Divining origins of diabetes

PHILADELPHIA—In a cross-border genome-wide association (GWA) study involving diabetic children and their families, researchers at Children's Hospital of Philadelphia (CHOP) and Montreal's McGill University have identified a new gene that increases a child's risk for developing type 1 diabetes (T1D)—formerly known as juvenile diabetes. They also confirmed several other gene markers identified in other studies. The results shine a brighter light on what is going wrong in this autoimmune disorder and offer insights to potential diagnostic and therapeutic targets.

As they describe in a paper published in Nature, the researchers performed a two-step GWA study. They first determined the SNP profiles of more than 2,000 healthy and diabetic children (and some parents) using the Illumina BeadChip system, which identified a novel marker that was closely associated with T1D, as well as previously identified markers. Further study showed the new marker resided in the gene KIAA0350. The researchers then confirmed the results with the Sequenom SNPlex platform.

"The KIAA0350 gene presents a highly interesting drug target given the potential role of the gene in natural killer cells, which is the cell type that ultimately destroys the insulin-producing [pancreatic] islet cells in susceptible individuals, since we envision an opportunity here to block the process and prevent T1D from happening," says Dr. Hakon Hakonarson, CHOP researcher and study lead.

The research is just the latest in a growing list of GWA studies that are making significant inroads into our understanding of disease.

"GWA studies do not identify genes per se; they identify markers associated with the disease," says McGill collaborator Dr. Constantin Polychronakos, whose lab is also studying another potential T1D drug target, the insulin locus INS. "These markers attract our attention to that genome region for searching the genetic variations involved in pathophysiology.

"Only after additional fine-mapping and functional studies we will be able to identify the gene functionally responsible; however, if all the markers that show association reside within a gene and the gene is the only gene in that particular linkage-disequilibrium block, as in the case of KIAA0350, it is going to be exceedingly unlikely that the gene is wrong."

GWA studies are also powerful tools for second-party validation of earlier results, according to Dr. Carsten Rosenow, Illumina's senior marketing manager for DNA analysis. "The CHOP studies are good examples of the ability to validate results in other patient cohorts, which was difficult to do in earlier linkage studies," he says (see Clarifying Crohn's).

Aside from being potential targets for drug development, the many T1D-associated markers could also be developed into an effective T1D diagnostic tool, Hakonarson says.

"It is worth noting that the T1D association in a T1D-associated locus can be from the combined effects of different genetic variations, which has been confirmed in the MHC class II locus, PTPN22, and CTLA4," he says. "Therefore, a combination of genetic markers for each locus is needed. In addition, for reliable diagnostics, the remaining unknown loci must be discovered. Our Stage 2, currently underway, will reveal most of the remaining important ones."

Moving forward, Hakonarson says, the researchers are further analyzing KIAA0350, looking for possible genetic variations that cause changes in amino acid sequence and tissue-specific gene expression changes that indicate the gene itself is responsible for T1D. They will also try to establish a direct link between KIAA0350 and T1D by examining the gene's impact in knock-in and knock-out studies in animal models of T1D