Gene Tied to Heightened Diabetes Risk.

People with certain common variations of a newly identified gene called CAPN10 face a sharply increased risk of getting adult-onset, or type II, diabetes, research now suggests. If further studies confirm that these variants contribute to the disease, the finding could have landmark implications for diabetes diagnosis and prevention, scientists say.

The gene, which sits on chromosome 2, encodes an enzyme called calpain-10. Calpains are proteases--proteins that cleave other proteins. Some proteases have well-defined roles, but scientists admit they know little about calpains.

The gene for calpain-10 can come with dozens of variations, or polymorphisms. In the October NATURE GENETICS, scientists report that having a particular form of CAPN10 tripled the diabetes risk in a group of Mexican-Americans. The most troublesome version of the gene contains three specific variations, or single nucleotide polymorphisms (SNPs), within its DNA sequence.

The three variations don't appear to change calpain-10 itself, but they may cause its supply to fluctuate, says study coauthor Nancy J. Cox, a geneticist at the University of Chicago. This, in turn, seems to influence susceptibility to type II diabetes, she says.

Chromosome 2 drew attention 4 years ago in a study of 346 pairs of Mexican-American siblings with diabetes. Because none of the siblings were identical twins, each pair should have shared roughly half their genes. So, geneticist Craig L. Hanis and his team at the University of Texas Health Science Center in Houston looked at 12 spots on various chromosomes where the siblings were inordinately similar, suspecting these might contribute to diabetes. The most likely culprit, the researchers concluded, is a region on chromosome 2.

Hanis, Cox, and her Chicago colleague Graeme I. Bell have now identified the gene CAPN10 within that region. To scrutinize CAPN10, the researchers compared gene variations in 110 diabetic Mexican-Americans with those in 112 people randomly chosen from the same ethnic group. When a person had a certain version of SNP 43, diabetes risk rose. A combination of that SNP and two others tripled the normal risk. Having these three variations could account statistically for 14 percent of type II diabetes among Mexican-Americans, the scientists estimate.

To test their theory in other populations, the researchers examined CAPN10 in 191 Finns with diabetes and 192 without it, as well as in 220 Germans with the condition and 90 healthy Germans. Combined, these data also yielded a three-fold risk associated with the gene version having the three problem SNPs.

The researchers "are treading new ground here," says geneticist Alan R. Shuldiner of the University of Maryland School of Medicine in Baltimore. "It's a purely statistical argument, but it's a tour de force."

To discern a role for SNP 43, Cox and Bell collaborated with molecular biologist Leslie J. Baier of the National Institute of Diabetes and Digestive and Kidney Disorders (NIDDK) in Bethesda, Md. The team worked in Arizona with Pima Indians, who have a high rate of diabetes.

Participants with the troublesome version of SNP 43 were missing up to half of the CAPN10 messenger RNA in their muscle tissues, the researchers report in the October JOURNAL OF CLINICAL INVESTIGATION. This indicates a shortage of the calpain-10 protein.

Reaching this point in understanding CAPN10 "took a tremendous amount of gene sequencing," says Allen M. Spiegel, director of NIDDK. Recognizing variations in the gene may provide physicians with "an opportunity to better identify those people at risk [of type II diabetes] and to target prevention," he says. Recommended measures might include exercise and a lean diet.

Many questions remain. The polymorphisms occur in DNA that lies between the protein-encoding sequences of CAPN10. Any changes these SNPs cause are likely to show up in the quantity, not quality, of calpain-10, Shuldiner says.

That makes it difficult to prove that these SNPs cause diabetes. "It's a lot easier to convince yourself ... if the [variations] are clearly functional, such as changing an amino acid," says Michael P. Stern, an epidemiologist at the University of Texas Health Science Center in San Antonio. Researchers still "have to show how [these variations] lead to biochemical changes that ultimately result in diabetes," he says.

Meanwhile, scientists suspect that SNPs in another gene, PPAR-gamma, may also hike diabetes susceptibility (SN: 9/9/00, p. 167). Some interplay between this gene and CAPN10 is possible. "I expect there would be dozens of diabetes-susceptibility genes [and that] specific combinations of these genes will define risk," Shuldiner says.