11 January 2010
Cancer cell genomes are known to contain a variety of structural rearrangements that are thought to play a role in the development of cancer. The involvement of structural changes in driving tumour formation has long been recognised, at scales ranging from whole chromosome translocations (such as the Philadelphia translocation associated with chronic myelogenous leukemia), down to much smaller-scale rearrangements. However, much remains unknown about the nature and role of these changes in driving cancer.
Now, a new paper reports high resolution mapping of the chromosome rearrangements in human breast cancer cells. In all, 24 different breast cancers were examined for the number, position and nature of these structural changes [Stephens PJ et al. (2009) Nature 462(7276):1005-10]. Most rearrangements occurred within chromosomes (as opposed to between different chromosomes) and the most common type were tandem duplications; the authors propose that this may due to a specific (unknown) defect in DNA maintenance in breast cancer cells that generates - or fails to correct - this type of rearrangement.
The researchers also observed correlation between the chromosomal rearrangements and different sub-types of breast cancer; for example, oestrogen- and progesterone-receptor negative, basal-like tumours typically had many tandem duplications, whereas oestrogen-receptor positive luminal type tumours had fewer rearrangements and mostly within repeat regions of the genome.
In many cases rearrangements affected genes and resulted in altered or novel gene products, but none of these were recurrent, suggesting that many different, individually rare mutations may be active in breast cancer cells. However, the authors themselves note that much larger numbers of tumours need to be studied before conclusions may be drawn.
Comment: The observation that breast cancer cell genomes, even from a small sample, show a higher diversity and number of structural rearrangements than expected suggests the need for major new research into how different genomic changes are involved in cancer. It will be interesting to see how far genomic features mirror other characteristics by which tumours are classified into these sub-classes; it may also help to identify key targets for new therapeutics common to some tumour sub-groups, improving prospects for more personalised treatment of breast cancer.