By Monica Heger
Despite whole-genome sequencing's falling costs, exome sequencing continues to be a good technique for studying Mendelian disorders, said researchers who for the first time used exome sequencing to identify de novo, dominant mutations in a rare disorder.
The researchers used the approach to identify the causative mutations in Schinzel-Giedion syndrome, a rare disorder that results in severe mental retardation, a high prevalence of tumors, and multiple congenital malformations. Most affected individuals die before the age of ten.
Reporting last month in Nature Genetics, the team used Agilent's SureSelect and Life Technologies' SOLiD to capture and sequence the exomes of four affected infants. In all four cases they found heterozygous mutations in the SETBP1 gene that were not present in the parents, which they confirmed with Sanger sequencing of the SETBP1 gene, suggesting that Schinzel-Giedion is a rare, dominant disease, caused by a de novo mutation. The researchers identified mutations in the same gene in an additional eight cases using Sanger sequencing.
"The mutation is not present in the parents, and it has not been seen in any normal individual anywhere else. So that gives it a very strong link to the disease," Jorvis Veltman, assistant professor in the department of human genetics at Radboud University and senior author of the paper, told In Sequence. "And now, we've seen 12 patients all with a mutation at that particular spot."
Several studies have shown that exome sequencing can be used to identify inherited recessive mutations in Mendelian disorders. For example, researchers from Cold Spring Harbor Laboratory used exome sequencing to identify the causative mutations in Joubert syndrome (IS 2/9/2010), and a University of Washington team used exome sequencing to pinpoint the disease genes in Miller syndrome (which were also independently verified through a whole-genome sequencing study from researchers at the Institute for Systems Biology) (IS 9/29/2009).
The Radboud study, which uncovered de novo, dominant mutations, is "a nice extension of the application of exome sequencing" said Lu Wang, program director of large scale sequencing at the National Human Genome Research Institute, who was not affiliated with the study. "This adds to the now growing list of disease disorders that have been identified by exome sequencing."
The researchers used SureSelect and SOLiD to sequence the exomes of the four patients to an average of 43-fold coverage with 50 base pair reads, generating between 2.7 and 3 gigabases of mappable sequence data per individual. They identified 5,351 non-synonymous SNPs per individual. They focused on 12 genes for which all four patients carried variants, eventually narrowing down their candidate genes to one.
Because Schinzel-Giedion was already thought to be caused by a de novo, dominant mutation, narrowing down the list of candidate genes was slightly different than for inherited, recessive disorders. Instead of looking for homozygous variants where each parent carried an allele, the researchers were looking for heterozygous variants that were also not present in the parents.
On the one hand, searching for heterozygous variants can be slightly trickier, said Veltman, because individuals have many more heterozygous variants than homozygous. But, on the other hand, because the mutation was also thought to be dominant, it was easier to compare to dbSNP and other catalogues of normal variants, he said. With recessive mutations, there is always the possibility that it will be present in normal, healthy individuals, said Veltman, so when you use dbSNP to narrow down candidate variants you could unknowingly exclude the causative mutation. With dominant mutations, that is much less likely to happen.
Veltman said that identifying the causative mutation could potentially enable clinicians to do prenatal testing for the disorder and also could shed light on other diseases with similar phenotypes. For instance, one characteristic of Schinzel-Giedion is a high prevalence of neuroepithelial neoplasia — tumors that affect the sensory and central nervous system. Veltman said they are now studying the gene in this disorder to see if it is involved.
Another interesting finding was that all the mutations in the SETBP1 gene occurred within 12 base pairs of each other, said Veltman. This is unusual, and suggests that the mutations have a gain-of-function effect, changing the function of the protein. Veltman said that if other related disorders are also affected by mutations in that gene, he would expect that those mutations would be more subtle. So, in the case of neuroepithelial neoplasia, he said, they are looking for mutations outside of the 12 base pair segment of the gene.
Aside from studying the effect of the gene in other, more common, syndromes, Veltman said his team is working on sequencing more Mendelian disorders as well as more complex disorders like mental retardation and autism.
Some have argued that the falling price of whole-genome sequencing will soon render exome sequencing obsolete (IS 4/16/2010). However, Veltman said that his team plans to continue using exome sequencing on the SOLiD platform.
"At the moment, [exome sequencing] is a very good approach," said Veltman. "We know how to interpret variations in the coding region much better than outside the coding region," he said. "When whole-genome sequencing becomes more affordable and the interpretation is clear, we will start using that."
NHGRI's Wang agreed that exome sequencing continues to have advantages over whole-genome sequencing, particularly for rare, monogenic disorders. Not only is the cost lower, but data interpretation is much more straightforward.
"Exome sequencing has a lot of advantages for the discovery of rare mutations," said Wang. "There are 7,000 assumed Mendelian disorders. I can see exome sequencing being used for quite awhile."
The technology for exome sequencing is also further along than for whole-genome sequencing, Wang added, including not only the sequencing, but the capture technologies and the data analysis. "The timing is right to apply exome sequencing for the discovery of disease-related genes," she said. "The fact that [Veltman's team] was able to identify de novo dominant mutations is a very nice example of how far exome sequencing can go in terms of identifying rare mutations."