Ever since life on earth began, a serious "oops" factor has been atwork in the endless unwinding and replicating of the DNA double-helix. From microbe to human, down through the ages, everyindividual being's genes have been reproducing after their kind _and making genomic mistakes along the way.

As the four base pairs, the building blocks of life, line up to swingtheir partners along the DNA strands, if one of the four loses itsfooting, the resulting slippage may leave purines or pyrimidines onone strand bereft of opposite numbers on the other. Result: Amismatch _ a loop of unpaired bases, bellying out from the DNAstrand, like an aneurysm.

To catch and correct such errors as they happen, swarms of DNArepair genes patrol each replicating double-helix. Their encodedprotein products replace the missing bases, or excise redundant ones,to make the strand whole again.

Such repair genes evidently go back to the dawn of life. Moleculargeneticists have found them in prokaryotes, notably, Escherichiacoli, in eukaryotic yeast cells, and in Homo sapiens. But even thesequality-control inspectors can acquire errors in their own genesequences.

Such mutated DNA repair genes have recently been found insuspicious intimacy with the genomes of cancer cells, namely,hereditary nonpolyposis colorectal carcinoma (familial colon cancer)and in the tumors of a variety of sporadic cancers.

`Junk DNA' Gets More Attention

This association calls for a close look at another genomic aberration_ microsatellite instability. Microsatellites are the apparentlyuseless short stretches of DNA that repeat over and over again, forwhich reason they used to be dismissed as "junk DNA." Now, therepeaters are taken more seriously.

The scientists who discovered the familial colon cancer gene, foundthat the microsatellites in the tumors varied in length from the "junk"in a patients' own healthy cells. This meant that the repeatingsequences must have gained or lost bases when their tumors wereforming.

As a commentary in the current Science (Nov. 4) notes: "That wasintriguing, because microsatellite instability is one of the defectsmismatch repair is supposed to prevent."

On this score, molecular biologist Thomas Kunkel told BioWorldToday, "The whole explosion of information on microsatelliteinstability is the first compelling evidence that there's some validityto the hypothesis that cancer is a mutated phenotype."

Kunkel heads a laboratory that studies DNA replication fidelity atthe National Institute of Environmental Health Sciences in TrianglePark, N.C. He is senior author on a paper in this same currentScience titled "DNA Loop Repair by Human Cell Extracts."

His long-range goal, Kunkel said, "if one is speaking exclusivelyabout cancer, is to look for other ways to get other biochemicaldefects associated with cancer cells, which could be responsible forthe multiple mutations required for full-blown tumorigenesis."

He mused over the fact that E. coli's DNA repair service "canrecognize and fix mismatches, but can't handle correction of a four-base mismatch loop" on one strand of a double helix. What abouthuman cells? "We wondered whether they might be different from E.coli, because of the tremendous richness of repetitive sequenceinformation in the human genome."

He explained that "Eukaryotes have much more opportunity to createstrands containing loops than do prokaryotes," and observed, "That'sactually a little bit of a new way to think. The knee-jerk reaction tothat would be: `What's good enough for E. coli should be goodenough for human cells.' We wondered whether human cells coulddo it, and lo and behold, they can, which is what our paper inScience is all about."

Kunkel and his co-workers had a colon-cancer cell line (sporadic,not familial) containing the mutated human genomic equivalent of anE. coli DNA repair gene. "We were expecting that this cell extractwouldn't do loop repair, but it did, as we reported in Science, andthat's a surprise. It suggests that the protein requirements formismatch repair and loop repair are different."

He surmised, "If that's the case, we may have discovered a newrepair pathway. It's my generic belief that if the cell has taken thetime to evolve such a repair pathway, then it's probably there for avery good reason. So it's likely that perturbations in that pathwaywould have biologic consequences, associated with some kind ofhuman misery."

Cautiously envisioning a possible clinical application, Kunkelsuggested that "Certainly, if we find a genetic defect in loop repair,closely associated with a human disease, it would be useful fordiagnostic purposes in the same way that that defective DNAmismatch repair gene is going to be used for colon cancer, orBRCA1 for familial breast cancer." But he emphasized, "We've notmade that association yet."

Kunkel raised the question of a person who inherits an inactivatedDNA mismatch repair gene from one parent: "Is the second genethereby at increased risk of being inactivated by mutations resultingfrom environmental insult, by virtue of being partially compromisedin repair capacity?" This, he concluded, "remains an open question,one of extreme interest to this Institute of Environmental Health." n

-- David N. Leff Science Editor

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