Laser gene-mapping yields clues in diabetes.

As difficult as geneticists found it to pinpoint the gene for Huntington's disease and the gene for cystic fibrosis (SN: 7/10/93, p.20), those searches seem simple compared to the quests for the genetic bases of more common disorders, such as diabetes and schizophrenia. Several genes and environmental factors seem to play a role in the development of these illnesses, making it difficult to tease out specific causes.

A refinement of a gene-mapping technique now promises to simplify these quests and has already enhanced the understanding of type I, or insulin-dependent, diabetes. "We have proved for the first time that type I diabetes is a polygenic disease," says John A. Todd of the University of Oxford in England.

Other studies had indicated that a form of a gene responsible for an immune-system molecule called HLA class II and a version of the insulin gene were important in this disease. The new data confirm HLA's key role -- HLA genes account for about 40 percent of an individual's risk of developing type I diabetes. The insulin gene seems no more important than about a dozen other genes, Todd said last week at the Jackson Laboratory in Bar Harbor, Maine, during the Short Course in Medical and Experimental Mammalian Genetics.

If these findings hold up, then testing for this HLA gene may help identify people at high risk of getting this disease, he adds.

To home in on a gene, scientists first identify ever smaller sections of the chromosome on which it resides. For several years, gene mappers have used DNA signposts called microsatellite repeats as markers to delineate these sections.

To streamline the analysis of these markers, Todd's group labeled 300 of them with fluorescent dyes. These markers divide the human genome into sections of 20 million nucleotides. As the labeled DNA sections migrate down a gel, a laser detects each one's color and sizes it more accurately than a person could, Todd explains. Then a computer stores and analyzes the data, comparing an individual's genome to those of other family members, Todd's group reports in the July NATURE GENETICS.

From these comparisons, the researchers compiled a "suspect list" consisting of all sections of DNA that show up in diabetic children more often than expected. When compared to lists generated from 100 families with diabetic members, "you begin to find out which sequences are cropping up all the time," Todd says.

Eighteen suspect sections, not counting those with the insulin and HLA class II genes, appeared on multiple lists. The team is checking the 18 against the genetic makeup of another 100 diabetic families.

The two sections evaluated so far do appear in these families, Todd notes. However, he expects that many of the 18 will prove uninvolved in diabetes and that new suspects will crop up.

Using dyes, a laser, and automated data analysis makes this mapping technique 10 times more efficient than existing ones, Todd says. These refinements make practical whole-genome screens that evaluate all of an individual's chromosomes at once.