Abstract<p>Context: Breast cancer treatment has experienced several changes in the last decades due to the innovation of specific prognostic and predictive biomarkers that facilitate the application of more personalized therapies to different molecular sub-groups. Presently, more women are be- ing treated with neoadjuvant (preoperative) therapy which involves chemotherapy or endocrine agents before surgery, for earlier-stage operable breast carcinoma. Following this mode of pre- operative systemic treatment could improve the surgical option and make inoperable tumors operable. It can also increase the breast conservation rate. Another key benefit of neoadjuvant therapy is monitoring response to the treatment. The good response to neoadjuvant therapy with complete pathological response (pCR) is a surrogate marker for overall survival.<p>Objective: 1) To investigate the association between early changes in several gene expression signatures, recapitulating several biological processes, and neoadjuvant letrozole (endocrine therapy), and to compare those to Ki67 values. 2) To interrogate the association between chemotherapy response (Pathological complete response (pCR) in this case) and gene expres- sion modules, recapitulating important biological processes such as the gene expression grade index (GGI) and ”druggable”oncogenic pathways in different breast cancer subtypes.<p>Data Sources: We collected publicly available gene expression data based on the review of selected literature on breast carcinoma after neoadjuvant therapy with the clinical and patho- logic characteristics.<p>Results: In this work we have shown, 1) Residual proliferation after short-term endocrine therapy can be used as an early surrogate marker of clinical to response to endocrine therapy in this population. 2) Different processes and pathways are associated with pCR in different BC subtypes.<p>Conclusions: Our analysis has several limitations such as: 1) Lacks of statistical power due to small dataset for endocrine treated patients, 2) We have included only anthracycline-based neoadjuvant chemotherapy regimens; therefore, it is not known if the associations between gene modules and pCR are anthracycline specific or indicate general chemosensitivity. More- over, patients with HER2-positive tumors did not receive preoperative trastuzumab, and it is not known how this could modulate the identified associations. But our results generate sev- eral hypotheses that should be tested in BC subtype - focused trials of targeted agents like IGF1, PARP inhibitors, and agents modulating immune response. If results are confirmed by additional validation studies, this may lead to a paradigm shift in early breast cancer treatment. / Doctorat en Sciences biomédicales et pharmaceutiques / info:eu-repo/semantics/nonPublished

Les cancers du sein de sous-type luminal B sont associés à un mauvais pronostic. Afin de mieux comprendre la biologie de ce sous-type, nous avons étudié au sein de 188 tumeurs mammaires de différents sous-types, les anomalies du nombre de copies, les méthylations de l'ADN, les profils d'expression génique et les mutations somatiques dans 9 gènes sélectionnés. Un total respectif de 237 et 101 oncogènes et gènes suppresseurs de tumeurs (TSG) candidats présentaient une dérégulation de l'expression en relation avec leur CNA. 88% des TSG potentiels étaient localisés sur le bras chromosomique 6q. 101 oncogènes candidats ont été validés sur une série publique de 5765 cancers du sein, et l'expression de 67 gènes était associée à un mauvais pronostic au sein des tumeurs luminales. 24 gènes présentaient une dérégulation de l'expression en relation avec un haut niveau de méthylation de l'ADN. FOXO3, PIK3CA et TP53 étaient les gènes les plus fréquemment mutés parmi les 9 testés. Dans une méta-analyse de séquençage de nouvelle génération regroupant 875 cancers du sein, les gènes les plus fréquemment mutés dans le sous-type luminal B étaient PIK3CA, TP53 et GATA3. Les nombreuses altérations moléculaires ciblaient des voies de signalisation communes, incluant 3 axes pouvant jouer un rôle majeur dans le sous-type luminal B : la voie TP53 et l'instabilité chromosomique, les voies de signalisation PI3K/AKT/MTOR/FOXO et MAPK/JNK, et les altérations des facteurs de transcription et épigénomiques. En conclusion, nous avons établi un répertoire de gènes candidats dans le sous-type luminal B qui pourraient être impliqués dans le développement et/ou l'hormonorésistance de ce sous-type. / Breast cancers (BCs) of the luminal B subtype have a poor prognosis. To better understand this subtype we studied in 188 BCs of various molecular subtypes, DNA copy number aberrations, DNA promoter methylation, gene expression profiles, and somatic mutations in nine selected genes. A total of 237 and 101 luminal B-specific candidate oncogenes and tumor suppressor genes (TSGs) presented a deregulated expression in relation with their CNAs. Interestingly, 88% of the potential TSGs are located within chromosome arm 6q. 101 candidate oncogenes were validated in a public series of 5,765 BCs and the overexpression of 67 was associated with poor survival in luminal tumors. 24 genes presented a deregulated expression in relation with a high DNA methylation level. FOXO3, PIK3CA and TP53 were the most frequent mutated genes among the nine tested. In a meta-analysis of next-generation sequencing data in 875 BCs, PIK3CA, TP53 and GATA3 were the most frequent mutated genes. Numerous molecular alterations targeted common signalling pathway, included 3 ways wich may play a major in the luminal B subtype: TP53 pathway and chromosomal instability, PI3K/AKT/MTOR/FOXO and MAPK/JNK pathway, and epigenomic and transcription factors alterations. In conclusion, we have reported a repertoire of luminal B candidate genes that may be involved in the development and/or hormone resistance of this subtype.

Polyploidization instantly doubles all genome content by combining two genomes that have markedly different methylation and gene expression levels. This process may be accompanied by genetic and epigenetic changes in each genome. Sequencing of the transcriptome (RNA-seq) and the methylome (bisulfite treated libraries whole genome libraries) were used to measure gene expression and methylation levels of genic regions of allopolyploid cotton petals and petals of their diploid relatives. Many differentially expressed genes detected by RNA-seq were consistent with expression levels previously detected by microarrays. RNA-seq results also reconfirmed the presence of general polyploid gene expression trends like expression level dominance and homoeologous expression biases in Gossypium polyploid species. Expression biases between A- and D-genome homoeologs and expression level dominance was characterized for thousands of genes in tetraploids and a diploid F1-hybrid. Unlike the results of microarray study previously done we found a slightly greater number of genes showing A-genome bias vs genes showing D-genome bias. More commonly the overall expression level from homoeologs of polyploid is heterotic i.e the expression level is greater than the average of the expression levels from the two parent genomes. In addition, genome methylation (CG, CHG, and CHH contexts) of each genome was assessed in the diploid and tetraploid samples. The A- and D-genomes had distinct levels of DNA methylation for each context. DNA methylation may be independently regulating homoeologous expression levels of a small number of genes.

Background: The genes coding for human (h) chorionic somatomammotropin (CS), hCS-A and hCS-B, and placental growth hormone (GH-V), hGH-V are located at a single locus on chromosome 17q22-24. Local regulatory (5´ P and 3´ enhancer) sequences and a remote locus control region (LCR) containing a placenta-specific hypersensitive site (HS) IV, have been implicated in the efficient expression of the placental hCS/GH-V genes, in part through gene transfer studies in placental and non-placental tumor cell lines. However, low levels of endogenous expression are reported in placental tumor cells compared to normal term placenta. Thus it was hypothesized that the hCS/GH-V chromatin structure in human choriocarcinoma cells is less accessible to regulatory regions essential for efficient expression due to DNA and/or histone modifications, specifically methylation and acetylation, respectively. Approach: To assess individual hCS-A, hCS-B and hGH-V gene expression in placental and non-placental tumor cells, and assess the effect of increasing “chromatin accessibility” on hCS/GH-V RNA levels by inhibiting DNA methylation and histone deacetylation using 5-aza-2´-deoxycytidine (azadC) and trichostatin A (TSA). Principal Findings: Low levels of hCS-A, hCS-B and hGH-V RNA were detected in placental and non-placental tumor cells compared to term placenta. A significant >5-fold increase in promoter activity was seen in placental but not non-placental cells transfected with hybrid hCS promoter luciferase genes containing 3´-enhancer sequences. Placental JEG-3 cells pretreated with azadC and TSA resulted in a significant >10-fold increase in hCS-A, hCS-B and hGH-V RNA levels compared to TSA treatment alone, however, a modest ~3-fold effect was seen in non-placental MCF-7 cells. By contrast to the effect of pretreatment with azadC, post-treatment with azadC mutes the stimulatory effects of TSA on hCS/GH-V transcripts. The specificity of the response suggests that azadC treatment, and presumably hypomethylation of DNA, results in an increase in response to TSA and histone hyperacetylation at the hGH/CS locus. An assessment of histone H3/H4 hyperacetylation in JEG-3 cells treated with azadC and TSA versus TSA alone revealed significant increases consistent with a more open chromatin structure including the hCS 3´-enhancer sequences and LCR. Conclusions: These observations suggest that accessibility of remote and local regulatory regions required for efficient placental hGH/CS expression can be restricted by DNA methylation and histone acetylation status. This includes restricting access of the hCS 3´-enhancer sequences to available placental enhancer transcription factors. / May 2016

Adipose tissue is not homogenous as individual fat depots display regional variation in their physiological properties. It follows that body fat distribution is increasingly being recognised as a major determinant of metabolic disease risk. At the cellular level, depot-specific properties are exhibited by adipocyte precursors during in vitro culture and persist for many generations, suggesting these cells retain an ‘intrinsic memory’ of their anatomical origin which is epigenetic in nature. A primary aim was to identify depot-specific genes whose expression may be regulated by DNA methylation in adipose precursors. Using two paired preadipocyte cell lines derived from human subcutaneous abdominal and gluteal adipose tissue - to represent upper and lower body fat with their opposing associations with cardiovascular disease and diabetes respectively - depot-specific gene expression and DNA methylation profiles were detected. Furthermore, the expression of certain genes in preadipocytes was found to change in response to treatment with the DNA demethylating agent 5-azacytidine, which suggests DNA methylation may regulate depot-specific gene expression. A secondary aim was to investigate whether glucocorticoids – which are important determinants of body fat distribution – exert their effects through DNA methylation. The synthetic glucocorticoid dexamethasone was found to modulate the expression of some of the differentially expressed genes in preadipocytes, with this effect possibly being mediated by DNA methylation. It has been postulated that depot-specific phenotypes in adipose tissue may arise from developmental differences. Several genes were found to be expressed in a depot-specific fashion during a differentiation time course, suggesting regional variation in adipogenesis may contribute to the generation of depot-specific phenotypes. Overall, the data presented suggests regional variation within subcutaneous white adipose tissue exists and supports the notion that DNA methylation patterns can, in part, determine adipose tissue heterogeneity.

DNA methylation refers to an addition of a methyl group to the 5 position of the cytosine pyrimidine ring. As the best characterized epigenetic mark, DNA methylation plays an important role in a plethora of biological functions, including gene repression, genomic imprinting, silencing of retro-transposons and X chromosome inactivation. Genomic imprinting refers to the mono-allelic expression of certain genes according to their parent-of-origin. In mammals, the expression of imprinted genes is controlled by the cis-acting regulatory elements, termed imprinted control regions (ICRs). ICRs are marked by parent-of-origin-specific DNA methylation and loss of DNA methylation at ICRs also causes aberrant expression of imprinted genes. Therefore it is believed that the genomic imprinting is a DNA methylation-associated epigenetic phenomenon. As accurate expression of imprinted genes is essential for normal embryonic growth, energy homeostasis, development of the brain and behaviour and abnormal expression of imprinted genes leads to numerous clinical phenotype and human disorders, it is important to investigate how the imprinted DNA methylation is stably maintained in mammals. DNA methyltransferases (DNMTs) are the main enzymes that play a in the establishment and maintenance of imprinted DNA methylation. In primordial germ cells (PGCs), DNMT3A and DNMT3L are involved in the establishment of imprinted DNA methylation. Whereas once established, the imprinted DNA methylation is maintained by DNMT1, DNMT3A and DNMT3B, but mainly by DNMT1. In addition, some other enzymes and DNA binding proteins also play a role in this process. One of the best examples is ZFP57, which forms a complex with KAP1 and SETDB1. ZFP57 maintains imprinted DNA methylation by recognizing a methylated hexa-nucleotide and recruits DNMTs to the ICRs in mammalian embryonic stem (ES) cells. Interestingly, DNA methylation analysis combined with promoter microarrays carried out in our lab suggested that imprinted DNA methylation is absent from some of the maternal ICRs in ES cells genetically null for G9a, a histone H3 lysine 9 methylase. This indicates that G9a might also play a role in the maintenance of imprinted DNA methylation. In my work, I found that the repressive H3K9me2 and imprinted DNA methylation are absent from several analysed ICRs in embryonic stem (ES) cells genetically null for either G9a or its partner histone methyltransferase GLP. A knockdown of G9a in ES cells reproduced these observations suggesting that G9a/GLP complex is required for the maintenance of imprinted DNA methylation. I also found that neither wild type nor catalytically inactive G9a can restore the loss of imprinted DNA methylation in G9a-/- ES cells. Chromatin immunoprecipitation (ChIP) combined with bisulfite DNA sequencing showed that imprinted DNA methylation was present on the H3K9me2-marked allele indicating a direct role for G9a in maintenance of genomic imprinting. Using a pharmacological inhibitor of G9a and mutagenesis analyses, I found that G9a maintains the imprinted DNA methylation independently of its catalytic activity and recruits DNMTs to the ICRs via its ankyrin repeat domain. Dimerization of G9a with GLP is also essential for the maintenance of genomic imprinting in ES cells. In summary, in addition to establish H3K9me2, histone methyltransferases G9a and GLP also play an essential role in the maintenance of genomic methylation imprints in ES cells.

Background: Mutator-like transposable elements, a class of DNA transposons, exist pervasively in both prokaryotic and eukaryotic genomes, with more than 10,000 copies identified in the rice genome. These elements can capture ectopic genomic sequences that lead to the formation of new gene structures. Here, based on whole-genome comparative analyses, we comprehensively investigated processes and mechanisms of the evolution of putative genes derived from Mutator-like transposable elements in ten Oryza species and the outgroup Leersia perieri, bridging similar to 20 million years of evolutionary history. Results: Our analysis identified thousands of putative genes in each of the Oryza species, a large proportion of which have evidence of expression and contain chimeric structures. Consistent with previous reports, we observe that the putative Mutator-like transposable element-derived genes are generally GC-rich and mainly derive from GC-rich parental sequences. Furthermore, we determine that Mutator-like transposable elements capture parental sequences preferentially from genomic regions with low methylation levels and high recombination rates. We explicitly show that methylation levels in the internal and terminated inverted repeat regions of these elements, which might be directed by the 24-nucleotide small RNA-mediated pathway, are different and change dynamically over evolutionary time. Lastly, we demonstrate that putative genes derived from Mutator-like transposable elements tend to be expressed in mature pollen, which have undergone de-methylation programming, thereby providing a permissive expression environment for newly formed/transposable element-derived genes. Conclusions: Our results suggest that DNA methylation may be a primary mechanism to facilitate the origination, survival, and regulation of genes derived from Mutator-like transposable elements, thus contributing to the evolution of gene innovation and novelty in plant genomes.

Integrin α6β4 is upregulated in pancreatic carcinoma, where signaling promotes metastatic properties, in part by altering the transcriptome. Such alterations can be accomplished through DNA demethylation of specific promoters, as seen with the pro-metastatic gene S100A4. I found that signaling from integrin α6β4 dramatically upregulates expression of amphiregulin (AREG) and epiregulin (EREG), ligands for the epidermal growth factor receptor (EGFR), and that these ligands promote pancreatic carcinoma invasion. To determine if AREG and EREG are regulated by DNA methylation, pancreatic cancer cells with low AREG and EREG expression were treated with the DNA methyltransferase inhibitor 5-aza-2’-deoxycytidine (5-Aza-CdR), resulting in stable overexpression of AREG and EREG, and this induction required signaling from integrin α6β4. Similarly, treatment of cells with high integrin α6β4 with the methyl donor S-adenosylmethionine inhibited gene expression of AREG and EREG. Whole genome bisulfite sequencing on pancreatic cancer cells reveled hypomethylation of the promoter regions of AREG and EREG when integrin α6β4 is high, and these regions correspond to H3K27Ac, indicative of enhancer location. Interestingly, I also observed genome-wide DNA demethylation, and a large proportion of altered CpGs correspond to potential enhancers. It is currently accepted that active DNA demethylation occurs via DNA repair. I tested this hypothesis by treating cells with Gemcitabine, which inhibits multiple components of DNA repair, including DNA demethylation mediated by GADD45A. Gemcitabine treatment resulted in marked reduction in AREG and EREG expression. To further test the involvement of GADD45A, I used RNAi-mediated knockdown or cDNA overexpression to alter GADD45A levels. In both instances, AREG and EREG expression positively correlated with GADD45A, particularly when integrin α6β4 is high, indicating that GADD45A is a rate-limiting step in AREG and EREG overexpression. Similarly, using stable shRNA, I show that Thymine DNA Glycosylase (TDG), and TET1 known modulators of DNA demethylation, are required for AREG and EREG expression in integrin α6β4 high cells, and nuclear localization of TDG is much higher in cells with high integrin α6β4. Using a specific inhibitor I found that AREG and EREG expression is dependent on Parp-1. Finally, I determined that integrin α6β4 signaling enhances cells ability to respond to and survive in the presence of DNA damage, and that active DNA repair is required for integrin α6β4 mediated DNA demethylation. Taken together, these data indicate that DNA repair is required to maintain overexpression of AREG and EREG in response to signaling from integrin α6β4 and that integrin α6β4 promotes this overexpression by enhancing DNA repair.

A controlled and efficient coordination of gene expression is the key for normal development of an organism. In mammals, a subset of autosomal genes is expressed monoallelically depending on the sex of the transmitting parent, a phenomenon known as genomic imprinting. The imprinted state of the H19 and Igf2 genes is controlled by a short stretch of sequences upstream of H19 known as the imprinting control region (ICR). This region is differentially methylated and is responsible for the repression of the maternally inherited Igf2 allele. It harbors hypersensitive sites on the unmethylated maternal allele and functions as an insulator that binds a chromatin insulator protein CTCF. Hence the H19 ICR, which plays an important role in maintaining the imprinting status of H19 and Igf2, was shown to lose the insulator property upon CpG methylation. Another ICR in the Kcnq1 locus regulates long-range repression of p57Kip2 and Kcnq1 on the paternal allele, and is located on the neighboring subdomain of the imprinted gene cluster containing H19 and Igf2, on the distal end of mouse chromosome 7. Similarly to the H19 ICR, the Kcnq1 ICR appears to possess a unidirectional and methylation-sensitive chromatin insulator property in two different somatic cell types. Hence, methylation dependent insulator activity emerges as a common feature of imprinting control regions. The protein CTCF is required for the interpretation and propagation of the differentially methylated status of the H19 ICR. Work in this thesis shows that this feature applies genomewide. The mapping of CTCF target sites demonstrated not only a strong link between CTCF, formation of insulator complexes and maintaining methylation-free domains, but also a network of target sites that are involved in pivotal functions. The pattern of CTCF in vivo occupancy varies in a lineage-specific manner, although a small group of target sites show constitutive binding. In conclusion, the work of this thesis shows that epigenetic marks play an important role in regulating the insulator property. The studies also confirm the importance of CTCF in maintaining methylation-free domains and its role in insulator function. Our study unravels a new range of target sites for CTCF involved in divergent functions and their developmental control.

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.