By David N. Leff

Fat-finger syndrome is a common complaint in people who haven't mastered touch-typing. Hitting two keys at once on the keyboard will often produce a typographical error, which the word processor's spell-checker instantaneously proofreads, and underlines in red for correction.

This high-tech performance dwindles to triviality compared with the human body's genomic DNA mismatch-repair system. "There are approximately 3 billion base pairs of DNA in the human genome," observed clinical cancer genomicist Steven Lipkin. "And every time a cell replicates, it needs to copy - like a Xerox machine - "those 3 billion base pairs. But in doing so, it sometimes makes mistakes. The DNA repair system's genes then goes into action, and express spell-checker proteins that fix some of those mistakes." (That goes as well for glitches in the repair genes themselves.)

Lipkin is a research fellow in the Genetics and Molecular Biology Branch of National Institutes of Health's National Human Genome Research Institute (NHGRI) in Bethesda, Md.

"Mutations in the six known human mismatch repair genes," he told BioWorld Today, "one of which is called MLH1, have been associated with susceptibility to human cancer - in particular, colorectal cancer. So based on some of the existing human and mouse data, we speculated that there would be an additional mammalian mismatch repair gene that interacts with MLH1.

"Using a strategy," Lipkin went on, "that enabled us to identify interacting DNA-repair proteins, we have cloned a new - seventh - DNA mismatch repair gene, named MLH3. A gene from yeast that looks like this gene had been cloned elsewhere about a year ago. That suggested to us that our MLH3 gene might also be similarly associated with mammalian susceptibility to colorectal cancer." He and his colleagues located their new gene to the long arm of human chromosome 14.

Discovery by NHGRI of a seventh colon-cancer susceptibility gene is reported in the January issue of Nature Genetics under the title: "MLH3: a DNA mismatch repair gene associated with mammalian microsatellite instability." Lipkin is the paper's lead author, and the institute's director, Francis Collins - a pioneer cancer-gene hunter - its senior author.

Spell-Checking Genes Go Back Eons

"The DNA mismatch repair system," Lipkin pointed out, "is very ancient in evolution. It actually goes back to the bacteria. And because it's so primitive, there seems to be a lot of correlation in the way most mammals handle or use the system, between humans and mice. Mice get intestinal cancer. When we looked for where MLH3 was located in the mouse and human genomes, in the mouse genome we found the gene in a particular region associated with mouse susceptibility to intestinal cancer.

"We didn't really see anything in the human DNA at that time," he went on, "but about two months ago, the group up at Roswell Park Cancer Institute in Buffalo, N.Y., did an analysis of colon tumors, to see what genomic regions are gained or lost in human cancers. They found, interestingly, that this MLH3 region of the human genome was also lost in approximately 30 percent of all colorectal cancers."

Lipkin made the point that "this is not a mutation a person carries with him, but arises somatically after you're born. Some sort of a mutational error that happens turns that part of a cell into a tumor. But roughly 60 percent of somatic colorectal cancers occur in individuals who get them before 50 years of age. So 50 is considered the start of normal or sporadic - as opposed to familial or inherited - colorectal cancers. That led the American Gastroenterologic Association to recommend people that age go to their doctor and get a fecal blood test or a flexible sigmoidoscopy, to see if they have the beginnings of colon cancer."

He added, "The MLH3 gene itself, of course, is inherited among the approximately 70,000 genes in the human genome. However, what seems to happen is that after birth, in some people, the cells somewhere in their colons go through a number of changes. When we look at them we say, 'Oh, they've become a colorectal cancer.' If you compared the DNA in their tumor to the DNA they were born with, you would see that one change is loss of this region in chromosome 14.

"In other words," Lipkin continued, "it could be because the cells are losing this gene, or because they are losing some other genes in the region, which are important in the progression to colorectal cancer. The best way for us to test that," he observed, "is to do a knockout mouse. Instead of the bunch of DNA repair genes in the region, we'd knock out only this one MLH3 gene, and see what the effects are. We're working on this right now. The other thing we're doing, of course, is looking for more human mutations."

In this endeavor, one of Lipkin's co-authors is the eminent Henry Lynch, of Creighton University in Omaha, Neb., who has been elucidating colon cancer genetics for the past two decades. "It was Henry," Lipkin recalled, "who clinically defined human non-polyposis colon cancer - known as the Lynch syndrome. And the bottom line is that two-thirds of Lynch syndrome patients have identifiable mutations in DNA mismatch repair genes.

A Lynch Syndrome Test For Ovarian Cancer?

"That could mean that MLH3 is involved," Lipkin went on. "So we're currently looking in families with colorectal cancer susceptibility, and also at individuals younger than 50, who've got the disease, to try to see if we can find MLH3 mutations that are not inherited.

"But Lynch syndrome," he added, "is also associated with other tumors - namely, ovarian, uterine and kidney. A DNA test for the MLH3 gene might be useful in families that have a hereditary component for ovarian cancer, but not yet for colorectal. We've just started to look into that."

He concluded: "Everyone should follow pre-colorectal cancer screening after age 50. And to minimize their risk of that disease, everyone should eat a diet high in fruits and vegetables, folate and calcium - that is, dairy products."