NEW YORK (GenomeWeb News) – A study online in Nature Genetics yesterday is offering new details about the contributions that copy number variants make to a range of conditions characterized by developmental or intellectual delays, providing a resource for those looking for clinically relevant CNVs in children.
"The way I view it is like a roadmap for exome studies," senior author Evan Eichler, a genome sciences researcher at the University of Washington, told GenomeWeb Daily News, explaining that genes and loci identified in the study and others like it should help determine which of the copy number changes and mutations found in exome sequencing studies are most likely to contribute to developmental delay, autism, epilepsy, and related heterogeneous conditions.
The team compared copy number patterns in sample from tens of thousands of individuals with or without developmental delay. In general, they found that copy number changes were enriched in the development delay group, with CNVs larger than 400,000 bases accounting for an estimated 14 percent of the cases.
"We report striking differences in the CNV landscape between cases and controls, highlight potentially pathogenic genes, refine known disease-causing mutations, and develop methods to opportunistically discover smaller disruptive CNVs from clinical datasets," the researchers wrote.
As recently as five years ago, it would have come as a surprise to find that such a large percentage of developmental delay cases were a consequence of copy number changes, Eichler said. And as the resolution of such studies increases, he expects to see even more disease-related CNVs.
"My feeling on this is sporadic duplications and deletions of DNA are responsible for a much greater fraction than even 14 percent," he said.
Even so, the researchers explained, causative CNVs are often unknown due to the rarity of each individual's alteration. And, they noted, the situation is further complicated by the fact that similar copy number changes are sometimes tied to very different phenotypes.
"[A]lthough their disease risk in general is well established, the phenotypic consequences for most large CNVs are not well characterized, nor have these effects been fine mapped," they wrote.
To explore the role of CNVs in developmental delay in more detail, the team did CNV analyses on 15,767 children with a wide range of intellectual disability or developmental delay conditions. Among the children tested, 73 percent had been classified as having an intellectual disability, developmental delay, and/or autism spectrum disorder, while the other cases had cognitive abnormalities or had conditions that weren't annotated.
The cases were collected and tested at Signature Genomic Laboratories in Spokane using customized array comparative gnomic hybridization with nine custom CGH platforms, primarily an Agilent array with 97,000 probes.
Using anonymized raw data for children, the researchers were able to track down CNVs and compare them with a control CNV map that they were building from Illumina microarray data on 8,329 unaffected adult controls. The control group was assessed using several platforms, typically at 550,000 probes or more.
"Because of the increased probe density," the study authors explained, "our control dataset provides increased CNV detection power and resolution when compared to the disease dataset, reducing the potential for spurious CNV enrichments within cases."
The control group was not specifically evaluated for neuropsychiatric traits, Eichler explained, but represents individuals from the general population in the US and UK.
Overall, the researchers tracked down 446,736 rare and common CNVs — an average of almost 54 per person — in the control group. As anticipated, the case group was much more CNV-heavy, especially when it came to large CNVs.
There, researchers saw CNVs of 400,000 bases or larger in nearly 26 percent of individuals tested, with these copy number changes causing an estimated 14.2 percent of the intellectual disability/developmental delay cases. In contrast, 11.5 percent of control individuals carried CNVs of that magnitude.
The increased CNV burden in the cases was noteworthy, but the team wanted to go further in finding regions of the genome where deletions and duplications lead to a development-related phenotype, since a significant percent of individuals in the general population carry large, private CNVs with no obvious phenotype or ties to disease.
"By having these large numbers of 15,000 [cases] versus 8,000 [controls], you can really ascribe some significance, get some sense of penetrance," Eichler said, "and more importantly get some confidence in terms of what is the likelihood that an event you see in these kids is actually pathogenic."
To look at this in more detail, the team focused in on 200,000 base pair windows across the genomes of cases and controls, identifying 80 loci that were enriched for deletions and/or duplications in the children with developmental delay conditions. Of these, 59 CNVs were predicted to be pathogenic, accounting for some 10 percent of the cases.
The search turned up 14 loci that had not been significantly associated with disease in the past and also implicated some loci that were previously dismissed as harmless, Eichler noted, though penetrance at the loci is variable.
"We've applied the basic models that you do in genetics in terms of the case controls to find at least nominal significance for these events," he said.
The researchers also refined associations within the regions known to contain disease-associated CNVs and identified 940 genes with dosage-dependent effects on development related phenotypes. Of these, 615 genes are typically deleted in cases, while 325 are generally duplicated.
Many of the larger CNVs in the cases appear to be de novo changes, researchers reported. And larger CNVs were more common in individuals with more severe developmental problems and additional congenital problems, such as unusual facial or cardiovascular features, than they were in children with autism spectrum disorder, epilepsy or seizure conditions.
"What this is telling us is that there are many organs, many systems that are affected by large CNVs," Eichler said, "but particularly sensitive during development is the development of organ systems, particularly the heart, and, our data suggests, also craniofacial [development]."
The researchers were also able to track down smaller deletions and duplications affecting exons within known genomic hotspots — narrowing in on the minimal regions associated with some phenotypes. For instance, within a chromosome 17 region linked to a microdeletion syndrome, they found a pair of candidate genes, Eichler said, including one that's associated with Alzheimer's disease and dementia.
"This is the kind of power you get if you can combine tens of thousands of patient materials together," he added. "Because what you really need when you're talking about something that has so many different genetic causes, like developmental delay, is huge numbers of samples."
The team is currently doing more research on specific CNVs and their consequences and looking at potential interactions between CNVs. In the future, they hope to look at 50,000 more cases and as many controls to find additional developmental delay and autism-related CNVs. They are also doing exome sequencing studies to look for additional genetic changes behind these conditions — an effort that they believe will be aided by the new CNV map.
"We predict that this map of CNVs and potentially dosage-sensitive genes will be invaluable for both clinical and research purposes in the future," the researchers wrote. "As genomic discovery efforts (especially exome sequencing) expand, the results described here should prove increasingly important to clinicians and researchers faced with the challenges of linking rare disruptive mutations to pediatric diseases."