By Monica Heger
This story was originally published Jan. 28.
Under several grants totaling more than $3.2 million, researchers at Washington University in St. Louis are using the Illumina HiSeq to perform pooled targeted sequencing of DNA from 450 children with acute lymphoblastic leukemia.
Led by Todd Druley, assistant professor of pediatrics and genetics at Wash U's Center for Genome Sciences and Systems Biology, the team is sequencing both germline and tumor DNA from children enrolled in high-risk ALL trials nationwide, with the goal of finding rare variants that increase a child's risk for developing the disease, act as driver mutations, and affect tumorigenesis.
The samples are from the Children's Oncology Group, a national consortium of pediatric treatment centers, and the team is also sequencing samples from an additional 450 healthy infants.
The researchers are using a pooled targeted sequencing approach, based on a method that Druley published in Nature Methods in 2009 to identify rare allelic variants from pooled DNA, and will pool up to about 200 to 250 samples for sequencing. They will use their own bioinformatics software, which is able to call SNPs at a high confidence with 30-fold coverage per allele.
Already, the researchers have sequenced 55 genes from 100 samples, and Druley told In Sequence that they plan to expand the number of genes sequenced to several thousand from each patient.
For the first 100 patients, the team used PCR for the enrichment step. Due to cost and time concerns of doing separate PCR reactions for each gene, they sequenced only 55 genes for the first 100 patients. Going forward, though, he said they planned to use custom-designed Agilent SureSelect arrays, which would enable them to sequence many more genes because the genes can be captured in one array, instead of setting up individual PCR reactions to enrich for each gene.
The 55 genes were chosen based on a genome-wide array study performed at St. Jude's Hospital in Memphis. Druley said that around 13,000 genes were implicated in the array, but they chose 55 that showed the strongest associations and for which there is other published data suggesting their role in pediatric cancer.
He added that targeted sequencing, as opposed to either whole-exome or whole-genome, is preferable for this study because the goal is to find rare variants. As such, sequencing a large number of samples is more important than a lot of sequencing.
"If you're going to find rare variants that are causing deleterious effects you have to sequence a lot of people," Druley said. For the same cost, "I could do exomes on a few dozen, or whole genomes on a handful, but I don't think we would find genes where a bunch of rare variants would pile up," he said.
Sequencing germline DNA is especially important in pediatric cancer because it will allow the researchers to pinpoint inherited rare variants versus acquired or sporadic mutations. Pediatric cancer is unique, said Druley, because typically even those with cancer-causing genes do not develop the disease until they become adults. "The definition of early-onset cancer is before the age of 50," he said. "So, why does a 4-year old get cancer? We don't have any genetic information to point us to why that's happening."
Druley said he expects both the sequencing and analysis to be completed in the next 18 months. Additionally, he is interested in looking at methylation differences between the samples, either by doing a pooled bisulfite treatment of the DNA, or by using a method that Agilent is developing for a methylation-specific hybridization capture array. "But that's still in the works," he said.
Finally, he is also working on adding molecular barcodes to the samples prior to pooled capture to make it easier to match the variants in an individual's germline DNA to those in the resulting tumor. Right now everyone's anonymous, so they are looking at trends across all samples, followed by individually validating subsets of variants with putative functional implications, which is costly and time consuming.
Funding for the project is through several grants from the St. Louis Children's Discovery Institute, the National Institutes of Health, the Alex Lemonade Stand, and the Hope Street Foundation. Each grant lasts between two to five years, for a total of just over $3.2 million.
The project is separate from the sequencing projects at Wash U's Genome Center, including one with St. Jude's Hospital, to sequence whole genomes of 600 tumor/normal samples of pediatric cancers. The Center for Genome Sciences and Systems Biology has its own next-gen sequencers, including two HiSeqs and two Genome Analyzers.
Have topics you'd like to see covered by In Sequence? Contact the editor at mheger [at] genomeweb [.] com.