Molecular biologist Norman Arnheim, Ph.D., and computational biologist Pavel A. Pevzner, Ph.D., both of the University of Southern California, call the new technique ExonPCR. The method allows speedy reconstruction of the way genes are written on human chromosomes, they reported at the Gene Discovery in Silicon Conference.

The new technique uses a series of questions to tease out the presence of genes concealed on chromosomes within a huge volume of  "junk" DNA. It is able to sort through the "junk" and find the useful information spread out over segments called exons. An exon is a segment of DNA that is transcribed into RNA and translated into protein. It specifies the amino acid sequence of a portion of the complete polypeptide. Such exons are fused into a continuous message when the gene is activated, creating an "mRNA" or "cDNA" form of the gene.

New DNA chip techniques have recently made it feasible to zero in on all cDNA from large regions on chromosomes suspected of carrying a disease-causing gene.  This  enables researchers to isolate a group of suspect genes for further testing. However, knowledge of the exon borders from the original chromosomal form of these genes is still required to quickly compare genes from large numbers of people in order to find disease-causing mutant forms.  Researchers first must know how the genetic message actually appears on the chromosomes -- that is, into how many exons the gene is broken up, and where the breaks occur.

This information normally is invisible in the cDNA form, so biologists are forced to do time-consuming chromosomal DNA sequencing to reconstruct the "hidden" boundaries. The cDNA form is too elusive and expensive to detect in mass screenings. This is where the new technique comes into play, explains Dr. Pevzner, a  professor of mathematics and computer science at USC.

The technique resembles the old parlor game of "20 Questions" in which a series of "yes" or "no"  answers is used to narrow the possibilities and solve a riddle, says Pevzner. In the molecular biological form of the game, questions are asked using a technique known as the Polymerase Chain Reaction (PCR).  Every query in the "game" measures the distance between two PCR primers in the genetic message.  However, this distance may differ in cDNA and chromosomal DNA since chromosomal DNA contains junk DNA between PCR primers.

If the chromosomal DNA distance is greater, reasoned Pevzner, it's because  the two "words" are separated by junk DNA on the chromosomes and must therefore be carried on different exons. By repeating the process and starting at different words, the boundaries of the exons can be narrowed down. Pevzner and colleague Sing-Hoi Sze have demonstrated mathematically that, on average, 30 such word queries can reveal an approximate map of the exons in cDNA. Then, another technique, "ligation mediated PCR," can establishx the exact boundaries.

The technique, which the researchers describe as "gene hunting without DNA sequencing," avoids expensive and time-consuming efforts to sequence the entire million-base-pair length of chromosomal DNA, says Arnheim:

"We believe that the technique will greatly speed the identification of mutations, with direct applications for research into genetically based human disease," says Arnheim, who holds USC's George and Louise Kawamoto Chair in Biological Sciences.