Factors Governing the Recurrence and Breakpoint Distribution of Chromosomal Translocations in Human Tumors (pp. 255-263)
Authors: (Francisco J. Novo, Department of Genetics, University of Navarra Spain)
Abstract: Chromosomal translocations, a frequent finding in cancer cells, are the result of
double-strand breaks that are repaired via non-homologous end-joining in somatic cells.
Since many translocations are recurrent and the breakpoints are non-randomly
distributed, a widely held view posits that local sequence factors are responsible for the
appearance of double-strand breaks at specific genomic sites. However, efforts to
identify such factors are complicated by their widespread occurrence throughout the
genome. Thus, establishing a convincing direct causal link between specific sequence
elements and nearby breakpoints has been difficult.
Other potential factors that could account for the recurrence of chromosomal
translocations are frequently overlooked. For instance, the importance of chromosome
localization inside the cell nucleus, and the relative positions of chromosomes with
respect to each other, could determine which genes participate in specific translocations.
Therefore, studies that aim to identify the causes of recurrence of chromosomal
translocations should take into account factors responsible for the nuclear co-localization
of specific genes.
Likewise, little attention has been given to the power of functional selection in
determining the identity of the genes translocated and the localization of breakpoints
within those genes. However, careful analysis of reciprocal translocations that create
chimeric fusion proteins could show whether recurrence is the result of strong cellular
selection pressures for the functions encoded by specific genes. Furthermore, the
requirement to keep the reading frame in the fusion product could explain, in these cases,
the non-random distribution of translocation breakpoints across those genes.
I propose that the generation of DNA double-strand breaks might be widespread
throughout the genome, resulting in a relatively high number of potential translocation
events. Most of these would never take place because the genes involved are far apart
within the nuclear space. Of those rearrangements that would be allowed, only a small
subset could generate an in-frame fusion product with oncogenic potential. Thus, the
non-random genomic distribution of translocation breakpoints in tumor samples might be
due to the fact that these samples contain the limited collection of rearrangements that
were able to survive distinct functional requirements imposed by cellular selection