NEW YORK (GenomeWeb News) – A new paper by Montreal researchers is providing evidence that the gene variants found in some non-cancerous tissues may differ from those present in blood samples from the same individual.
"The usual dogma is that your DNA is the same all over the place," senior author Morris Schweitzer, an endocrinologist and lipidologist with McGill University and the affiliated Lady Davis Institute for Medical Research at Montreal's Jewish General Hospital, told GenomeWeb Daily News. But, he said, his team's work suggests that isn't the case.
The researchers, who were studying a condition called abdominal aortic aneurysm, or AAA, found that SNPs in a gene called BAK1 were different in aortic tissue than in blood samples, even in samples taken from the same individuals. The work appears in this month's issue of the journal Human Mutation. Based on the findings, those involved in the study are urging caution in interpreting genetic associations based on DNA from blood samples alone.
"Traditionally when we have looked for genetic risk factors for, say, heart disease, we have assumed that the blood will tell us what's happening in the tissue," lead author Bruce Gottlieb, a researcher affiliated with McGill and the Lady Davis Institute for Medical Research, said in a statement. "It now seems this is simply not the case."
AAA is a rare cardiovascular disease characterized by a ballooning of the abdominal aorta, which provides blood for the abdomen and much of the lower body. Although there are no obvious symptoms, the disease can lead increase the risk of aortic ruptures, a potentially fatal event. AAA affects roughly five to nine percent of the North American population — particularly men over 60 years old — and is more common in individuals with cardiovascular risk factors such as smoking, hypertension, or high cholesterol.
Because chronic apoptosis has been implicated in AAA, the researchers decided to investigate a gene called BAK1, which codes for an apoptosis-activating protein, in AAA patients. They used Sanger sequencing to sequence BAK1 cDNA from diseased abdominal aortic tissue and matching blood samples from 31 AAA patients.
When they looked at these sequences, the researchers were surprised to find that the BAK1 sequences in the aortic tissues differed from that in the matched blood samples. The aortic tissue carried a version of the gene containing three SNPs that are rare in the blood. In contrast, the matched blood samples contained a version of BAK1 that did not contain any of the three SNPs found in the aortic tissue samples.
On the other hand, when the team sequenced BAK1 cDNA from healthy aortic tissue obtained from a Quebec transplant service, they found the same three SNPs as in the aortic tissue from the AAA cases. The researchers verified their findings by sequencing both strands of DNA and repeating the sequencing several times.
So far, Schweitzer said it's unclear whether these BAK1 differences in the blood and aortic tissue are the consequence of RNA editing, which changes the messenger RNA but not the gene, or DNA editing, which involves differences in the gene itself.
Down the road, he and his team intend to use pyrosequencing to look at BAK1 sequences from healthy and diseased abdominal aortic tissues in more depth — an approach that could provide insights into whether that tissue contains both majority and minority BAK1 sequences. If that's the case, it would bolster the idea that different genetic sequences can arise through selective pressures in different tissue, Schweitzer added, "You may have different tissue selectivity for different DNA phenotypes."
The team is also interested in investigating whether the pattern they detected holds true in BAK1 sequences from genomic DNA and learning more about whether there are differences in the expression or activity of different BAK1 variants.
Based on the evidence so far, Schweitzer believes the BAK1 differences his team detected resulted from developmental rather than somatic DNA alterations. Such a pattern may not hold true for all genes, he said, but the BAK1 story suggests there could be other genes that vary slightly between blood and other tissues.
That, in turn, highlights the need to assess genetic profiles specifically in tissues of interest, Schweitzer argued, though he noted that that is a lofty goal given the fact that it is difficult or ethically impossible to collect some types of tissue from living individuals.
He and his colleagues also suggested that their findings raise questions about GWAS, many of which rely on DNA profiles obtained from blood samples.
"Genome-wide association studies were introduced with enormous hype several years ago, and people expected tremendous breakthroughs," Gottlieb said in a statement. "Unfortunately, the reality of these studies has been very disappointing, and our discovery certainly could explain at least one of the reasons why."
In an e-mail message to GenomeWeb Daily News, Navigenics Co-founder and Chief Science Officer Dietrich Stephan said the team's work is interesting and deserves further investigation.
"Differences between the germ-line genome and somatic cells is well established in cancer. It is also well described that chimeras can result from early DNA changes in early embryonic development that propagate to form regional differences in the genome across the body," Stephan noted. "It is intriguing to think that such mechanisms could result in common phenotypes, and is a topic that warrants deeper study."
Even so, he does not believe the findings are a blow to the results or rationale behind GWAS in general. Researchers have gained "incredible insight" into disease mechanisms using GWAS over the past few years, Stephan said in his message, adding, "It is much more likely that the missing heritability that we are all searching for will be accounted for by rare DNA variants, copy number variants, and heritable epigenetic modifications than by this mechanism."