Function of the ATP-dependent chromatin remodeler Mi-2 in the regulation of ecdysone dependent genes in Drosophila melanogaster

The development of the fruitfly Drosophila melanogaster is regulated by the steroid hormone ecdysone. Ecdysone is released at the onset of metamorphosis and initiates a cascade of transcriptional events. First, it leads to the heterodimerisation of the Ecdysone receptor (EcR) with its binding partner ultraspiracle. This complex recruits the transcription machinery to ecdysone inducible genes and thereby initiates transcription of genes that contribute to pupariation and metamorphosis. ATP-dependent chromatin remodelers regulate transcription by altering DNA accessibility and often reside in multimeric protein complexes. Mi-2 is a member of the CHD family of ATP-dependent chromatin remodelers and can function both as co-repressor and co-activator in transcription regulation. The results described in this thesis investigate the function of the chromatin remodeler Mi-2 in the regulation of ecdysone dependent genes. Further, they provide a model by which Mi-2 is targeted to and influences transcription of ecdysone dependent genes. In the first part of this thesis, genome-wide Mi-2 binding sites were mapped by chromatin immunoprecipitation followed by DNA-Sequencing (ChIPSeq) in untreated and ecdysone treated Drosophila S2 cells. This led to the identification of 103 Mi-2 binding sites that show increased binding of Mi-2 upon hormonal stimulation. Further analyses showed that a significant proportion of these binding sites resides in the close proximity of ecdysone inducible genes, implicating that Mi-2 functions in the regulation of these loci. Six ecdysone induced Mi-2 binding sites at two ecdysone dependent genes, the vrille and the broad loci were investigated in more detail. Here, depletion of Mi-2 resulted in a strong increase in expression of these genes in untreated and ecdysone treated cells. However, depletion of a different ATP-dependent chromatin remodeler, Iswi, did not result in derepression of broad and vrille, indicating that Mi-2 function is specific at the broad and vrille genes. In the second part of this thesis, interaction studies revealed that Mi-2 can bind to EcR. This interaction was found to be independent of the hormone ecdysone. Further, the interaction between Mi-2 and EcR was mapped to the ATPase domain of Mi-2. These results demonstrated the first described interaction between the catalytic domain of Mi-2 and a nuclear receptor. In addition, the activation function 2 (AF2 domain) of EcR was found to be important for the interaction with Mi-2. The finding that Mi-2 and EcR can physically interact led to the hypothesis that EcR can recruit Mi-2 to specific sites in the genome. Indeed, a significant overlap between EcR and Mi-2 binding sites was found in both untreated and ecdysone treated cells. In agreement with this hypothesis, depletion of EcR led to decreased ecdysone induced Mi-2 recruitment to the vrille and broad genes. These findings established a new recruitment model for Mi-2 by EcR to chromatin. Finally, Micrococcal nuclease (MNase) mapping demonstrated that Mi-2 functions at the vrille gene by maintaining a closed chromatin structure at this locus. Here, depletion of Mi-2 resulted in a more open chromatin structure, which correlated with an increase in expression of vrille