To the hazards of junk food, junk mail, junk bonds and junk ingeneral, now add the cancer risk factor in "junk" DNA.

Last week's New England Journal of Medicine (NEJM) spelledout in its lead article "An Association Between the Risk ofCancer and Mutations in the HRAS1 Minisatellite Locus."

Minisatellites and microsatellites comprise a vast part of thenon-coding junk DNA that litters the human genome withapparently useless repetitive sequences. Molecular biologists,probably in the 1970s, dismissed as garbage the stutteringstrands of the double helix that served no known purpose ingene structure or regulation.

Although scientists are now learning more and more aboutrepetitive DNA, the junk label has stuck, said oncologicalgeneticist Theodore Krontiris because "it's still a black box."

Krontiris, a professor at Tufts University School of Medicine, isthe principal author of the paper in NEJM, which links satelliteDNA to at least four of the main killer cancers: breast,colorectal, bladder and acute leukemia.

"Mutant alleles of the H-RAS-1 minisatellite locus represent amajor risk factor for common types of cancer, including one in11 breast, colorectal and bladder cancers," he said.

His evidence -- comparison of junk DNA mutations in cancerpatients and controls -- is circumstantial. But "it's real," saidcancer geneticist Alfred Knudson, a senior member of Fox ChaseCancer Center in Philadelphia. "I'm not in doubt about it."

Krontiris described his study as "one of the largest molecularanalyses of any human proto-oncogene locus." It concluded thatcertain rarely occurring mutations in the alleles -- genevariants -- in an individual's genome doubled the person'sstatistical likelihood of contracting cancer.

For starters, the Tufts team focused on a gene sequence calledH-RAS-1, which encodes a protein involved in mitogenic signaltransduction and differentiation. H-RAS-1 is tightly linked to aminisatellite about 1,000 bp downstream from its codingsequence. This comprises 30 to 100 units of a 28-base-pairrepeating stretch of DNA, and contains 30 alleles of 1,000 to3,000 bp each.

What this junky segment has to do with the cancer-related rasoncogene, Krontiris told BioWorld, "hasn't yet been established.It's our working hypothesis that a mutation of the minisatelliteinteracts with these proteins in a fashion that in some waydisrupts the ras gene's regulation."

Minisatellites consist of core repeating units 10 to 100 baseslong; microsatellites in their simplest form have only two corebases, repeated dozens of times. Taken together, they are called"tandem repeats" because their bases tend to pair up head totail.

Tandem repeats usually mutate, explained an editorialaccompanying the NEJM paper, by altering the number of theirrepeat units in polymorphic patterns that differ, likefingerprints, from one individual to another. They areincreasingly used of late in polymerase chain reaction-amplified DNA probes to determine paternity and forensicidentity.

Krontiris and his colleagues based their cancer-connectionsurvey on the fact that the highly unstable alleles in tandemrepeats tend to mutate frequently, as often as from parent tochild. Four of these aberrant progenitor sequences are commonand unremarkable, but a few are rare and sinister.

In recent years at medical centers associated with Tufts, thegroup screened leukocyte DNA from 900 white patients,hunting for rare alleles in their H-RAS-1 tandem repeats. Ascontrols, they examined 882 hospital patients with otherdiseases than cancer.

Three times as many of the cancer patients (105) possessedrare H-RAS-1 alleles as did the 53 non-cancer controls. (Theprecise odds ratio was 2.07-1.00.) To double-check this strikingrisk factor, the Tufts investigators analyzed 22 roughly similar,but smaller studies from the literature, totaling 3,730 cancercases and 4,406 controls. After adjusting to maximize statisticalconfidence, the odds ratio came out 1.90 to 1.00, ranging from1.14 to 4.17. Taken together, these 23 studies typed 8,500 H-RAS-1 alleles.

As Krontiris wrote in NEJM, "more than 50,000 cases of cancera year may be attributed to the possession of these alleles."

He calls "unlikely" the interpretation that the rare mutationsare simply markers for the risk of malignancy, but uninvolvedin the actual pathogenesis. Rather, he proposes that "newmutations of the H-RAS-1 minisatellite disrupt the controlledexpression of nearby genes, including H-RAS-1, by interactingdirectly with transcriptional mechanisms."

Krontiris is now attempting to figure out the underlyingmechanism of this disruption. "We're trying to establish therelationship between the minisatellite risk factor, which isabout double that of people who don't have a rare mutation,and other known environmental factors in cancer." The grouphas studies under way to see if there is genetic interactionbetween the minisatellite and the familial breast-cancer geneBRCA-1.

For lung cancer, the odds-on risk was only 1.55 to 1.00. FoxChase's Knudson told BioWorld that "lung cancer could be acontributor, but so drowned out by smoking it might cause younot to notice."

Knudson has closely followed Krontiris' work and finds his 2.07relative risk ratio, though statistically sound, "rather slightepidemiologically."

He pictures the presumed mechanism of the rare alleles' cancereffect this way: "You have some alleles at a particular site thatare unusual in the population; the body's been dealing withthem for a long time. Maybe the regulators can learn to workaround it, so to speak. Now, you introduce a variant, and itdoesn't do so well."

He said that despite the primitive language of this explanation,"it is very much worthy of further pursuit, and with all of theinterest these days in satellite DNA, it's just another one ofthese data that fit into some interesting puzzle."

Knudson pointed out that some repetitive DNA is transposableand can work their way into the middle of a gene and destroyits function. Specifically, Krontiris noted, minisatellites show upin the introns of retinoblastoma, and their genetic damage hasbeen documented in chromosome translocation, factor VIIIdeficiency, possibly Huntington's disease, and other disorders.

On the plus side, some pro bono tandem repeats scatteredthrough the genome organize the centromere regions ofchromosomes.

H-RAS-1 is one of thousands of minisatellites, dozens of whichhave been sighted in or near various genes. Krontirisconcluded, "I think one clear implication of this work is that ifthis minisatellite influences the ras gene in cancer, others mayinfluence other genes, and consequently the genetic aspects ofother diseases."

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.