Chromosome dynamics and genomic instablity in neuroblastoma. Three genomic pillars: MYCN amplification, numerical and structural changes.

In this thesis, the main focus has been on the childhood cancer neuroblastoma, one of the most common and lethal childhood tumours. Neuroblastoma has througout the years continued to be a clinical and biological enigma. Our first focus was on one of the most important biological risk factors in neuroblastoma -- amplification of the oncogene MYCN in the tumour cells. Because amplified MYCN typically reside in ring-formed chromatin structures lacking centromeres (so-called double minutes, DMs) it remained unknown for a long time how the amplified sequences were maintained in the growing tumour. We could show that MYCN-carrying DMs in neuroblastoma cells translocate from the nuclear interior to the periphery at the interphase-prophase transition and that they are preferentially anchored to human chromosomes at sites adjacent to the telomeres, resulting in a random segregation pattern of DMs to post-mitotic neuroblastoma cells. Furthermore, by making human/murine hybrids we showed that DMs do not bind to specific positional elements in human chromosomes. Our data explain the vast intercellular variety of MYCN copy number in neuroblastoma. Moving on from here, in our next study we found that telomeres without detectable TTAGGG-repeats were associated with MYCN amplification and the generation of chromosomal breakage-fusion bridge cycles and could confirm that MYCN amplification was associated with reduced tumour telomere length in vivo. We also found a possible association between poor survival and elongated telomeres, which needs to be studied further. Our third pillar was that of whole chromosome changes, with a focus on intratumoural diversity. We demonstrated a previously unreported high degree of intercellular variation in chromosome copy number and found indications that loss of chromosomes from a tetraploid state is a major route towards this prominent intra-tumour genomic diversity in aneuploid neuroblastomas. Taken together, these studies suggest that neuroblastoma genomes are highly plastic, which may to some extent explain the poor response to oncological treatment for some of these tumours