JQ1 treatment (Figure 3.1G). This suggests that the decrease in number of cells and Alamar Blue metabolism is likely the result of cells having a slower growth rate in high concentrations of JQ1. This was also observed in a large panel of luminal breast cancer lines where lines with high JQ1 sensitivity underwent apoptosis while resistant lines had slower cell cycle progression and therefore multiplied at a slower rate [71]. Our lab has previously shown in LAC that JQ1 causes increased apoptosis and cell-cycle arrest in sensitive lines, and does not induce either apoptosis or cell-cycle arrest in a primary resistant line [37]. Here we have only assessed the induction of apoptosis in one of the acquired resistant lines. Given that previous papers have reported differences in JQ1 inducing apoptosis or cell-cycle arrest, or both [38, 51, 65, 72], it may be of interest to investigate if resistant lines are progressing through the cell-cycle at a slower rate and if this (and lack of apoptosis) is consistent between both acquired JQ1 resistant lines. Another common mechanism of resistance is the up-regulation/ over-expression of efflux pumps/ transporters, which can limit the intracellular concentration of drug and can give the allusion of resistance. The RNA expression data supported this as a possible mechanism of resistance as the cell-membrane transporter ABCA3 (ATP-binding cassette sub-family A member 3) was identified as one of the top 10 up-regulated genes common to both resistant lines (Figure 3.7B). In addition ABCA3 has previously shown to confer multidrug resistance in leukemia [73]. However, dose response experiments, used later for targeting certain genes and pathways, never showed increased resistance to any of the inhibitors used (Trametinib, Afatinib, Crizotinib, and BEZ-235), even if the drug was non-specifically targeting that line (Extended Figure 6.6). If efflux pumps were integral 68		to the mechanism if resistance it would be expected that this would lead to increased resistance to a variety of inhibitors. Still, we are currently looking to support this claim through knockdown of ABCA3 using RNAi experiments and identifying if this contributes to resistant lines sensitivity to BETis.   4.3 Acquired resistance to BETis in two LAC lines is independent of MYC or FOSL1 Similar to our labs previous work [37] both JQ1 sensitive lines showed increased levels of MYC and decreased levels of FOSL1 upon JQ1 treatment (Figure 3.2, 6hr). This confirms that JQ1 sensitivity in LAC is MYC independent and likely to be at least partly due to the down-regulation of FOSL1. Since sensitive lines showed FOSL1 down-regulation upon JQ1 treatment and our lab has previously shown decreased FOLS1 levels  mimic JQ1 treatment in LAC lines [37] it was predicted that FOSL1 would be re-expressed to acquire resistance to BETi treatment similar to resistant mechanisms in MYC-dependent cancers [32, 33, 53]. Interestingly, both resistant lines did not show re-expression of FOSL1 at the RNA or protein level (Figure 3.2A and B, respectively). It should be noted that for the RNA expression data and subsequent blots the resistant lines were cultured in 10µM JQ1 media and the control lines in 0.1% DMSO media prior to lysis. This may explain why in Figure 3.1G FOSL1 levels are present in the JQ1 untreated H23 resistant lane (0µM) while are absent in the western blot in Figure 3.2B. However, FOSL1 levels do disappear in the H23 resistant line upon JQ1 treatment (Figure 3.1G) indicating that FOSL1 likely does not play a role in resistant to BETis in these lines. These results suggest that the down-regulation of FOSL1 may not be the only 69		downstream target of BET inhibition in LAC and that other genes may play a role in BETis function in sensitive lines. This is additionally supported by the fact that over-expression of FOSL1 was unable to  rescue sensitive LAC lines from JQ1 treatment [37]. Overall, it seems likely that both resistant lines have acquired resistance to BETis independent of MYC or FOSL1, suggesting a novel mechanism of resistance in LAC.  4.4 EMT does not confer resistance to BETis in LAC EMT is the process by which epithelial cells shift towards a more mesenchymal phenotype and can be seen at a morphologically level with change from a polarized, epithelial shaped cell to a more spindle shaped cell [62]. This transition is often associated with invasion and metastasis and has been linked to resistance of both chemotherapy and other targeted therapies [60, 61]. In addition, BRD4 has been linked to playing an important role in EMT and metastasis, though exact mechanisms are yet to be elucidated in detail [30]. This transition from an epithelial cell to a mesenchymal one can also be characterized based on the expression of certain genes, termed “EMT marker” genes. For instance, the loss of certain genes; such as E-cadherin and β-catenin, or the increase of others; such as N-cadherin and vimentin, are indicators of changing from a epithelial cell to a more mesenchymal phenotype [62]. During acquisition of JQ1 resistance H1975 went through EMT as evident by the changes to its morphology and protein expression (Figure 3.3A & B). However, the H23 line, which is EMT-like, showed no morphological change upon JQ1 resistance acquisition. In addition, protein markers were conflicting in the H23 resistant line as decreased β-catenin levels indicated EMT, while N-cadherin levels also decreased suggesting a shift to a more epithelial 70		morphology. As EMT features were present in both resistant lines and EMT has been linked to resistance to other targeted therapies, such as tyrosine kinase inhibitors [60], I examined whether induction of EMT could confer resistance to BETis in LAC sensitive lines.  To induce EMT H23 and H1975 parental cells were treated with TGFβ1, which has previously been shown to induce EMT in epithelial lines by binding to serine/ threonine transmembrane kinase receptors [63, 74]. Treatment with 10ng/ml TGFβ1 over a two-week period showed indications of cells transitioning to a more mesenchymal phenotype, as evident by increased levels of vimentin (H1975) and decreased levels of β-catenin (H1975 and H23) (Figure 3.3E), as well as a morphological change in the H1975 treated line resembling the phenotype of the H1975 resistant line (Figure 3.3 F). However, TGFβ1 treatment did increase resistance to JQ1 in either treated line (Figure 3.3C and Extended Figure 6.2C), suggesting that this transition is not the driving force in BETi resistance. In addition, there is no trend in LAC lines of primary JQ1 resistant lines having a more mesenchymal phenotype and sensitive lines having more of an epithelial phenotype. Therefore, these results conclude that EMT does not induce BETi resistance in LAC lines and is likely a passenger phenotype of the resistant lines.            71		4.5 JQ1 resistant cell lines remain dependent on BRD4 To validate BRD4 as the primary BET protein targeted by JQ1 and that the resistant lines had acquired resistance to this proteins inhibition RNAi experiments were performed. It was expected that the acquired resistant lines would be resistant to BRD4 knockdown, as this had been previously seen in other acquired resistant lines [33]. In addition, BRD4 is the main target of the inhibitor the cells were now resistant to. I presumed that resistant lines would adapt to down-regulation of genes normally under the control of BRD4 as it would be unable to localize to acetylated lysine sites due to JQ1 binding to the bromodomains. Resistance could involve mechanisms to cause re-expression of down-regulated genes [32, 33] or could involve increased signalling through other pathways to circumvent previous signaling pathways [65]. In either case, BRD4 would be predicted to not play a role in the resistant lines. However, this was not observed as BRD4 knockdown in both resistant lines significantly inhibited the viability of the cells as compared to the NonT control (p value<0.001 for H23 and <0.005 for H1975) and furthermore, BRD4 knockdown showed no significant difference between the resistant and control lines (two-way ANOVA, n=3 for both) (Figure 3.4A and C). This suggests that BRD4 must still be important and functioning in the resistant lines independent of its ability to bind to acetylated lysine residues of histone tails.  A similar result was described during acquired BETi resistance in a TNBC line that was published during the course of this study [52]. The authors discovered that the TNBC line acquired resistance through phosphorylation of BRD4 by CK2. This allowed BRD4 (now pBRD4) to bind to the mediator protein MED1, resulting in BRD4 being able to once again regulate the expression of MYC independent of its acetylated lysine-72		binding function. This mechanism of resistance therefore explained how BETi resistant lines could still be dependent on BRD4. Still, with this mechanism it would be expected that BRD4 would localize back to its initial regulatory gene targets in sensitive cells, which is not the case here given the lack of FOSL1 re-expression. However, it was not fully explained why MED1 would be specific to areas that BRD4 previously bound to due to specific acetylated lysine residues. One possible explanation is that BRD4 has shown to localize to SE regions [26, 43], and that this localization is often accompanied by the mediator complex during initiation of transcription, of which MED1 is a part of. However, it is still unclear if the mediator complex acts in co-localization to SE regions or is recruited to these regions by BRD4. In addition, the phosphorylation of BRD4 has only recently been identified and though phosphorylation of BRD4 has shown to cause binding of different proteins, such as p53 [28], its exact role in the cell is still largely unknown [26, 28]. Therefore, further work is needed to determine the role of MED1 in localizing pBRD4 to areas previously regulated by BRD4 and if these binding partners change upon phosphorylation. However, as pBRD4 links why BETi resistant lines are still dependent on BRD4 I decided to explore this further as a possible mechanism of resistance for LAC lines.  4.6 Phosphorylation of BRD4 by CK2 is a possible mechanism of acquired resistance in LAC lines Through western blot analysis and densitometry calculations pBRD4 levels were compared between control and resistant lines (Figure 3.5A). Strikingly, pBRD4 levels were greatly enhanced (11 and 28-fold) in both H23 and H1975 resistant lines, 73		respectively. BRD4 levels were also found to be elevated in both resistant lines through RNA expression analysis, but only when detecting the long isoform of the protein (Figure 3.5B, red arrows). This suggests that the resistant lines could be positively selecting this isoform to acquire resistance to JQ1 and that this isoform may be preferentially phosphorylated. Further work is needed to identify if this is true, however, it is still clear that phosphorylation of BRD4 is enhanced in both resistant lines as H23 and H1975 resistant lines have 8.9 and 12-fold greater pBRD4 levels as compared to their controls even when BRD4 levels are accounted for (Figure 3.5A, red values)  As CK2 has previously been identified as a kinase of BRD4 [28, 52] its protein and RNA expression levels were also compared between control and resistant lines. CK2 exists as a tetrameric complex that contains two regulatory subunits (CK2β) and can contain two catalytic subunits (CK2α and CK2α’) in either a homozygous or heterozygous composition [66, 75].  Interestingly, CK2α was shown to be up-regulated in both resistant lines at both the protein and RNA level (Figure 3.5A & B, respectively), while the CK2α’ subunit did not show any concise results at the RNA level. However, due to the possible homozygous nature of the CK2 complex the increased levels of CK2α in both resistant lines could correspond to increased pBRD4 levels in these lines. To test if CK2 could be regulating pBRD4 levels H23 resistant cells were treated with CX-4945, an inhibitor previously shown to decrease CK2 kinase activity [66, 74], and previously shown to reduce pBRD4 levels in the TNBC acquired resistant line [52]. Interestingly, CK2 levels also decreased upon CX-4945 treatment, which was a little unexpected as cells were only treated for 2hrs and CX-4945 is known to only block the kinase activity and has been shown to not decrease CK2α expression (67.2µM for 1-24hrs) [66]. 74		However, CX-4945 treatment in another study (10µM for 72hrs) did result in decreased CK2α protein levels similar to the results seen here [74].  Upon treatment with 10µM, 20µM and 40µM of CX-4945 for 2hr it was observed that pBRD4 levels did indeed decrease (Figure 3.4C) as hypothesized and previously shown [52]. However, this decrease was only observed for the 150kDa band (showing a slight decrease), and the 84kDa band, thought to represent BRD4-SF (Figure 3.5C). The pBRD4 200kDa band, which was used in the rest of this thesis to represent BRD4-LF and pBRD4-LF, was not detectable. This problem had arisen before with the 200kDa band being unable to be detected while still detecting the other two bands during western blot analysis. This could be a result of protein degradation during lysate sample preparation, and is something that I am currently investigating. However, it is worth noting that a similar result was seen in the TNBC paper [52]. In their results of pBRD4 all immunoblots focused on the 200kDa band except after CX-4945 treatment, as the 200kDa bands were unable to be detected and instead the lower bands were used to represent pBRD4, which for them was around 130-140kDa [52]. Also, it is unclear if they observed any bands around 84kDa as this region was not presented. Overall, the results of the TNBC paper support the indication that CX-4945 treatment is able to decrease phosphorylation of BRD4 in the resistant. However, further work is needed to validate this connection between CK2 and phosphorylation of BRD4 in the context of LAC lines. For example, the phosphatase PP2A was reported to de-phosphorylate pBRD4 and that inhibition of PP2A by siRNA knockdown could lead to increased levels of pBRD4 [52]. In addition, PP2A was shown to be activated by phenothiazine (PTZ) treatment and this 75		resulted in decreased pBRD4 levels [52]. Therefore, further experiments like these are needed to validate pBRD4’s role in acquired BETi resistance in LAC lines.   4.7 CK2 inhibition synergizes with JQ1 treatment in JQ1 sensitive and acquired resistant lines To help evaluate if CK2 could be phosphorylating BRD4 and this being a mechanism of resistance to BETI treatment I decided to use a combinational assay approach, using the two inhibitors CX-4945 and JQ1. If phosphorylation of BRD4 was important for resistant lines it would be expected that decreasing these levels through inhibition of its kinase (CK2) would re-sensitize the cells to BETi treatment. For the H23 resistant line this was exactly what was seen in both the 72hr (Figure 3.5D) and 10-day (Figure 3.6A) combination growth assays. Mechanistically, using the 72hr 500nM CX-4945 and JQ1 treated well as an example, what is predicted to have occurred is that by inhibiting CK2, which has minimal effects on growth inhibition itself (3%), the cells are more sensitized to JQ1 treatment, causing 12% growth inhibition to increase to 34% (Figure 3.5E).  Therefore, it would be expected that the CI score for this well would be less than 0.75, indicating a synergistic relationship. However, synergistic values in a well or two are not enough to infer an overall synergistic effect of the two inhibitors. Instead, a synergistic effect is better exemplified by increasing synergistic values (decreasing CI scores) as both inhibitor concentrations increase, with the most indicative wells having some inhibition but not a complete lack of cells [76]. This synergistic effect is better exemplified in the H23 resistant line 72hr combination assay as CI values decrease as the 76		concentration of both drugs increase (Figure 3.5F). In addition, synergy between the two inhibitors is even more present after 10-day treatment as all CI values are less than 0.75, as calculated using CompuSyn software (see materials and methods for full details). For the H1975 resistant synergy between the two inhibitors was only seen with certain inhibitor combinations and this synergy was more apparent in the 10-day combinations plates (Extended Figure 6.4A) than the 72hr combination plates (Extended Figure 6.3A). Overall, these results suggest that CK2 inhibition may re-sensitize acquired resistant cells to BETi treatment and provides further evidence that CK2 may be phosphorylating BRD4 and that this is important in acquiring resistance to BET inhibition in LAC. Synergy was also observed in the H23 control line (Extended Figure 6.3C and 6.4C) and to a lesser extent in the H1975 control line (Extended Figure 6.3B and 6.4B). This was unexpected as it was expected that the inhibition of CK2 would only have an effect on the resistant lines as they both elevated levels of pBRD4 (Figure 3.5A). This result suggests that there could be basal activity of pBRD4 in control lines and that inhibition of this could enhance the action of JQ1 even in sensitive lines. In addition, pBRD4 levels were present for BRD4-SF and the 150kDa band of BRD4-LF in equal amounts between control and resistant lines (data not shown), suggesting a downstream target of CK2 inhibition, as previously shown (Figure 3.5C) and providing an explanation of why CX-4945 would synergize with JQ1 in JQ1 sensitive lines. However, this result could also be due to CK2 having other roles in the cell, as CK2 has been shown to phosphorylate Akt [66] as well as target proto-oncogenes, such as MYC and JUN, and transcriptional activators, such as NF-	κB [75]. In addition, CK2 has been implicated in tumor development, proliferation, cellular transformation, and apoptosis suppression in 77		cancers and has been documented as overexpressed in several cancers including; kidney, prostate, and even the lung [66, 75]. Lastly, it should be noted that there are some variations in these combinational assays due to multiple media/drug renewal, leading to negative growth inhibition and resulting in antagonistic CI values for some wells. However, CI values that result from <5% growth inhibition fall into the uncertainty of the assay and should not be taken as significant.  More importantly, the overall trends remained the same between replicates with certain combinations of inhibitors always having a resounding synergistic impact on growth inhibition. Overall, these results illustrate that JQ1 synergizes with CX-4945 in both JQ1 sensitive and acquired resistant LAC lines, though further experiments are needed to better elucidate if this inhibition is due to decreased levels of pBRD4 and better understand what the 150kDa and 84kDa bands are representing.    4.8 Synergistic effects are also present in a primary JQ1 resistant LAC line An initial goal of this thesis was to identify mechanisms of resistance that might be integral for primary resistant lines. I therefore continued investigating if pBRD4 may be important for these lines and if the CK2 inhibitor CX-4945 might sensitize these cells to JQ1 treatment. Unfortunately, problems arose in identifying the 200kDa band for pBRD4, although the 150kDa and 84kDa bands were again detected at similar levels to the other resistant and control lines. Despite this, combinational growth experiments showed synergy between the two inhibitors (Figure 3.6B), suggesting that CK2 inhibition is able to sensitize a primary resistant LAC line to JQ1 treatment; hypothesized to be due 78		to decreased levels of BRD4 phosphorylation. Overall, these results indicate that a combinational approach inhibiting CK2 kinase activity in conjunction with BETi treatment could lead to better patient response and a broader range of patients responding to BETi therapies.   4.9 RNA expression analysis identifies AXL and SPOCK1 as possible targets of pBRD4 in H1975 and H23 resistant lines Previous results in this thesis have indicated that pBRD4 plays a role in resistance to BETis; however, it does not clarify what genes may be regulated by phosphorylation of the long isoform of BRD4, as the initial down-regulated gene FOSL1 is not re-expressed in either resistant line. To investigate this genome-wide RNA expression analysis was conducted, with the goal of identifying common genes in both resistant lines. The RNA expression data was analysed by selecting genes with at least 2-fold increase or decrease only in the resistant lines as compared to their controls, and a BH corrected p-value<0.005 (unpaired t-test). These gene lists were then compared and from this 101 candidate genes were identified as being common between the two resistant lines (Figure 3.7A). Of these, the top 10 up and down-regulated genes were plotted (Figure 3.7B). Interestingly, it was observed that the top up-regulated gene in each line (AXL and SPOCK1 for H1975 and H23, respectively) were more than four and five-fold higher than the second most up-regulated in that line. I therefore hypothesized that the resistant lines could have become dependent on signaling through these genes or their associated pathways. Dose response experiments using Crizotinib to target AXL [67] in H1975 and BEZ-235 to target the PI3K pathway, which SPOCK1 has shown to play a role in [68, 79		69], both showed selective inhibition of the resistant line (Figure 3.7D and F). These results supported the previous hypothesis indicating that the resistant lines may have become dependent on these genes/ pathways upon acquiring resistance to BETis. In addition, the over-expression of both of these genes has previously been shown to induce EMT as a secondary phenotype in lung cancer [77, 78] as well as in several other cancers [69, 79, 80]. This supports my results that a transition to a more mesenchymal phenotype is not driving resistance to BETi treatment but instead suggests that this change in phenotype may be a result of overexpression of other genes required for acquired resistance.  What is interesting; however, is the fact that both these genes are conversely regulated between the two lines, with AXL being up-regulated in H1975 and down-regulated in H23, and SPOCK being up in H23 and down in H1975 resistant line (Figure 3.7C). This contradicts the notion of a common mechanism of resistance as it would be expected that a gene important to acquiring resistance would be analogously regulated in the two resistant lines. However, as BRD4 is an epigenetic protein with regulation of genetic transcription at a global level [30], and has been shown to change binding partners upon phosphorylation [28], it could be postulated that co-regulators like the mediator complex may influence cell specific gene regulation by BRD4. For instance, there may be a select number of common genes regulated by BRD4 in every cell, possibly in proximity to regions of super enhancers, but that this regulation differs between lines in the requirement of co-regulators for their transcription and the requirement of phosphorylation of BRD4 to bind to these co-regulators. In the context of the acquired resistant lines, this hypothesis would suggest that under normal conditions 80		AXL would be regulated by BRD4 in the H23 line whereas in the H1975 line AXL would be require co-regulators for transcription and that binding to these regulators would require phosphorylation of BRD4. Therefore, upon BETi treatment the expression of AXL would be inhibited in the H23 line as BRD4 would be unable to bind in its normal fashion while AXL expression would increase in the H1975 lines as this regulation would be acetylated-lysine binding independent. This hypothesis would also explain why JQ1 sensitive cells would be additionally sensitive to combinatorial treatment with a CK2 inhibitor as this hypothesis would suggest that BRD4 may be phosphorylated at basal levels in the cell and that this phosphorylation may be important for regulation of certain genes under normal conditions or as a mechanism to compensate for conditions of stress. Though this mechanism of action is speculative at best right now and further work is needed, the results in this thesis suggest that phosphorylation of BRD4 is common mechanism of resistance to BETis in LAC. Furthermore, this work provides evidence for an effective therapeutic combinational approach with CX-4945 and JQ1 in treating a larger range of LAC lines and lines that have acquired resistance to BETi treatment.        81		Chapter 5: Conclusions 5.1 Summary of research Our understanding of the functions of epigenetic proteins, specifically the BET family protein BRD4, continues to evolve as the understanding of cellular mechanisms and epigenetic regulation continues to grow. Recently, phosphorylation sites were identified for BRD4, with phosphorylation of BRD4 causing a switch in protein interactions [28] . In addition, Shu et al. showed that BRD4 was able to be phosphorylated by CK2, and that increased levels of pBRD4 played a role in acquiring resistance to BETis in TNBC [52]. Here I establish BETi resistance in two JQ1 sensitive LAC lines and confirm resistance is not due to EMT or multi-drug resistant mechanisms. Furthermore, I identify pBRD4 as being elevated in both resistant lines and give evidence for CK2 as the protein kinase phosphorylating BRD4.  I also identified a synergistic effect of combination treatment with the CK2 inhibitor CX-4945 in combination with JQ1 in both resistant and sensitive lines, suggesting that treatment with a CK2 inhibitor can re-sensitize resistant cells to BETi treatment and that pBRD4 may have a role in JQ1 sensitive lines as well. In addition, I identify that this combinational treatment can also selectively target JQ1 primary resistant LAC lines, implicating this combinational treatment as a possible therapeutic strategy in the future for LC patients. Lastly, I identify two conversely regulated genes as being considerably upregulated in their respective resistant line and confirmed each resistant lines dependence on these genes. I therefore hypothesize that BRD4 may regulate similar genes between cell lines, possibly in proximity to super enhancer regions, but that regulation of the same gene may differentiate in the requirement for a co-regulator between different lines. I further hypothesize that 82		phosphorylation of BRD4 is important for interaction with certain co-regulators and that phosphorylation of BRD4 has an either basal or stress response function in LAC lines that are sensitive to JQ1 treatment. As the epigenetic landscape of a cell can be drastically different, thought to be major reason for dramatically different clinical outcomes for patients with similar stages of NSCLC [5], it is important to understand the different roles of epigenetic proteins in these different contexts and to try and identify common mechanisms between lines irrespective of their epigenetic landscape. Through this work I hope to further the understanding of how LAC lines respond to BETi treatment and identify specific targets and pathways essential for resistance, while also offering a potential strategy for combination-based therapies to circumvent BETi resistance in lung cancer.   5.2 Future directions  This thesis provides evidence for phosphorylation of pBRD4 by the protein kinase CK2 as being a common mechanism of resistance to BETi treatment in LAC lines. However further work is needed to validate some of these results and provide further evidence to possible roles of pBRD4. Currently, I am looking into the banding pattern of pBRD4 in western blot analysis and trying to identify what the 150kDa and 84kDa bands represent. To help determine if these two bands are a form of BRD4 I will be blotting BRD4 knockdown lysate (similar to Figure 3.4B &D) with the pBRD4 antibody. This will allow me to determine if these bands are non-specific bands bound by the antibody or are a form of BRD4 as it would be expected that if the bands do represent different forms of BRD4 their expression would decrease equally to the 200kDA band as observed 83		previously. It is also possible that the siRNAs could be specific for just the long isoform of the protein and not cause decreased expression of all the other bands. However, this is likely not the case as a mixture of 4 siRNAs is provides as the reagent, though this will be looked at in greater detail depending on the results. In addition, I also look to determine if the absence of the 200kDa pBRD4 band in certain blots is due to lysate preparation. As these blots show banding patterns for BRD4 at the same molecular weight the missing bands could be a result of degradation during the lysate sample preparation process. To test this, I will be lysing H23 resistant cells in 6cm plates (8 samples) separately followed by half the samples being put in the -80°C overnight, as is the normal protocol, and half the samples being immediately sonicated, prepped, and loaded into gels. I will additionally be sonicating the 4 samples differentially using 0, 5, 15(standard), and 30sec sonication times to determine if this plays a role in the eventual detection of pBRD4 in western blot analysis and sonicating the 4 samples for each condition. Lastly, I look to further identify pBRD4s role in BETi resistance through investigating other mechanisms of regulating the phosphorylation of BRD4 and measuring this impact of cell sensitivity to BETi treatment. For instance, PP2A has previously shown to be a phosphatase of pBRD4 with PTZ treatment activating PP2A activity [52]. Therefore, it would be hypothesized that PTZ treatment or acquired resistant lines would decrease pBRD4 levels and make lines more sensitive to BETi treatment, using dose response experiments to test this. Conversely, PP2A levels could also be decreased through RNAi methodology in JQ1 sensitive lines. This would by predicted to increase pBRD4 levels in these lines and therefore make the lines more resistant to BETi treatment.  84		Further work may also be required to validate certain findings if this work is to be published. For instance, further validation is needed to confirm that acquired resistant lines are not resistant to BETi treatment due to elevated levels of the cellular transporter ABCA3 through RNAi experiments. One way to test this is to use siRNAs to knock down ABCA3 protein expression followed by evaluation of BETi sensitivity using JQ1 does response experiments. This RNAi approach could also be used to validate AXL and SPOCK1 as required genes for their resistant lines survival, with resistant cell viability being measured after protein knockdown. Lastly, it may be a requirement to validate that decreased resistant cell numbers at high concentrations in long-term growth assays (Figure 3.1B) and dose response experiments (Figure 3.1C & E) are not due to increased apoptosis in resistant lines. This can be tested by treating the H1975 resistant and control lines with four concentrations of JQ1 for 48hrs and assessment of cleaved-PARP levels by western blot analysis, as done previously on H23 resistant and control lines (as shown in Figure 3.1G). It may also important to determine if resistant cells are growing normally or are getting held up during certain stages of the cell cycle upon JQ1 treatment. This can be done using FACS (fluorescence associated cell sorting) flow cytometry followed by analysis using FlowJo after growing resistant cells in DMSO or 10µM JQ1 as previously described [37].  Lastly, the results found in this thesis provide support for recent findings such as CK2 phosphorylation of BRD4 [28, 52]  and the hypothesis that pBRD4 may regulate other genes separate from BRD4 due to binding to different co-regulators [28]. The latter of these however, is still in need of identification and validation. Similarly, it would be of interest to determine if pBRD4 and BRD4 localize to different regions of the genome and 85		therefore regulate different genes within a given cell line. To test these two things variations of the immunoprecipitation (IP) technique, namely protein complex immunoprecipitation (Co-IP) and chromatin immunoprecipitation (ChIP) could be used. Co-IP is a technique used to analyze protein-protein interactions. This could be performed in a specific cell line in parallel with an antibody specific for BRD4 and another specific for pBRD4. Protein binders could then be compared to determine if phosphorylation of BRD4 does indeed cause a switch in interacting partners as previously described [28]. Another way to test if pBRD4 is binding different proteins would be to treat a line known to have high levels of pBRD4 (such as the two LAC acquired resistant lines in this thesis) with or without an inhibitor to decrease pBRD4 levels (such as CX-4945 or PTZ) [52]. Protein binding profiles could then be compared between the two conditions with binders found only in the untreated condition being inferred to bind to pBRD4. However, this experimental design could result in inconclusive results due to leakiness of the experiments and would be dependent on if CK2 inhibitors or PTZ treatment inhibits all pBRD4 forms, which is still in need of validation.  Another important experiment would be to determine the localization of pBRD4 and BRD4 through ChIP analysis. Similar to above the idea of these experiments would be to compare binding locations on the genome to try and identify if BRD4 and pBRD4 have different roles in the cell and may further allude to pBRD4s function depending on where it binds. Overall, this thesis adds support to some previous claims for pBRD4 and identifies pBRD4 as playing a role in BETi resistance in LAC lines independent of re-expression genes down-regulated through initial BETi treatment.   86		Chapter 6: Extended Figures and Tables  											 							 Extended Figure 6.1: JQ1 resistant lines are permanently changed and are confirmed to be isogenic offspring of parental lines. Established H23 and H1975 resistant lines were cultured for 2 ½ months in regular media (without JQ1 presence) and imaged using a phase-contrast microscope (scale bar = 400µM) (A). 10-day growth assays were used to assess resistance of these lines (n=2) (B). Separately, JQ1 resistant lines and controls were identified to be isogenic colonies of the parental lines through authentication by STR profiling and calculation of “percent match,” in which >80% is consistent with the two lines being related (D). Western blot analysis also confirmed the H1975 resistant line to express the EGFR mutant driver gene (C). Raw STR probe data for H23 (E) and H1975 (F) are given. Percent Match Cell lines In relation to ATCC line In relation to “in lab” line H23 100% N/A H23 Control 100% 100% H23 Resistant 100% 100% H1975 100% N/a H1975 Control 100% 100% H1975 Resistant 100% 91.3%  H1975 Cell lines STR probes NCI-H1975 (from ATCC) H1975  (in lab) H1975 Control H1975 Resistant D5S818 11,12 11,12 11,12 11,12 D13S317 10,13 10 10 10,13 D7S820 8,11 8,11 8,11 8 D16S539 9,12 9,12 9,12 9,12 vWA 18 18 18 18 TH01 7 7 7 7 AMEL X X X X TPOX 8,11 8,11 8,11 8,11 CSF1PO 12 12 12 12 D3S1358   15 15 14,15 D21S11   28 28 28 D8S1179   13,16 13,16 13 Penta E   12,16 12,16 12 Penta D   12,13 12,13 12,13 D18S51   13 13 13 FGA   21,24 21,24 21,24  H23 Cell lines STR probes NCI-H23 (from ATCC) H23  (in lab) H23 Control H23 Resistant D5S818 12,13 12,13 12,13 12,13 D13S317 12 12 12 12 D7S820 9,10 9,10 9,10 9,10 D16S539 11 11 11 11 vWA 16,17 16,17 16,17 16,17 TH01 6 6 6 6 AMEL X X X X TPOX 8,9 8,9 8,9 8,9 CSF1PO 10 10 10 10 D3S1358  15 15 15 D21S11  30 30 30 D8S1179  15 15 15 Penta E  7,17 7,17 7,17 Penta D  8 8 8 D18S51  14 14 14 FGA  24 24 24 400um	400um	H23	Resistant	in	regular	media	H1975	Resistant	in	regular	media	10µM	JQ1										DMSO	H23	Resistant	in	regular	media	 H23	Control	10µM	JQ1										DMSO	H1975	Resistant	in	regular	media	 H1975	Control	A	 B	 C	D	E	 F	87		   							 	 Extended Figure 6.2: TGFβ1 treatment does not induce increased resistance to JQ1 treatment in H23 line. H23 parental lines were treated with 10ng/ml of TGFβ1 for approximately 2 weeks and morphology observed through phase-contrast microscopy (scale bars = 400µm) (A). Induction of EMT was assessed through western blot analysis through blotting for EMT markers (B). Resistance to JQ1 was determined by 72hrs dose response experiments (C) with IC50 values (D) representing average of two experiments ± SEM.            0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0H 2 3J Q 1  (u M )Relative Fluorescence to controlH 23H 2 3  T G F B 1H 2 3  R e s is ta n tCell line IC50 (µM) H23 0.8 ± 0.4 H23 TGFB1 0.9 ± 0.2 H23 Resistant >10 H23	TGFB1	treated	A	 B	C	 D	H23		88		H1975	Control									-			10	100	500	JQ1	(nM)	CX4945	(nM)							-								10						100						500			1000		5000				 H23	Control				H1975	Resistant		Growth	inhibition	(%)				 Combination	Index	Scores				CX4945	(nM)							-								10						100				500			1000		5000				CX4945	(nM)				10							100						500				1000			5000				-								10						100						500			1000			5000				 -									10					100						500			1000			5000				-								10						100				500			1000		5000				 -								10						100				500			1000		5000				10							100						500				1000			5000				 10							100						500				1000				5000									-			10	100	500									10	100	500	JQ1	(nM)	JQ1	(nM)			 																																																										 	 	 	 								             Extended Figure 6.3: Both resistant and control lines show synergistic values upon JQ1 and CX-4945 combination treatments. Combination growth assays for H975 resistant line (A), H1975 control line (B), and H23 control line (C) treated with a combination of JQ1 and CX-4945 for 72hrs. Plate images are one of two replicates. Percent growth inhibition was calculated and plotted in plated format from the mean of the two replicates as compared to the control well. CompuSyn was used to calculate combination index scores as a way of calculating synergy for the two doses, also shown in plate layout format. Legends for percent growth inhibition and combination index values shown at the bottom.     A	 B	E	C	E	89		H1975	Resistant		 H1975	Control		 H23	Control		CX4945 (nM)          -              500         1000      JQ1	(nM)	      -   100   500 2500 -             500         1000      -             500         1000      JQ1	(nM)	JQ1	(nM)	      -   100   500 2500          100   500 2500 Not	Available	CX4945 (nM)       -            500      1000      -         500      1000      -         500      1000      CX4945 (nM)       500          1000      500          1000      Growth	inhibition	(%)				 Combination	Index	Scores				A	 B	 C	 	 																														Extended Figure 6.4: 10-day combination growth assays show synergy for H23 control and H1975 resistant and control lines. 10-day combination growth assays for H1975 resistant (A), H1975 control (B), and H23 control (C). Plate image is one of three replicates (top). Percent growth inhibition was calculated as the mean of replicates as compared to the control well (middle). CI values, as calculated using CompuSyn, show only synergistic values for both lines (bottom). Legends for percent growth inhibition and combination index values shown at the bottom.  		90		A	 B	aaLine       IC50 (µM)   H2030        >10 aaLine       IC50 (µM)   H2030        >20 									 Extended Figure 6.5: H2030 confirmed to be resistant to BETis. Primary resistant line (H2030) validated to be resistant to BETis through 72hrs does response curves with JQ1 (A) and I-BET762 (B). Graphs are representative of one experiment, 4 replicate values for each dose, while accompanying IC50 values (D) are averaged from two experiments.                 0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0J Q 1J Q 1  (u M )Relative Fluorescence to controlH20300 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0I-B E TI-B E T  7 6 2  (u M )Relative Fluorescence to controlH203091		A	 B	C	 D											 					  Extended Figure 6.6: Dose response experiments with Trametinib, Afatinib, Crizotinib and BEZ-235. 72hr does response experiments of H23 and H1975 control and resistant lines, with inhibitors targeting different pathways including; MEK pathway (Trametinib (A)), HER2/ EGFR (Afatinib (B)), AXL (Crizotinib (C)), and PI3K pathway (BEZ-235 (D)). Graphs are representative of one experiment, 4 replicate values for each dose, while accompanying IC50 values (D) are averaged from two experiments ± SEM.   0 .0 0 0 0 1 0 .0 0 0 1 0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0T r a m e t in ibT ra m e tin ib  (u M )Relative Viability to Control H 2 3  C o n tro lH 2 3  R e s is ta n tH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n t0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0A fa tin ibA fa tin ib  (u M )Relative Viability to ControlH 2 3  C o n tro lH 2 3  R e s is ta n tH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n t0 .0 0 1 0 .0 1 0 .1 1 1 00 .00 .51 .0C r iz o t in ibC r iz o t in ib  (u M )Relative Viability to ControlH 2 3  C o n tro lH 2 3  R e s is ta n tH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n t0 .0 0 1 0 .0 1 0 .1 10 .00 .51 .0B E Z -2 3 5B E Z -2 3 5  (u M )Relative Viability to ControlH 2 3  C o n tro lH 2 3  R e s is ta n tH 1 9 7 5  C o n tro lH 1 9 7 5  R e s is ta n tCell line IC50 (µM) H23 Control 0.0005 ± 0.0001 H23 Resistant 0.00011 ± 0.00004 H1975 Control 2.4 ± 0.7 H1975 Resistant 0.0009 ± 0.0006 Cell line IC50 (µM) H23 Control 2.3 ± 0.4 H23 Resistant 0.083 ± 0.008 H1975 Control 0.011 ± 0.003 H1975 Resistant 0.009 ± 0.005 Cell line IC50 (µM) H23 Control 0.07 ± 0.02 H23 Resistant 0.019 ± 0.0006 H1975 Control 0.051 ± 0.008 H1975 Resistant 0.014 ± 0.009 Cell line IC50 (µM) H23 Control 0.34 ± 0.05 H23 Resistant 0.42 ± 0.04 H1975 Control 2.3 ± 0.7 H1975 Resistant 0.35 ± 0.05 92		B	A	A	                                   Extended Figure 6.7: Heat-maps of the top 20 deregulated genes in only the resistant lines. Top 20 up and down-regulated gene found only to be deregulated in the H23 (A) and H1975 (B) resistant line and not in the 6hr time point gene list as well. Genes are at least ≥2 fold differentially regulated as compared to the control lines with a BH corrected p-value<0.005 (unpaired t-test). No common genes were identified between the two lists. 93		Analogously	regulated	genes	of	the	101	common	genes				 			Extended Table 6.1: Gene set list of analogously regulated genes. 70 genes of the 101 common genes are up-regulated in both resistant lines or down-regulated in both. Numbers are log2 fold changes as compared to the control for both the resistant line (cultured in 10µM JQ1) or control lines treated for 6hrs with 10µM JQ1 (6hr). 				    H1975 H23 Gene Name Resistant 6hr Resistant 6hr ABCA3 2.555732 0.221681 2.496724 0.216226 ALDH4A1 1.440575 0.468657 1.16937 0.263361 BARX2 -2.47518 -1.4896 -2.16937 -1.06944 BCKDHB 1.756708 0.008942 1.260362 -0.00388 C8orf44-SGK3 -2.27148 -0.48627 -1.77862 -0.64386 C8orf44-SGK3 -2.11629 -0.3604 -1.60107 -0.7859 CAMTA1 2.084665 0.121777 1.699531 -0.74372 CAMTA1 1.834841 0.248745 1.615375 -0.71387 CHMP4A 1.349949 0.004452 1.302181 -0.02962 CLYBL 2.483807 0.460682 1.529669 0.18492 CNPY2 -2.07116 -0.53502 -1.06566 0.154777 COQ3 -1.12688 -0.74172 -1.09489 -1.39857 CRIM1 2.030422 0.396363 1.67237 0.700717 CRIM1 1.916598 0.311757 1.881345 0.85055 DERA -1.52231 0.620701 -1.24616 -0.17753 DMRTA1 -1.6039 -1.39509 -2.80232 -1.82469 EPCAM -6.06287 0.192027 -2.23219 -0.74965 EPCAM -6.04025 0.328493 -2.22334 -0.88001 ERCC6 1.741979 0.267107 1.602496 0.379245 FAM149B1 1.156355 0.234361 1.211381 0.241801 FAM69A -2.59211 -0.03148 -1.49203 0.341593 FOSL1 -1.62245 -0.46224 -3.13009 -1.77356 FOXC1 1.5269 0.102271 1.943103 0.93323 GGPS1 2.01933 0.224975 1.080855 -0.11568 GINS2 -2.00967 -0.2156 -1.18933 -0.55999 GPR56 -3.56341 -0.29043 -2.24957 -1.0966 HEG1 1.010154 -0.53147 1.52306 -0.25137 HMGB3 -3.53917 0.04423 -2.28251 -0.45806 HMGB3 -2.83871 0.13126 -1.83426 0.041955 HOOK3 1.087496 0.003515 1.307883 -0.17988 IL17RD 1.348346 -0.0701 1.135875 0.002478 IMPA1 -1.28164 0.118827 -1.46056 -0.07357 IMPA1 -1.18394 0.192273 -1.45387 -0.14028 KIAA1147 -1.108 -0.86634 -1.00745 -0.62962 LACTB 2.389499 0.26709 1.133193 0.46698 LOC100128551 1.098932 0.216393 1.187601 0.521265 LONRF3 -1.30944 -0.09965 -1.40248 -0.39587 LSM6 -1.31812 -0.10855 -1.04825 -0.2749 MAP7 -5.11709 -0.02461 -5.55791 -0.46352 MAP7 -3.84603 0.104927 -5.66471 -0.37815 MARCH2 1.727864 0.183563 1.415208 -0.02684 MARCH2 1.835602 0.227106 1.258737 0.024798 MEF2C -2.92771 -1.28993 -1.83506 0.03093 MET -1.45168 -0.4087 -4.32819 -1.20203 MGST2 -4.40705 -0.40595 -3.19687 -0.33868 MTMR2 -1.19705 -0.37377 -1.68986 -0.93199 MUC1 -1.22042 0.187942 -1.23257 -0.22079 NIN 1.703647 0.057069 1.234349 0.150957 NT5E -1.32368 0.183937 -3.37009 0.027477 PALLD 1.583999 0.401239 1.892756 -1.02391 PIGK -1.26411 0.419091 -1.89403 -0.32352 PNRC1 1.917132 0.605774 2.076833 0.935421 PON2 -2.78451 -0.18493 -1.98698 -0.31514 PON2 -2.81548 -0.15059 -2.07908 -0.32677 PPP1R1C -2.07862 -0.48723 -3.16907 -2.08747 PQLC3 2.360558 0.630731 2.234225 0.123724 PRDM8 -1.70215 -1.85858 -2.37352 -1.65991 PRMT2 1.433867 0.315339 1.071919 0.203737 PROS1 -1.89483 -0.06157 -1.13557 0.074391 PRR13 -1.42276 -0.19996 -1.1218 -0.34293 PRTFDC1 -3.2271 -0.07307 -1.59865 -0.41353 RAB31 -1.16437 -0.00978 -1.42338 -0.85861 RIBC2 -2.39804 -1.46155 -1.43201 -1.32066 SLC44A3 -1.9592 -0.11519 -2.1466 -0.72218 SLITRK6 -3.54604 -1.91657 -4.60112 -3.19546 SSH2 -2.32757 0.14112 -2.68837 -0.49181 STEAP1 -4.76615 -0.81298 -1.75932 -0.39429 STEAP1 -4.79812 -0.82482 -1.58757 -0.40972 STEAP2 -2.3334 -1.1588 -2.45906 -1.36447 STX11 1.704335 0.155893 1.364074 -0.34865 STX11 1.608255 0.331138 1.411876 -0.41299 SYT1 -1.12439 -1.02337 -2.39028 -0.42883 TCAIM 2.465037 0.367294 1.032526 0.054028 TMEM117 -1.20631 0.383966 -1.20891 -0.65747 TMEM194A -1.18429 -0.73822 -1.17637 -0.86464 TMEM50B 1.288956 0.961294 1.496549 0.852938 TNC -4.20213 0.114524 -2.89019 -0.29849 TPM3 -1.92166 0.228428 -1.32073 0.057252 TROAP -2.50938 -0.01139 -1.83049 -0.36724 TSPAN6 -1.65044 -0.03878 -1.09219 -0.42335 ZFAND1 -1.29135 -0.24536 -1.81035 -0.08577 94		Conversely	regulated	genes	of	the	101	common	genes		          																		 Extended Table 6.2 Gene set list of conversely regulated genes. 31 genes of the 101 common genes are up-regulated in one line and down in the other or vice versa. Numbers are log2 fold changes as compared to the control for both the resistant line (cultured in 10µM JQ1) or control lines treated for 6hrs with 10µM JQ1 (6hr).            H1975 H23 Gene Name Resistant 6hr Resistant 6hr AXL 5.639425 -0.45468 -1.64657 -0.48569 CASK -1.06844 0.417729 1.126168 0.427475 CD44 1.154102 0.505192 -2.09404 -0.28129 CD44 1.618407 0.252497 -2.88022 0.026587 CD44 1.390007 0.416735 -2.47518 -0.10625 CPS1 2.487073 0.001245 -1.71178 -1.06708 DENND5B 1.254705 -1.05938 -1.57921 -0.74163 DUSP4 2.117123 -0.15871 -2.72144 -0.31415 DVL3 -1.04031 0.245613 2.004215 0.057664 FAT4 2.330595 -0.03346 -2.59202 -0.4345 FHL1 2.581158 0.273041 -4.6629 -0.28627 FHL1 2.174903 0.065608 -4.11703 -0.25497 FHL1 2.056448 -0.00835 -3.59634 -0.33059 FZD7 -2.30265 0.220794 1.724781 0.429294 GATM -2.45216 0.291936 1.087296 0.860456 IRX2 -1.77218 -0.60536 1.364379 0.186967 JAG1 1.248887 -0.08224 -3.80265 -0.54641 KCMF1 1.885547 -1.38519 -1.12502 -1.44972 KLHL4 1.138949 0.131306 -2.29498 -0.97928 MAMLD1 1.921724 -0.10265 -1.63041 0.599452 MARCH6 1.047853 -0.55946 -1.10184 0.171004 NEBL -2.97857 0.075023 1.707967 -0.29724 NR2F2 1.925674 0.667016 -1.22616 -0.22665 PCDH7 -1.65635 0.036624 1.038787 -0.11994 PEX13 -1.19276 0.062871 1.192561 1.019168 PROCR 2.284053 0.059039 -1.22364 -0.28214 PTP4A2 1.753385 0.409895 -1.00646 -0.25939 S100A4 2.829169 -0.09439 -1.76454 0.063871 SNN -1.32877 0.069991 1.217833 -0.42576 SPOCK1 -3.83813 0.147237 5.830128 0.601139 SRI 1.412671 -0.40294 -1.69364 -0.42177 TAF9B 1.256623 -0.06258 -1.35146 -0.58709 TMEM56 1.357481 -0.28376 -1.01511 -0.90967 VGF -1.14959 1.174524 3.263561 0.614033 VGF -1.08547 1.343336 3.376363 0.600096 VTI1B 1.222209 -0.00773 -1.0653 0.008323 95		References 1.	 Cheng,	T.Y.,	et	al.,	The	International	Epidemiology	of	Lung	Cancer:	Latest	Trends,	Disparities,	and	Tumor	Characteristics.	J	Thorac	Oncol,	2016.	11(10):	p.	1653-71.	2.	 Siegel,	R.L.,	K.D.	Miller,	and	A.	Jemal,	Cancer	Statistics,	2017.	CA	Cancer	J	Clin,	2017.	67(1):	p.	7-30.	3.	 Miller,	K.D.,	et	al.,	Cancer	treatment	and	survivorship	statistics,	2016.	CA	Cancer	J	Clin,	2016.	66(4):	p.	271-89.	4.	 Sun,	S.,	J.H.	Schiller,	and	A.F.	Gazdar,	Lung	cancer	in	never	smokers--a	different	disease.	Nat	Rev	Cancer,	2007.	7(10):	p.	778-90.	5.	 Herbst,	R.S.,	J.V.	Heymach,	and	S.M.	Lippman,	Lung	cancer.	N	Engl	J	Med,	2008.	359(13):	p.	1367-80.	6.	 Meza,	R.,	et	al.,	Lung	cancer	incidence	trends	by	gender,	race	and	histology	in	the	United	States,	1973-2010.	PLoS	One,	2015.	10(3):	p.	e0121323.	7.	 Aberle,	D.R.,	et	al.,	Reduced	lung-cancer	mortality	with	low-dose	computed	tomographic	screening.	N	Engl	J	Med,	2011.	365(5):	p.	395-409.	8.	 Stratton,	M.R.,	P.J.	Campbell,	and	P.A.	Futreal,	The	cancer	genome.	Nature,	2009.	458(7239):	p.	719-24.	9.	 Weinstein,	I.B.	and	A.K.	Joe,	Mechanisms	of	disease:	Oncogene	addiction--a	rationale	for	molecular	targeting	in	cancer	therapy.	Nat	Clin	Pract	Oncol,	2006.	3(8):	p.	448-57.	10.	 Mancini,	M.	and	Y.	Yarden,	Mutational	and	network	level	mechanisms	underlying	resistance	to	anti-cancer	kinase	inhibitors.	Semin	Cell	Dev	Biol,	2016.	50:	p.	164-76.	11.	 Waddington,	C.H.,	Preliminary	Notes	on	the	Development	of	the	Wings	in	Normal	and	Mutant	Strains	of	Drosophila.	Proc	Natl	Acad	Sci	U	S	A,	1939.	25(7):	p.	299-307.	12.	 Dawson,	M.A.,	The	cancer	epigenome:	Concepts,	challenges,	and	therapeutic	opportunities.	Science,	2017.	355(6330):	p.	1147-1152.	13.	 Arrowsmith,	C.H.,	et	al.,	Epigenetic	protein	families:	a	new	frontier	for	drug	discovery.	Nat	Rev	Drug	Discov,	2012.	11(5):	p.	384-400.	14.	 Campbell,	R.M.	and	P.J.	Tummino,	Cancer	epigenetics	drug	discovery	and	development:	the	challenge	of	hitting	the	mark.	J	Clin	Invest,	2014.	124(1):	p.	64-9.	15.	 Baylin,	S.B.,	DNA	methylation	and	gene	silencing	in	cancer.	Nat	Clin	Pract	Oncol,	2005.	2	Suppl	1:	p.	S4-11.	16.	 Gros,	C.,	et	al.,	DNA	methylation	inhibitors	in	cancer:	recent	and	future	approaches.	Biochimie,	2012.	94(11):	p.	2280-96.	17.	 Horn,	P.J.	and	C.L.	Peterson,	Molecular	biology.	Chromatin	higher	order	folding--wrapping	up	transcription.	Science,	2002.	297(5588):	p.	1824-7.	18.	 Hansen,	J.C.,	Conformational	dynamics	of	the	chromatin	fiber	in	solution:	determinants,	mechanisms,	and	functions.	Annu	Rev	Biophys	Biomol	Struct,	2002.	31:	p.	361-92.	19.	 Cole,	P.A.,	Chemical	probes	for	histone-modifying	enzymes.	Nat	Chem	Biol,	2008.	4(10):	p.	590-7.	96		20.	 Filippakopoulos,	P.	and	S.	Knapp,	The	bromodomain	interaction	module.	FEBS	Lett,	2012.	586(17):	p.	2692-704.	21.	 Chaidos,	A.,	V.	Caputo,	and	A.	Karadimitris,	Inhibition	of	bromodomain	and	extra-terminal	proteins	(BET)	as	a	potential	therapeutic	approach	in	haematological	malignancies:	emerging	preclinical	and	clinical	evidence.	Ther	Adv	Hematol,	2015.	6(3):	p.	128-41.	22.	 Filippakopoulos,	P.	and	S.	Knapp,	Targeting	bromodomains:	epigenetic	readers	of	lysine	acetylation.	Nat	Rev	Drug	Discov,	2014.	13(5):	p.	337-56.	23.	 Shu,	S.	and	K.	Polyak,	BET	Bromodomain	Proteins	as	Cancer	Therapeutic	Targets.	Cold	Spring	Harb	Symp	Quant	Biol,	2017.	24.	 Taniguchi,	Y.,	The	Bromodomain	and	Extra-Terminal	Domain	(BET)	Family:	Functional	Anatomy	of	BET	Paralogous	Proteins.	Int	J	Mol	Sci,	2016.	17(11).	25.	 Filippakopoulos,	P.,	et	al.,	Selective	inhibition	of	BET	bromodomains.	Nature,	2010.	468(7327):	p.	1067-73.	26.	 Chiang,	C.M.,	Phospho-BRD4:	transcription	plasticity	and	drug	targeting.	Drug	Discov	Today	Technol,	2016.	19:	p.	17-22.	27.	 Wu,	S.Y.	and	C.M.	Chiang,	The	double	bromodomain-containing	chromatin	adaptor	Brd4	and	transcriptional	regulation.	J	Biol	Chem,	2007.	282(18):	p.	13141-5.	28.	 Wu,	S.Y.,	et	al.,	Phospho	switch	triggers	Brd4	chromatin	binding	and	activator	recruitment	for	gene-specific	targeting.	Mol	Cell,	2013.	49(5):	p.	843-57.	29.	 Alsarraj,	J.,	et	al.,	BRD4	short	isoform	interacts	with	RRP1B,	SIPA1	and	components	of	the	LINC	complex	at	the	inner	face	of	the	nuclear	membrane.	PLoS	One,	2013.	8(11):	p.	e80746.	30.	 Devaiah,	B.N.,	A.	Gegonne,	and	D.S.	Singer,	Bromodomain	4:	a	cellular	Swiss	army	knife.	J	Leukoc	Biol,	2016.	100(4):	p.	679-686.	31.	 Shi,	J.	and	C.R.	Vakoc,	The	mechanisms	behind	the	therapeutic	activity	of	BET	bromodomain	inhibition.	Mol	Cell,	2014.	54(5):	p.	728-36.	32.	 Rathert,	P.,	et	al.,	Transcriptional	plasticity	promotes	primary	and	acquired	resistance	to	BET	inhibition.	Nature,	2015.	525(7570):	p.	543-7.	33.	 Kumar,	K.,	et	al.,	GLI2-dependent	c-MYC	upregulation	mediates	resistance	of	pancreatic	cancer	cells	to	the	BET	bromodomain	inhibitor	JQ1.	Sci	Rep,	2015.	5:	p.	9489.	34.	 Mazur,	P.K.,	et	al.,	Combined	inhibition	of	BET	family	proteins	and	histone	deacetylases	as	a	potential	epigenetics-based	therapy	for	pancreatic	ductal	adenocarcinoma.	Nat	Med,	2015.	21(10):	p.	1163-71.	35.	 Puissant,	A.,	et	al.,	Targeting	MYCN	in	neuroblastoma	by	BET	bromodomain	inhibition.	Cancer	Discov,	2013.	3(3):	p.	308-23.	36.	 Delmore,	J.E.,	et	al.,	BET	bromodomain	inhibition	as	a	therapeutic	strategy	to	target	c-Myc.	Cell,	2011.	146(6):	p.	904-17.	37.	 Lockwood,	W.W.,	et	al.,	Sensitivity	of	human	lung	adenocarcinoma	cell	lines	to	targeted	inhibition	of	BET	epigenetic	signaling	proteins.	Proc	Natl	Acad	Sci	U	S	A,	2012.	109(47):	p.	19408-13.	38.	 Baker,	E.K.,	et	al.,	BET	inhibitors	induce	apoptosis	through	a	MYC	independent	mechanism	and	synergise	with	CDK	inhibitors	to	kill	osteosarcoma	cells.	Sci	Rep,	2015.	5:	p.	10120.	97		39.	 French,	C.A.,	et	al.,	BRD4	bromodomain	gene	rearrangement	in	aggressive	carcinoma	with	translocation	t(15;19).	Am	J	Pathol,	2001.	159(6):	p.	1987-92.	40.	 LeRoy,	G.,	B.	Rickards,	and	S.J.	Flint,	The	double	bromodomain	proteins	Brd2	and	Brd3	couple	histone	acetylation	to	transcription.	Mol	Cell,	2008.	30(1):	p.	51-60.	41.	 Deeney,	J.T.,	et	al.,	BET	Bromodomain	Proteins	Brd2,	Brd3	and	Brd4	Selectively	Regulate	Metabolic	Pathways	in	the	Pancreatic	β-Cell.	PLoS	One,	2016.	11(3):	p.	e0151329.	42.	 Fowler,	T.,	et	al.,	Regulation	of	MYC	expression	and	differential	JQ1	sensitivity	in	cancer	cells.	PLoS	One,	2014.	9(1):	p.	e87003.	43.	 Shi,	X.,	et	al.,	Loss	of	TRIM33	causes	resistance	to	BET	bromodomain	inhibitors	through	MYC-	and	TGF-β-dependent	mechanisms.	Proc	Natl	Acad	Sci	U	S	A,	2016.	113(31):	p.	E4558-66.	44.	 Chiang,	C.M.,	Brd4	engagement	from	chromatin	targeting	to	transcriptional	regulation:	selective	contact	with	acetylated	histone	H3	and	H4.	F1000	Biol	Rep,	2009.	1:	p.	98.	45.	 Wang,	R.,	et	al.,	Bromodomain	protein	Brd4	associated	with	acetylated	chromatin	is	important	for	maintenance	of	higher-order	chromatin	structure.	J	Biol	Chem,	2012.	287(14):	p.	10738-52.	46.	 Wadhwa,	E.	and	T.	Nicolaides,	Bromodomain	Inhibitor	Review:	Bromodomain	and	Extra-terminal	Family	Protein	Inhibitors	as	a	Potential	New	Therapy	in	Central	Nervous	System	Tumors.	Cureus,	2016.	8(5):	p.	e620.	47.	 Belkina,	A.C.	and	G.V.	Denis,	BET	domain	co-regulators	in	obesity,	inflammation	and	cancer.	Nat	Rev	Cancer,	2012.	12(7):	p.	465-77.	48.	 Bhadury,	J.,	et	al.,	BET	and	HDAC	inhibitors	induce	similar	genes	and	biological	effects	and	synergize	to	kill	in	Myc-induced	murine	lymphoma.	Proc	Natl	Acad	Sci	U	S	A,	2014.	111(26):	p.	E2721-30.	49.	 Ott,	C.J.,	et	al.,	BET	bromodomain	inhibition	targets	both	c-Myc	and	IL7R	in	high-risk	acute	lymphoblastic	leukemia.	Blood,	2012.	120(14):	p.	2843-52.	50.	 Dawson,	M.A.,	et	al.,	Inhibition	of	BET	recruitment	to	chromatin	as	an	effective	treatment	for	MLL-fusion	leukaemia.	Nature,	2011.	478(7370):	p.	529-33.	51.	 Ma,	Y.,	et	al.,	The	MAPK	Pathway	Regulates	Intrinsic	Resistance	to	BET	Inhibitors	in	Colorectal	Cancer.	Clin	Cancer	Res,	2017.	23(8):	p.	2027-2037.	52.	 Shu,	S.,	et	al.,	Response	and	resistance	to	BET	bromodomain	inhibitors	in	triple-negative	breast	cancer.	Nature,	2016.	529(7586):	p.	413-7.	53.	 Fong,	C.Y.,	et	al.,	BET	inhibitor	resistance	emerges	from	leukaemia	stem	cells.	Nature,	2015.	525(7570):	p.	538-42.	54.	 Settleman,	J.,	Cancer:	Bet	on	drug	resistance.	Nature,	2016.	529(7586):	p.	289-90.	55.	 Chou,	T.C.,	Drug	combination	studies	and	their	synergy	quantification	using	the	Chou-Talalay	method.	Cancer	Res,	2010.	70(2):	p.	440-6.	56.	 Irizarry,	R.A.,	et	al.,	Exploration,	normalization,	and	summaries	of	high	density	oligonucleotide	array	probe	level	data.	Biostatistics,	2003.	4(2):	p.	249-64.	57.	 Lockwood,	W.W.,	et	al.,	Integrative	genomic	analyses	identify	BRF2	as	a	novel	lineage-specific	oncogene	in	lung	squamous	cell	carcinoma.	PLoS	Med,	2010.	7(7):	p.	e1000315.	98		58.	 Reid,	Y.,	et	al.,	Authentication	of	human	cell	lines	by	STR	DNA	profiling	analysis.	2013,	NCBI:	Bethesda	(MD).	59.	 Soldani,	C.	and	A.I.	Scovassi,	Poly(ADP-ribose)	polymerase-1	cleavage	during	apoptosis:	an	update.	Apoptosis,	2002.	7(4):	p.	321-8.	60.	 Thomson,	S.,	et	al.,	Epithelial	to	mesenchymal	transition	is	a	determinant	of	sensitivity	of	non-small-cell	lung	carcinoma	cell	lines	and	xenografts	to	epidermal	growth	factor	receptor	inhibition.	Cancer	Res,	2005.	65(20):	p.	9455-62.	61.	 Yang,	A.D.,	et	al.,	Chronic	oxaliplatin	resistance	induces	epithelial-to-mesenchymal	transition	in	colorectal	cancer	cell	lines.	Clin	Cancer	Res,	2006.	12(14	Pt	1):	p.	4147-53.	62.	 Lian,	H.,	et	al.,	CK2	inhibitor	CX-4945	destabilizes	NOTCH1	and	synergizes	with	JQ1	against	human	T-acute	lymphoblastic	leukemic	cells.	Haematologica,	2017.	102(1):	p.	e17-e21.	63.	 Xu,	J.,	S.	Lamouille,	and	R.	Derynck,	TGF-beta-induced	epithelial	to	mesenchymal	transition.	Cell	Res,	2009.	19(2):	p.	156-72.	64.	 Wojcik,	E.J.,	et	al.,	Kinesin-5:	cross-bridging	mechanism	to	targeted	clinical	therapy.	Gene,	2013.	531(2):	p.	133-49.	65.	 Kurimchak,	A.M.,	et	al.,	Resistance	to	BET	Bromodomain	Inhibitors	Is	Mediated	by	Kinome	Reprogramming	in	Ovarian	Cancer.	Cell	Rep,	2016.	16(5):	p.	1273-86.	66.	 Ferrer-Font,	L.,	et	al.,	Targeting	Protein	Kinase	CK2:	Evaluating	CX-4945	Potential	for	GL261	Glioblastoma	Therapy	in	Immunocompetent	Mice.	Pharmaceuticals	(Basel),	2017.	10(1).	67.	 Okimoto,	R.A.	and	T.G.	Bivona,	AXL	receptor	tyrosine	kinase	as	a	therapeutic	target	in	NSCLC.	Lung	Cancer	(Auckl),	2015.	6:	p.	27-34.	68.	 Yang,	J.,	et	al.,	SPOCK1	promotes	the	proliferation,	migration	and	invasion	of	glioma	cells	through	PI3K/AKT	and	Wnt/β-catenin	signaling	pathways.	Oncol	Rep,	2016.	35(6):	p.	3566-76.	69.	 Shu,	Y.J.,	et	al.,	SPOCK1	as	a	potential	cancer	prognostic	marker	promotes	the	proliferation	and	metastasis	of	gallbladder	cancer	cells	by	activating	the	PI3K/AKT	pathway.	Mol	Cancer,	2015.	14:	p.	12.	70.	 Maurus,	K.,	et	al.,	The	AP-1	transcription	factor	FOSL1	causes	melanocyte	reprogramming	and	transformation.	Oncogene,	2017.	71.	 Marcotte,	R.,	et	al.,	Functional	Genomic	Landscape	of	Human	Breast	Cancer	Drivers,	Vulnerabilities,	and	Resistance.	Cell,	2016.	164(1-2):	p.	293-309.	72.	 Fiskus,	W.,	et	al.,	BET	protein	antagonist	JQ1	is	synergistically	lethal	with	FLT3	tyrosine	kinase	inhibitor	(TKI)	and	overcomes	resistance	to	FLT3-TKI	in	AML	cells	expressing	FLT-ITD.	Mol	Cancer	Ther,	2014.	13(10):	p.	2315-27.	73.	 Chapuy,	B.,	et	al.,	Intracellular	ABC	transporter	A3	confers	multidrug	resistance	in	leukemia	cells	by	lysosomal	drug	sequestration.	Leukemia,	2008.	22(8):	p.	1576-86.	74.	 Kim,	J.	and	S.	Hwan	Kim,	CK2	inhibitor	CX-4945	blocks	TGF-β1-induced	epithelial-to-mesenchymal	transition	in	A549	human	lung	adenocarcinoma	cells.	PLoS	One,	2013.	8(9):	p.	e74342.	99		75.	 Duncan,	J.S.	and	D.W.	Litchfield,	Too	much	of	a	good	thing:	the	role	of	protein	kinase	CK2	in	tumorigenesis	and	prospects	for	therapeutic	inhibition	of	CK2.	Biochim	Biophys	Acta,	2008.	1784(1):	p.	33-47.	76.	 Manchado,	E.,	et	al.,	A	combinatorial	strategy	for	treating	KRAS-mutant	lung	cancer.	Nature,	2016.	534(7609):	p.	647-51.	77.	 Zhang,	Z.,	et	al.,	Activation	of	the	AXL	kinase	causes	resistance	to	EGFR-targeted	therapy	in	lung	cancer.	Nat	Genet,	2012.	44(8):	p.	852-60.	78.	 Miao,	L.,	et	al.,	SPOCK1	is	a	novel	transforming	growth	factor-β	target	gene	that	regulates	lung	cancer	cell	epithelial-mesenchymal	transition.	Biochem	Biophys	Res	Commun,	2013.	440(4):	p.	792-7.	79.	 Song,	X.,	et	al.,	Up-regulation	of	SPOCK1	induces	epithelial-mesenchymal	transition	and	promotes	migration	and	invasion	in	esophageal	squamous	cell	carcinoma.	J	Mol	Histol,	2015.	46(4-5):	p.	347-56.	80.	 Gjerdrum,	C.,	et	al.,	Axl	is	an	essential	epithelial-to-mesenchymal	transition-induced	regulator	of	breast	cancer	metastasis	and	patient	survival.	Proc	Natl	Acad	Sci	U	S	A,	2010.	107(3):	p.	1124-9.	 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0365542/manifest

Comment

Related Items