By Monica Heger
Combining the old with the new, two separate research groups used linkage analysis and high-throughput targeted sequencing to identify the disease gene in the second most common form of human microcephaly, a rare, genetic, neurodevelopmental disease that causes a small brain.
The two groups, one from the University of Cambridge and the other from Children's Hospital Boston, each used slightly different strategies, but came to the same conclusion, in the studies published this week in Nature Genetics.
The University of Cambridge group had been searching for the causative gene in the disease for about 10 years, Geoff Woods, human geneticist at the University of Cambridge and senior author of the study, told In Sequence. Using linkage analysis, "we had the region narrowed down to about 3 megabases, but hadn't been able to find the gene."
Prior to the advent of next-generation sequencing technologies, the group had sequenced one gene at a time. "We thought we had sequenced everything that could possibly be a candidate gene," Woods said. "We thought we had either missed it, or it was an unusual mechanistic pathway, or it was a gene that was completely undefined."
Woods noted that it was difficult to decide whether to make use of the new sequencing technology and sequence the entire region because, at the time, it was still early days for capture technology and they weren't sure if they would get good results.
The team made use of the Eastern Sequence and Informatics Hub, located within the University, which used NimbleGen's capture microarray and Roche's 454 GS FLX to sequence the region in two affected individuals from two well-characterized, multi-affected, consanguineous families.
Only one gene in both individuals contained potentially pathogenic mutations — the WDR62 gene. One individual contained an insertion, causing a frameshift, while the other individual had a point mutation that affected a highly conserved amino acid.
Screening the gene in five additional consanguineous families identified four additional homozygous mutations in WRD62: three missense mutations and one protein-truncating mutation.
None of the mutations were present in genomic databases or in ethnically matched controls, and segregation analysis predicted the mutations as both recessive and pathogenic.
The team is now trying to further characterize the gene and its function. Woods said the gene appears to be involved in forming mitotic spindle proteins, which are important for neurogenesis, and in particular for creating neural precursors, which determine how big the human brain will be.
The results could have important implications both for understanding the biology of neurogenesis, and also for prenatal diagnosis, and in assessing a couple's risk of passing the disease on to their children.
Woods said that while the combination of linkage analysis and targeted sequencing proved to be useful, they have since moved on to primarily using whole-exome sequencing. "It's more tried and tested, and also cheaper," he said. They are continuing to use the Cambridge sequencing hub for their sequencing needs, and entrust the scientists there to use whichever sequencing platform they think is the most suitable. "We've decided to outsource the sequencing to experts," Woods said. "We say to the hub, 'Sequence it in the most reliable method.'"
Woods' team studies diseases that affect human brain size and also genetic disorders of pain. Currently, their samples are being sequenced on Life Technologies' SOLiD machine, and aside from the 454 they have also used the Illumina Genome Analyzer. The hub has all three platforms, said Woods, and its researchers are trying to work out which platform is best for which situation.
Woods said that advances in sequencing technology have enabled his team to spend less time identifying causative genes and instead "leap on to figuring out what those genes do."
The fact that the Children's Hospital Boston group independently came to the same conclusion provides even more evidence that the correct gene has been identified and also demonstrates the power of the sequencing technologies used, said Woods.
"Both of us were interested in diseases of neurogenesis," he said. "This is a new gene involved in neurogenesis, and we're now trying to figure out how it does what it does."
The Children's Hospital group also used linkage analysis between two affected families, one from Mexico and one from Turkey, identifying a 7.5-megabase region homozygous in the affected individuals.
The team used an array capture approach to target 3.5 megabases of the region containing 148 genes, and paired-end sequencing on the Illumina GA in two affected individuals, one from each family.
They found over 2,000 high-quality variants, including single-nucleotide variants, microinsertions, and microdeletions from each individual. They then filtered out variants present in dbSNP or the 1,000 Genomes Project, and used the Genomic Mutation Consequence Calculator to predict pathogenicity. Only five variants from one individual and seven from the other were not in either database and potentially pathogenic. Furthermore, only a single gene contained new, potentially pathogenic variants: WDR62.
The team then used Sanger sequencing to confirm mutations in the WDR62 gene in the other affected family members, and also identified four additional alleles in four other families.
In the family from Mexico, all three affected members had a homozygous single-base-pair deletion, creating a frameshift mutation. The sequenced individual from the Turkish family contained a single-base-pair deletion, also causing a frameshift mutation. And the other mutations included two missense mutations in highly conserved regions, a 4-base-pair deletion that removed a highly conserved region, and a 17-base-pair deletion. All the mutations were homozygous and found in the WDR62 gene.
The Boston group also did an immunocytochemistry analysis to study WDR62's expression in mouse cells, revealing widespread expression of the gene in the developing brain, with the highest expression occurring in the forebrain. Expression of the gene also followed a similar pattern as other proteins known to be involved in microcephaly. Additionally, a postmortem analysis of a 27-week old fetus affected by the disease revealed a much smaller than normal brain, lending further evidence to the role of WDR62 in neuronal proliferation.