GNN - Genome News Network  
  Home | About | Topics
Quirks of Genomic Disease
Palindromes, hairpins, and breakpoints
Edward R. Winstead

Featured Article.

At a recent laboratory meeting in Philadelphia, a researcher wrote "no lemon, no melon" on a blackboard. He then deleted the spaces and comma, leaving the palindrome "nolemonnomelon." Finally, he turned the letters into the shape of a vertical hairpin, with the palindrome halved and the same letters facing each other. The point of the exercise was to explain the concept of palindromic hairpins in the human genome.

Image of human chromosomes showing balanced translocation. "11" is normal copy of chromosome 11 (red marker) and "22" is normal copy of chromosome 22 (green marker). "der11" is derived from C11 with a bit of C22, and "der22" is derived from C22 with a bit of C11.

The researchers at the Philadelphia meeting have published findings that implicate palindromic hairpins in a puzzling disorder that is also an example of genomic disease. Stretches of palindromic DNA in the shape of hairpins may be a structural flaw in the genome that makes certain chromosomes prone to breakage; this instability could be the origin of a gene swap between two chromosomes that has occurred in hundreds of unrelated families. The rearrangement, or translocation, has been linked to mental retardation and physical anomalies in the offspring of individuals with the defect. A paper describing the model appears in the current issue of Human Molecular Genetics.

Detail from image showing chromosome shift. "Der11" is an abnormal chromosomes 11 that includes DNA from chromosomes 22 (indicated by green).

Since 1980, researchers have identified hundreds of families with examples of the same rearranged genome. In a clean geographic swap known as balanced translocation, a small piece of chromosome 11 switches places with a small piece of chromosome 22. Remarkably, almost no genetic material is lost and the relocated genes work just fine in their new home.

"Year after year we've been looking at this translocation trying to understand the mechanism," says Beverly S. Emanuel, of the Children's Hospital of Philadelphia, who led the research. Her team of researchers focused on a single translocation, but chromosomal abnormalities are relatively common. She predicts that similar mechanisms are likely to be involved in other translocations.

The first translocation linked to DNA patterns

A carrier of the balanced translocation is perfectly healthy, but the flaw can reveal itself and cause problems during the individual's reproductive years. Translocations are associated with multiple miscarriages, and the offspring of a carrier can inherit an extra chromosome derived from the translocation. As in Down syndrome, the extra copies of certain genes can cause distinct physical features and mental retardation in a child.

Previous DNA analyses of this translocation have shown that the breakpoints and rearrangements appear to be quite similar in most cases. Another interesting aspect of the translocation is its spontaneous occurrence—it turns up in individuals whose parents are unaffected, and genetic studies have shown that the abnormality is not due to a single shared ancestor.

How a child inherits an extra chromosome from a parent with a balanced translocation.

Everything about this translocation—its spontaneous nature, relative frequency, and similarity of breakpoints—led Emanuel to theorize that the human genome contains inherent structural flaws. Studies in other organisms have linked chromosomal shifts to repetitive sequences of genetic code, which are often unstable. These stretches of DNA chemical letters can lose their classic double-helix shape and become prone to breakage.

To test the genomic instability theory, Emanuel teamed up with Bruce Roe, head of the Advanced Center for Genome Technology at the University of Oklahoma. Roe's lab was part of the international collaboration that sequenced human chromosome 22 and published the result in December. Using DNA samples from carriers of this translocation and their offspring, the Oklahoma group began looking at the breakpoints.

Computers predict shifts in the genome

The breakpoint on chromosome 22, it turns out, lies within one of the "gaps": a region of chromosomal DNA that is extremely difficult to sequence. Repeated failed attempts to characterize the region forced the researchers to take a backdoor approach. First they characterized the breakpoint region on chromosome 11; then they went back to the DNA of patients that had a combination of genetic material from both chromosomes and compared it with DNA from chromosome 11 alone.

The comparison produced enough of the missing sequence from chromosome 22 to yield an interesting result: The breakpoint on each chromosome appears to be surrounded by essentially the same palindromic DNA sequence. Each palindromic sequence consists largely of two chemical letters, A and T.

A computer analysis of the DNA sequence flanking the breakpoints predicts the formation of abnormal structures such as hairpins. "We think that points at the base of the hairpins are very sensitive to breaks," says Emanuel. The similar DNA sequences on chromosomes 11 and 22 make them likely to recombine if breakage occurs.

Improvising to fill sequence gaps

"This result indicates that there are regions of the genome that seem to be more prone to these rearrangements," says Roe, who has conducted DNA sequencing for other studies involving disease-related translocations. This is the first example, he says, of chromosomal shifts associated with a particular region or pattern in the genome.

"I'm convinced we have these translocations occurring all the time," says Roe. "But the only ones we study are those that lead to severe characteristics or direct alterations of the properties of a cell." Given that genomes might be full of translocations, he says, "It was very exciting to see that you can predict one with a computer."

. . .

Kurahashi, H. et al. Regions of genomic instability on 22q11 and 11q23 as the etiology for the recurrent constitutional t(11;22). Hum Mol Genet 9, 1665-1670 (July 2000).

Back to GNN Home Page