Scientists from the Institute of Cancer Research in Sutton, UK, have discovered a gene responsible for a fatal form of neonatal diabetes that has devastated at least two families studied in the project. Researchers can now screen potential parents for the gene and can begin developing new treatments for those affected.

Neonatal diabetes is a rare inherited disease characterized by unusually high blood sugar levels within the first six weeks of life. The disease can disappear within weeks following insulin therapy, can go into long remission periods only to reemerge years later, may be a permanent part of the patient's life, or may even cause death within hours or days.

Dr. Richard Houlston's research group at the Institute of Cancer Research in Sutton, UK discovered one of the first genetic links to the disease; the team published their findings in the October 2003 issue of the journal Diabetes (see "Neonatal Diabetes: SNPing it in the Bud"). Houlston's group used the GeneChip® Mapping 10K Array to sift through the 3.1 billion bases in the human genome and find the parts that differed between family members diagnosed with neonatal diabetes mellitus and those who were healthy.

Combining this analysis with a linkage test, the scientists focused in on a locus of chromosome 10p13-12.1 containing 46 hypothetical genes, only about 0.1% of the entire human genome. Of the 46 genes, the group focused on three likely suspects: PIP5K2A, PTF1A, and CACNB2. These genes were chosen because they are expressed in human or mouse pancreatic and cerebellar tissue, and they contribute a function that would imply a possible role in the disease.

One year later, in a study published in Nature Genetics in November 2004, Dr. Houlston's group honed in on the gene responsible. The researchers sequenced all three candidate genes in members of two families stricken with neonatal diabetes. They found two mutations in the PTF1A gene in individuals with the disease. Both of these mutations result in a shortened version of the PTF1A protein that lacks a domain known to be highly important in its function. In both families, these mutations were only found in affected individuals and carriers of the disease—not in any healthy noncarriers.

To verify that this shortened version of PTF1A can actually cause neonatal diabetes, the researchers analyzed mice which produce a similarly mutated protein. They found that mice expressing mutated PTF1A protein died within 2-3 days of birth, had below-average birth weight, and completely lacked pancreatic development. These findings are similar to those people afflicted with neonatal diabetes from the two families in this study. The research group also reported a lack of normal cerebellar development in these mutated mice, raising speculation that PTF1A may play a role in both cerebellar development and the disease process of permanent neonatal diabetes mellitus.

This work demonstrates the value of using the Mapping 10K Array to identify the genetic regions linked to disease. Scientists had previously used microsatellite markers to study the same samples, but failed to identify the gene because they couldn't scan the genome at a high enough of resolution. Using the 10K array, in a relatively short time, these researchers were able to weed through about 30,000 possible genes and focus there studies on three—or about 0.01% of the genes in the human genome. Now, researchers can study the effects of PTF1A mutations in humans, and the role of these mutations in both neural and diabetic disorders.

The findings reported in these two papers pave the way for development of new and better treatments for neonatal diabetes. In addition, suspected carriers of the disease may someday seek genetic counseling to determine the odds of passing this devastating disease onto their children, and genetic prenatal tests might be performed to test for the disease, allowing treatments to begin before a child is even born.