Sex differences in mammalian gene expression result from differences in genotypic sex as well as in hormonal regulators between males and females. In rat, mouse and human liver, ~1000 genes are expressed in a sex-dependent manner, and contribute to sex differences in metabolism of drugs, steroids and lipids, and in liver and cardiovascular disease risk. In rats and mice, sex-biased liver gene expression is primarily dictated by the sexually dimorphic pattern of pituitary growth hormone (GH) release and its STAT5-dependent transcriptional activities. Studies presented in this thesis include the following. (1) A computational approach based on DNA sequence and phylogenetic conservation was developed and used to identify novel functional STAT5 binding sites - both consensus and non-consensus STAT5 sequences - near prototypic GH-responsive genes. (2) Global gene expression analysis of livers from pituitary-ablated male and female mice identified four major classes of sex-biased genes differing in their profiles of GH dependence. (3) Sex-differences in DNase-hypersensitive sites (DHS, corresponding to open chromatin regions) were identified genome-wide in mouse liver. These sex-differential DHSs were enriched for association with sex-biased genes, but a majority was distant from sex-biased genes. Furthermore, many were responsive to GH treatment, demonstrating that GH-mediated regulation involves chromatin remodeling. Analysis of sequence motifs enriched at sex-biased DHSs implicated STAT5 and novel transcription factors such as PBX1 and TAL1 in sex-biased gene regulation. (4) Genome-wide mapping of histone modifications revealed distinct mechanisms of sex-biased gene regulation in male and female liver: sex-dependent K27me3-mediated repression is an important mechanism of repression of female-biased, but not of male-biased, genes, and a sex-dependent K4me1 distribution, suggesting nucleosome repositioning by pioneer factors, is observed at male-biased, but not female-biased, regulatory sites. STAT5-mediated activation was most strongly associated with sex-biased chromatin modifications, while BCL6-mediated repression primarily occurs in association with sex-independent chromatin modifications, both at binding sites and at target genes. The relationships between sex-dependent chromatin accessibility, chromatin modifications and transcription-factor binding uncovered by these studies help elucidate the molecular mechanisms governing sex-differential gene expression, and underscore the utility of functional genomic and epigenetic studies as tools for elucidating transcriptional regulation in complex mammalian systems.

The development of Chromatin immunoprecipitation followed by sequencing (ChIP-seq) technology has enabled the construction of genome-wide maps of protein-DNA interaction. Such maps provide information about transcriptional regulation at the epigenetic level (histone modifications and histone variants) and at the level of transcription factor (TF) activity. This dissertation presents novel computational methods for ChIP-seq data analysis and applications. The work of this dissertation addresses four main challenges. First, I address the problem of detecting histone modifications from ChIP-seq cancer samples. The presence of copy number variations (CNVs) in cancer samples results in statistical biases that lead to inaccurate predictions when standard methods are used. To overcome this issue I developed HMCan, a specially designed algorithm to handle ChIP-seq cancer data by accounting for the presence of CNVs. When using ChIP-seq data from cancer cells, HMCan demonstrates unbiased and accurate predictions compared to the standard state of the art methods. Second, I address the problem of identifying changes in histone modifications between two ChIP-seq samples with different genetic backgrounds (for example cancer vs. normal). In addition to CNVs, different antibody efficiency between samples and presence of samples replicates are challenges for this problem. To overcome these issues, I developed the HMCan-diff algorithm as an extension to HMCan. HMCan-diff implements robust normalization methods to address the challenges listed above. HMCan-diff significantly outperforms another state of the art methods on data containing cancer samples. Third, I investigate and analyze predictions of different methods for enhancer prediction based on ChIP-seq data. The analysis shows that predictions generated by different methods are poorly overlapping. To overcome this issue, I developed DENdb, a database that integrates enhancer predictions from different methods. DENdb also integrates several experimental data including ChIP-seq data for TF binding sites. Finally, I present an extensive computational comparison of different ab-initio motif identification methods based on TF ChIP-seq data. The comparison included 10 different methods over 159 different TF datasets. Recommendations of this comparison indicate that the usage of simple methods outperforms the usage of high order models.

High­-throughput sequencing has become pervasive in all facets of genomic analysis. I developed computational methods to analyze high­-throughput sequencing data and derive biological conclusions in two research areas -- transcriptional regulation in mammals and evolution of virus under immune pressure. To investigate transcriptional regulation, I integrated data from multiple experiments performed by the ENCODE consortium. First, my analysis revealed that Transcription Factors (TFs) prefer to bind GC-­rich, histone­-depleted regions. By comparing in vivo and in vitro nucleosome dynamics, I observed that while histones have an innate preference for binding GC-­rich DNA, TF binding overrides this preference and produces a negative correlation between GC content and histone enrichment. In the next project, I found that the binding events of multiple TFs co-­occur at genomic regions enriched in activating histone marks that are typically associated with gene enhancers and promoters, suggesting that these regions may be enhancers or have TSS-­distal transcription. Lastly, I used supervised machine ­learning techniques to train histone enrichment signals and sequence features to predict transcriptional enhancers to be validated in mouse-­transgenic assays. In a post­-clinical trial exploratory analysis of Hepatitis C Virus (HCV), I traced the evolutionary path of the envelope proteins E1 and E2 in HCV-infected liver transplant patients, in response to a novel antibody. I developed a systematic amino acid­-level analysis pipeline that quantifies differences in amino acid frequencies in each position between two time points. Upon applying this method across all positions in the E1/E2 region and comparing pre-­liver­-transplant and post­-viral­-rebound time points, mutations in two positions emerged as being key to antibody evasion. Both these mutations--N415K/D and N417S--were in the epitope targeted by the antibody, but surprisingly, did not co­-occur. In post­-rebound viral genomes that contain the N417S mutation but retain the wild-­type variant at 415, N-­linked glycosylation of 415 is another possible escape mechanism. Using the same analysis pipeline, I also identified additional candidate escape mutations outside the epitope, which could be potential therapeutic targets.

<p>The accumulation of high-throughput data from vast sources has drawn a lot attentions to develop methods for extracting meaningful information out of the massive data. More interesting questions arise from how to combine the disparate information, which goes beyond modeling sparsity and dimension reduction. This dissertation focuses on the innovations in the area of heterogeneous data integration.</p><p>Chapter 1 contextualizes this dissertation by introducing different aspects of meta-analysis and model frameworks for high-dimensional genomic data.</p><p>Chapter 2 introduces a novel technique, joint Bayesian sparse factor analysis model, to vertically integrate multi-dimensional genomic data from different platforms. </p><p>Chapter 3 extends the above model to a nonparametric Bayes formula. It directly infers number of factors from a model-based approach.</p><p>On the other hand, chapter 4 deals with horizontal integration of diverse gene expression data; the model infers pathway activities across various experimental conditions. </p><p>All the methods mentioned above are demonstrated in both simulation studies and real data applications in chapters 2-4.</p><p>Finally, chapter 5 summarizes the dissertation and discusses future directions.</p> / Dissertation

Obesity and type 2 diabetes (T2D) are major risk factors for cardiovascular and other diseases and are currently undergoing an increase in global prevalence. The work presented in my thesis addresses the role epigenetics, specifically DNA methylation, plays in the susceptibility to obesity and T2D and deals with methodological issues in the analysis of DNA methylation data. I first combined epigenome-wide DNA methylation data across 38 adipose tissue samples with corresponding SNP and mRNA data for the same subjects. At 5&percnt; false discovery rate (FDR), methylation of 149 regions associated with at least one cis-SNP. When 19 of the 149 regions were tested for association in an additional 181 independent samples, five regions replicated. These results indicate a genetic influence on DNA methylation in adipose tissue. I then analysed 90 epigenome-wide methylation samples taken from 15 South Asian controls and 30 T2D cases participating in the LOLIPOP study at two time points &sim;7 years apart. I found global differences at both follow-up and baseline between the normal glucose tolerant and T2D groups, as well as strong differences with aging. I further used the main EpiMigrant data from 2,687 individuals, with 36 samples measured in duplicate to assess approaches to quality control, data normalisation and batch correction through control probe adjustment. A null hypothesis for epigenome-wide association studies (EWAS) by permutation testing and I investigated the effects of correlation between individual methylation markers. Using the developed methods, I carried out an EWAS of body mass index (BMI) with subsequent meta-analysis amongst 10,261 individuals of European and South Asian ancestry. DNA methylation markers at 187 genetic loci were associated with BMI. Mendelian randomisation experiments suggested that association of DNA methylation with BMI is the consequence of BMI. Lastly, I tested haplotypes of 85 SNPs currently known to be associated with T2D and 118 SNPs associated with obesity traits for an enrichment of CpG creating or abrogating SNPs and found that 9 T2D and 23 obesity SNPs showed a significant difference in CpG count between the SNP alleles as established by permutation testing. Amongst these is FTO, a locus which has been previously been shown to have a haplotype-specific methylation effect. My work provides novel insights into the role of DNA methylation in metabolic diseases. The methods that I developed to robustly detect association are flexible and scalable and will further be useful for larger, future EWAS.

Orientador: Flávia Lombardi Lopes / Resumo: A meningite se caracteriza pela inflamação da aracnóide, pia-máter e líquor; causando danos às estruturas corticais e subcorticais. As meningites bacterianas estão intimamente relacionadas ao nível de desenvolvimento socioeconômico do país e são consideradas endêmicas no Brasil. Apesar dos esforços para o desenvolvimento de medicamentos e vacinas, a doença ainda possui altos índices de morbidade. Ela ocorre quando bactérias atravessam barreiras de proteção do corpo e alcançam o sistema nervoso central, desencadeando uma resposta imunológica. Sabe-se que durante o curso da doença os níveis de homocisteína do líquor aumentam, promovendo desmielinização e danos neuronais, e que a vitamina B12 é um tratamento utilizado para diminuir danos. Os microRNAs (miRNAs) são instrumentos de resposta fisiológica, tendo sua expressão modificada em diferentes tecidos, em decorrência de diferentes estímulos fisiológicos e patológicos. Estão associados ao controle da expressão de diferentes mediadores inflamatórios e sua ausência é capaz de causar danos severos a resposta imune. Considerando a importância dos miRNAs para regulação de processos imunes o presente estudo visou elucidar os padrões de expressão dos miRNAs durante o processo inflamatório da meningite pneumocócica (MP), bem como observar esses padrões em resposta ao tratamento adjuvante da doença com vitamina B12. Observamos um total de 37 miRNAs diferencialmente expressos, a infecção regulou positivamente 22 deles, e outros 7 negativ... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Meningitis is characterized by an arachnoid, pia mater and cerebrospinal fluid inflammation; causing damage to the cortical and subcortical structures. Bacterial meningitis is closely related to the level of socioeconomic development of country and is considered endemic in Brazil. Despite efforts to develop drugs and vaccines, the disease still has high rates of morbidity and mortality. Occurring when bacteria cross body protection barriers and reach the central nervous system, triggering immune response. It is known that during disease course the the levels of homocysteine increase in cerebrospinal fluid, leading to demyelination and neuronal damage, and that vitamin B12 is a treatment used to reduce those damages. MicroRNAs (miRNAs) are instruments of physiological response, having their expression modified in different tissues, due to different physiological and pathological stimuli. They are associated with expression control of different inflammatory mediators and their absence is capable of causing severe damage to the immune response. Considering the importance of miRNAs in regulation of immune processes, the present study aimed to elucidate miRNA expression patterns during the inflammatory process resulting from pneumococcal meningitis (PM), and to observe these patterns in response to adjuvant treatment of vitamin B12 in infected rats. We observed a total of 37 differentially expressed miRNAs; the infection positively regulated 22 and negatively regulated 7 of them, ... (Complete abstract click electronic access below) / Mestre

Somatic mutations are the main triggers that initiate the formation of cancer. Large sequencing data sets in recent years revealed a substantial number of mutational processes, many of which are poorly understood or of completely unknown aetiology. These mutational processes leave characteristic sequence patterns, often called "signatures", in the DNA. Characterisation of the mutational patterns observed in cancer patients with respect to different genomic features and processes can help to unravel the aetiology and mechanisms of mutagenesis. Here, we explored the effects of DNA modifications and DNA replication on mutagenesis. The most common mutation type, C&GT;T mutations in a CpG context, is thought to result from spontaneous deamination of 5-methylcytosine (5mC), the major DNA modification. Much less is known about the mutational properties of the second most frequent modification, 5-hydroxymethylcytosine (5hmC). Integrating multiple genomic data sets, we demonstrate a twofold lower mutagenicity of 5hmC compared to 5mC, present across multiple tissues. Subsequently, we show how DNA modifications may modulate various mutational processes. In addition to spontaneous deamination of 5mC, our analysis suggests a key role of replication in CpG > TpG mutagenesis in patients deficient in post-replicative proofreading or repair, and possibly also in other cancer patients. Together with an analysis of mutation patterns observed in cancers exposed to UV light, tobacco smoke, or editing by APOBEC enzymes, the results show that the role of DNA modifications goes beyond the well-known spontaneous deamination of 5mC. Finally, we explored which of the known mutational processes might be modulated by DNA replication. We developed a novel method to quantify the magnitude of strand asymmetry of different mutational signatures in individual patients followed by evaluation of these exposures in early and late replicating regions. More than 75 % of mutational signatures exhibited a significant replication strand asymmetry or correlation with replication timing. The analysis gives new insights into mechanisms of mutagenicity in multiple signatures, particularly the so far enigmatic signature 17, where we suggest an involvement of oxidative damage in its aetiology. In conclusion, our results suggest that DNA replication or replication-associated DNA repair interacts with most mutagenic processes.

Breast cancer is a clinically and molecularly heterogeneous disease displaying distinct therapeutic responses. Although recent studies have explored the genomic and transcriptomic landscapes of breast cancer, the epigenetic architecture has received less attention. To address this, an optimised Reduced Representation Bisulfite Sequencing protocol was performed on 1482 primary breast tumours (and 237 matched adjacent normal tissues). This constitutes the largest breast cancer methylome yet, and this thesis describes the bioinformatics and statistical analysis of this study. Noticeable epigenetic drift (both gain and loss of homogeneous DNA methylation patterns) was observed in breast tumours when compared to normal tissues, with markedly higher differences in late replicating genomic regions. The extent of epigenetic drift was also found to be highly heterogeneous between the breast tumours and was sharply correlated with the tumour’s mitotic index, indicating that epigenetic drift is largely a consequence of the accumulation of passive cell division related errors. A novel algorithm called DMARC (Directed Methylation Altered Regions in Cancer) was developed that utilised the tumour-specific drift rates to discriminate between methylation alterations attained as a consequence of stochastic cell division errors (background) and those reflecting a more instructive biological process (directed). Directed methylation alterations were significantly enriched for gene expression changes in breast cancer, compared to background alterations. Characterising these methylation aberrations with gene expression led to the identification of breast cancer subtype-specific epigenetic genes with consequences on transcription and prognosis. Cancer genes may be deregulated by multiple mechanisms. By integrating with existing copy number and gene expression profiles for these tumours, DNA methylation alterations were revealed as the predominant mechanism correlated with differentially expressed genes in breast cancer. The crucial role of DNA methylation as a mechanism to target the silencing of specific genes within copy number amplifications is also explored which led to the identification of a putative tumour suppressor gene, THSZ2. Finally, the first genome-wide assessment of epigenomic evolution in breast cancer is conducted. Both, the level of intratumoural heterogeneity, and the extent of epiallelic burden were found to be prognostic, and revealed an extraordinary distinction in the role of epiclonal dynamics in different breast cancer subtypes. Collectively, the results presented in this thesis have shed light on the somatic DNA methylation basis of inter-patient as well as intra-tumour heterogeneity in breast cancer. This complements our genetic knowledge of the disease, and will help move us towards tailoring treatments to the patient's molecular profile.

Les cancers colorectaux (CCR) constituent un véritable problème de santé publique. En France, en 2015, ils se situent au 3ème rang des cancers et au 2ème rang des décès par cancer, en raison d’un diagnostic souvent tardif et de réponses variables aux thérapeutiques. L’amélioration des tests de dépistage ces vingt dernières années n’a pas significativement réduit la morbi/mortalité liée au CCR et une meilleure caractérisation des différents sous-types moléculaires reste importante pour l’identification de nouveaux biomarqueurs. Les tumeurs colorectales sont hétérogènes tant sur le plan clinico-pathologique que moléculaire, issues de différentes voies de cancérogenèse. Les tumeurs de la voie festonnée représentent 20 à 30% des CCR. Elles sont caractérisées par une instabilité épigénétique (CpG Island Methylation Phenotype, CIMP), une instabilité microsatellitaire (MSI) et des mutations fréquentes du gène BRAF. Sur le plan phénotypique, ce sont des tumeurs volontiers mucisécrétantes surexprimant les mucines MUC2 et MUC5AC. L’identification précoce de leurs précurseurs reste actuellement un enjeu, d’autant que la séquence carcinogénétique des tumeurs festonnées serait particulièrement rapide.Dans cette étude, nous avons déterminé les caractéristiques clinico-pathologiques et moléculaires d’une série de cancers et polypes colorectaux et évalué l’intérêt des gènes de mucines MUC2 et MUC5AC pour classer les polypes festonnés et identifier les lésions précurseurs des cancers CIMP/MSI.L’étude a porté sur 418 CCR et 330 polypes coliques, incluant 218 polypes festonnés et 112 adénomes conventionnels, dont nous avons déterminé le profil moléculaire (mutations KRAS/BRAF, MSI, CIMP, méthylation MGMT, MLH1), ainsi que le profil d’expression et de méthylation des gènes de mucines MUC2 et MUC5AC. Les résultats ont été comparés aux données cliniques et anatomo-pathologiques.Nous montrons que l’hypométhylation du gène MUC5AC est un marqueur spécifique des CCR CIMP/MSI, indépendant des données clinico-pathologiques. De plus, nous montrons que l’hypométhylation de MUC5AC est un évènement précoce qui est spécifiquement associé aux polypes hyperplasiques de type microvésiculaire (MVHP) et adénomes festonnés sessiles (SSA), suggérant une filiation entre ces deux lésions, qui seraient les précurseurs des CCR CIMP/MSI. L’hypométhylation de MUC5AC était par ailleurs très spécifique des lésions festonnées mutées BRAF, CIMP ou MSI.En conclusion, nos résultats suggèrent un rôle de la mucine MUC5AC dans le développement des tumeurs coliques de la voie festonnée. L’hypométhylation du gène MUC5AC est un évènement précoce qui pourrait avoir un intérêt pour le diagnostic des polypes festonnés, en particulier lorsque la morphologie est ambiguë, et pour l’identification précoce des lésions à potentiel malin. De plus, nos résultats suggèrent que certains MVHP pourraient progresser en SSA et en CCR CIMP/MSI, d’où l’importance de réaliser des biopsies des polypes festonnés lors de l'endoscopie afin de repérer le sous-type microvésiculaire parmi les polypes hyperplasiques colorectaux, qui pourrait, bénéficier d’une surveillance minimale. / Colorectal cancer (CRC) is a major public health problem yet it remains the third most common and the second deadliest cancer for both men and women in the United States and in Europe, owing to diagnosis at advanced stage and variable response to the treatments. However, the incidence and mortality appear to be steadily declining in countries with programmatic screening and it remains critical to characterize CRC molecular subtypes to identify new biomarkers. CRC display a wide clinicopathological and molecular heterogeneity and arise from different carcinogenesis pathway. Approximately 20% to 30% of CRC occur by the means of the serrated neoplasia pathway. CRC from the serrated pathway are associated with frequent epigenetic instability (CIMP, CpG island methylator phenotype) which causes most sporadic microsatellite instability (MSI) CRC through epigenetic inactivation of MLH1 and frequent BRAF mutation. CRC from the serrated pathway frequently display a mucinous pattern and are prone to express secreted mucins MUC2 and MUC5AC. Early identification of serrated pathway precursor lesions is currently challenging since the progression to adenocarcinoma in this pathway is particularly brief.We assessed in this study clinicopathological and molecular features of colorectal carcinomas and polyps and we analyzed the interest of mucin genes MUC2 and MUC5AC to classify serrated polyps and to identify the precursor lesions of CIMP/MSI carcinomas._x000D_A series of 418 CRC and 330 polyps was included in the study, with 218 serrated polyps and 112 conventional adenomas. We assessed the molecular profile (KRAS/BRAF mutations, MSI, CIMP, MGMT methylation, MLH1) and the expression and methylation profile of MUC2 and MUC5AC genes and further correlation with clinical and pathological data were performed.We show that MUC5AC hypomethylation is a specific marker of CIMP/MSI CRC, independently of other clinical or pathological factors. We show moreover than MUC5AC hypomethylation is an early event of carcinogenesis, specific to microvesicular hyperplastic polyps (MVHP) and sessile serrated adenoma (SSA) suggesting a filiation between these two lesions and also a progression of MVHP to SSA, SSA with dysplasia, and then to MSI/CIMP CRC. MUC5AC hypomethylation was moreover highly specific of BRAF mutated, CIMP or MSI serrated lesions.In conclusion, our results suggest that the MUC5AC mucin is involved/plays a role in the development/progression of colorectal tumors of the serrated pathway. MUC5AC hypomethylation occurs early during carcinogenesis and may be of interest for the diagnosis of serrated polyp especially in the presence of ambiguous morphology and to early detection of polyps carrying a malignant potential. Our results suggest moreover that some MVHP may progress to SSA and CIMP/MSI CRCs highlighting the importance to perform biopsy of serrated polyps during endoscopy to identify microvesicular subtype among colorectal hyperplastic polyps, which could require a minimal surveillance.

<p>In eukaryotic systems, the genetic material of the cell –DNA– is packed into a protein-dense structure called chromatin. Chromatin structure is critical for preservation of the genetic material as well as coordination of vital processes such as DNA replication, transcription and DNA damage repair. The fundamental repeating unit of chromatin is nucleosome which is composed of an octamer of small alkaline proteins called histones and the DNA wrapped around this octamer. The nucleosomes are then packed into higher-order structures leading to formation of 3D chromatin architecture. The chromatin is a dynamic structure; the spacing between nucleosomes, or the folding of the larger chromatin segments is subjected to alterations during embryonic development, tissue specifications or <i>simply during any event that require gene expression changes</i>. Failure in proper regulation of chromatin structure has been associated with embryonic defects and disease such as cancer. </p> <p>This work has focused on a class of ATP-dependent chromatin remodeling complexes known as switch/sucrose-non-fermentable (SWI/SNF) or BRG-associated factors (BAF) complex. This family of complexes act on chromatin and alter its physical structure by mobilizing histones or nucleosome particles through the activity of its ATPase –BRG1 or BRM, enabling more accessible DNA for the other factors such as transcription factors to localize and recruit transcription machinery. In particular, we discovered and biochemically defined a novel version of this family of chromatin complexes that we named as GLTSCR1/1L-BAF (GBAF). GLTSCR1 and GLTSCR1L are two uncharacterized paralogous proteins that have been identified as BRG1-interacting proteins. Biochemically surveying the essence of this interaction, we realized that these proteins incorporates into a previously unknown SWI/SNF family complex that lacks well-characterized SWI/SNF subunits such as ARID1/2, BAF170, BAF47; instead, uniquely comprise GLTSCR1/1L and bromodomain-containing protein BRD9. Focusing on the GLTSCR1 subunit, we observed that its absence is well-tolerated by many different cell types except slight growth retardation by prostate cancer cells. Expanding the cohort of prostate cancer cells, we realized that not the paralogous subunits GLTSCR1 or GLTSCR1L but unique and non-redundant subunit BRD9 is the major GBAF-dependence in prostate cancer cells. We observed that especially the androgen-receptor positive cell lines have severe growth defects upon <i>BRD9 </i>knockdown or inhibition. <i>In vivo, </i>we showed that xenografts with <i>BRD9 </i>knockdown prostate cancer cells (LNCaP) have smaller tumor size. We demonstrated that BRD9 inhibition can block the expression of androgen-receptor targets. Similarly, <i>BRD9 </i>knockdown and treatment with antiandrogen drug (enzalutamide) has overlapping transcriptional effects. Mechanistically, we showed that BRD9 interacts with AR and it colocalizes with AR in subset of AR -binding sites. Surprisingly, we realized that BRD9 depletion has similar transcriptional and phenotypic effects as BET protein inhibitors. BET protein family contains 4 bromodomain containing proteins (BRD2, BRD3, BRD4, BRDT). These proteins were previously shown to be critical for AR-dependent gene expression. We detected interaction between BRD9 and BRD2/4. We demonstrated that BRD4 and BRD9 had shared binding sites on genome, a fraction of which are co-bound by AR. At particular target sites we showed that BRD9 localization is dependent on BET proteins, but not the other way around. Taking together, we provided some evidences that GBAF targeting through BRD9 can be a novel therapeutic approach for prostate cancer. Growing body of reports suggested that current therapy options targeting the androgen receptor is failing due to acquired resistance. Therefore, targeting the AR pathways via its coregulators such as BET proteins or SWI/SNF complexes can serve as potent alternative approaches. Further research is needed to elucidate the roles of GBAF and BET proteins in androgen receptor independent prostate cancer cells, which are still responsive to GBAF or BET manipulations although to a lesser extent.</p>