Duplicated Gene May Explain Rare Cancer in Some Families
For more than a decade, researchers at NCI and their collaborators have collected DNA from seven families with an inherited form of chordoma, a type of bone cancer. The researchers have now identified a genetic change that may lead to the disease in four of the families.
The change was not a mutation in the DNA sequence of a gene, as the researchers had expected, but rather the duplication of an entire gene. In each family, an extra copy of the gene was created by a large structural change in the genome—a rearrangement—that occurred once and was passed down to future generations.
While the type of the genetic change was perhaps unexpected, the identity of the gene was not. As the researchers reported online in Nature Genetics on October 4, the gene is active in the chordoma tumor cells of people with the noninherited form of the disease. The gene is called T, or brachyury (for the protein it encodes).
“We have found the gene for a subset of families with an inherited predisposition to chordoma,” said lead investigator Dr. Rose Yang of NCI’s Division of Cancer Epidemiology and Genetics (DCEG). “We had thought this gene may be important in predisposing to the disease, and for the first time we have direct evidence from a study of high-risk families.”
About 300 cases of the noninherited form of chordoma are diagnosed in the United States each year. The disease causes a tumor to develop anywhere along the spinal column, from the base of the skull to the tailbone, and the tumor can be fatal. Researchers believe the tumors arise from remnants of the notochord, an embryonic precursor to the spinal cord.
The discovery provides a new clue to the biology of chordoma and could have broader implications. It suggests an alternate strategy for finding susceptibility genes when traditional methods fail, the researchers said. In this case, genetic mapping studies had linked a region of chromosome 6 to the disease, but sequencing DNA in the region, including the T gene, revealed no suspicious changes.
“Sequencing DNA doesn’t find everything, and it won’t tell you if a whole gene is deleted or duplicated,” said co-author Dr. David Ng of the National Human Genome Research Institute (NHGRI). “So if a gene appears to play a role in a disease but you don’t find any mutations by sequencing, looking for genomic rearrangements, including duplications or deletions in the gene, is a reasonable next step.”
To survey the structural changes in the genome, the researchers used a technique called array comparative genomic hybridization (CGH), which revealed a different genomic rearrangement in each of the four families. Yet nearly all of the affected family members had tumors in the same area (the base of the skull).
Most human DNA is present in two copies per cell (one from each parent), yet in each cell, certain regions are deleted or present in multiple copies—what are known as copy number variations, or CNVs. These changes are usually harmless, but certain types of rearrangements can lead to cancer and other diseases.
In the largest survey to date of frequent CNVs in the human genome, reported this month in Nature, researchers said that any two human genomes differ by more than 1,000 CNVs (about 0.8 percent of a person’s DNA sequence). The study authors predicted that genomic rearrangements are likely to play a role in rare as well as common diseases.
The chordoma study certainly supports this idea. It also underscores just how important advances in technology and persistence are in the search for susceptibility genes.
“We almost gave up hope of finding the gene on several occasions,” said Dr. Dilys Parry of DCEG, who launched the project in 1996 after identifying a chordoma family that now has 10 affected members spanning several generations. “But in the last 13 years, the technology has changed enormously, and we’ve been able to take advantage of the advances at every step along the way.”
This was perhaps most evident 2 years ago, when the team used new markers and tools to locate the region of chromosome 6 with the T gene. After negative results from DNA sequencing, the researchers used the array-CGH technology to screen 13 members of the 7 families (including 11 with chordoma) for genomic rearrangements.
The result made sense immediately to the researchers because the T gene is active in some chordoma tumor cells and it regulates the development of the notochord. However, they do not yet understand how the gene might cause chordoma, according to co-author Dr. Michael Kelley of Duke University Medical Center and a former NCI investigator.
Another question is what caused the disease in the three families without the T gene duplication. The researchers predict that mutations in other genes or an as-yet-unknown mechanism involving the T gene will eventually provide the answer. They are actively enrolling chordoma families to participate in an ongoing study to identify the other genes.
In some rare inherited cancers, there are not enough families—or the families are too small—to use traditional linkage studies to locate chromosome regions of interest. But in some of these cases, noted Dr. Yang, testing for genomic rearrangements may be informative.
“This study really highlights what we can do with new genomic technologies,” said Dr. Joan Bailey-Wilson, who studies the genetics of inherited diseases at NHGRI and was not involved in the research. “CNVs are another item in our tool box for locating genetic susceptibility variants, and most of us are excited about any new tool we can get.”
The discovery of a susceptibility gene always leads to important questions about biological processes that can take many years to answer, Dr. Bailey-Wilson added. “Once we find a gene that we believe is responsible for a linkage signal, we have to realize that the true work is only starting,” she said.
—Edward R. Winstead