Mechanism of Hox-gen regulation by the protein SatB2

The correlation between the subnuclear localisation of a gene or locus and its expression status is becoming an intensive field of study in molecular biology. The nucleus is by no means a random mixture of DNA and proteins but rather a highly ordered and compartimentalized entity. Chromosomes, for an instance, assume preferred positions within the nucleus and their non-random distribution in interphase is heritable over cell divisions. In general it is believed that gene rich chromosomes preferrably localise in the interior of the nucleus whereas gene poor chromosomes localizes at the periphery. This points at a correlation between radial postioning of loci and their expression state. Indeed, loci like the immunoglobulin cluster or the beta-globin locus migrate away from the nuclear periphery upon activation and assume central postions when their full activity can be observed. The gross re-localisation of entire genes however is not the ultimate means by which active and inactive sequences can be physically seperated. Recently it was shown that upon activation individual genes of a multigene cluster - the HoxB cluster - loop out of their chromosome territory and fold back towards it when they are not being transcribed any longer. This looping out of the "solid" and inaccessible bluk chromosome territory into the so called interchromatin compartment is believed to facilitate transcription factor accessibility which can change the transcriptional state of this particular gene and lead to the epigenetic reprogramming of this locus. Proteins mediating such higher order chromatin reorganisations and subnuclear movements are still largely unknown. The SatB2 protein was identified as a protein binding to the immunoglobulin heavy chain enhancer. Its close relative SatB1 is known to regulate T cell specific gene expression by modulating long range chromatin interactions, loop formation and epigenetic landscaping of its target genes. Loss of function experiments have revealed that SatB2 deficient mice have severe skelletal defects and that SatB2 deficient osteoblasts show deregulated Hox gene expression. In addition, SatB2 binds to the HoxB cluster and directly represses its expression. Given the close molecular relationship between SatB2 and SatB1, we consider SatB2 a candidate regulator of the HoxB multigene cluster. We are using both osteoblastic cells as well as ES cells to analyse SatB2`s role in the regulation of the HoxB cluster. ES cells have the tremendous advantage that upon differentiation they activate the HoxB cluster in a co- linear manner. Using this system it was shown that the activity of HoxB genes correlates with their subnuclear localisation. Using these two cell types we will address the mechanisms by which SatB2 regulated the HoxB cluster. We are mainly interested in the effect of SatB2 on the subnuclear localisation of the HoxB cluster as well as on the accompanying changes in the chromatin configuration of this locus.