The breaking and rejoining of different chromosomes can lead to serious medical conditions, but the cause of these translocations was thought to be largely random. It now emerges that sequence variation is probably responsible for the fracture-prone sites in our DNA.

Translocations are among the most frequent genetic aberration in humans — in fact, many of us carry them around without any noticeable effect on health. Because translocations occur sporadically in so many individuals, the molecular mechanism by which they arise has been difficult to pinpoint. Kato and colleagues have now concentrated on a recurrent translocation between human chromosomes 11 and 22; they noticed that many unrelated translocations occurred within palindromic AT-rich repeats (PATRRs).

The authors then took a closer look at the palindrome on chromosome 11 (PATRR11). In most normal individuals the palindrome is about 450 bp; however, other shorter variants are also present in the healthy population, which probably arise by deletions within the 450-bp allele. But are these size variants functionally important? To find out, the authors looked to see how frequently translocations occurred de novo in sperm that carried various combinations of the PATRR11 alleles. What they found was a startling variation of three orders of magnitude in the frequency of translocation, depending on the repeat size — from 1 in 104 in the homozygote for the longest PATRR11 variant to 1 in 107 in a heterozygote for shorter variants.

This is the first time that sequence variation has been found to underlie human translocations. The fact that PATRR-like sequences have been found at the breakpoints of translocations between other chromosomes reinforces the causal role of palindromes in these aberrations.