Two rival groups of researchers are reportingtoday and Friday that they have isolated hMLH1,the "human mutation L homolog 1" gene. ThisDNA sequence is suspected of implication insome 30 percent of inherited colon cancer casesa type that comes on without priorproliferation of non-cancerous polyps.Human non-polyposis colon cancer (HNPCC)affects an estimated one in 200 Americans, andis one of the most common genetic diseases inhumans. Experts estimate that one-sixth of the156,000 colon cancers diagnosed each year areHNPCC.This discovery of genes that drive hereditarycolon cancer marks the first clinical success in J.Craig Venter's controversial drive to stockpilepartial gene sequences for future patenting.Venter is one of 20 co-authors from four U.S.research centers, plus one in Europe, whosepaper, "Mutation of a mutL Homolog inHereditary Colon Cancer," will appear in Scienceon Friday. Its principal authors are BertVogelstein and Kenneth Kinzler of the JohnsHopkins Oncology Center in Baltimore.Today's issue of Nature publishes a strikinglysimilar report, titled "Mutation in the DNAmismatch repair gene homologue hMLH1 isassociated with hereditary non-polyposis coloncancer." Its 18 co-authors represent four otherU.S. centers in Europe. That team's leaders areR. Michael Liskay of Oregon Health SciencesUniversity and Richard Kolodner of Harvard'sDana Farber Cancer Institute.When Science learned of the impending Naturearticle, it moved up the media embargo on itsown story by one day.Dead Heats in the Gene RaceLast December, the same two teamssimultaneously announced a gene calledhMSH2, allegedly responsible for 60 percent ofHNPCC cases. (See BioWorld, Dec. 6, 1993, p.1.) Both groups independently located hMSH2on region 21 of chromosome 3.Combining the two gene percentages, 30 and60, suggests that the large majority of suchcancers will in the near future be predictable anddiagnosable. In fact, Kinzler of Dana Farberforesees that "within the next 12 months adiagnostic blood test to screen for defects inthese genes should be available to families athigh risk and allow for early intervention."These genes act as quality assurance andcontrol inspectors. As a cell divides, they patrolthe replication fidelity of its DNA and encodeenzymes that correct errors in the matching upalong the evolving new double-helix of thenucleic acid's four basic building blocks A, G,C and T (adenine, guanine, cytosine, thymine). Ifthis mismatch repair process skips a beat, as inHNPCC, the progeny cells inherit one parentalgene copy in defective form while the secondone is lost, say, to exposure or environmentalcarcinogens. This double loss means the cellscan't copy their DNA accurately, so geneticmutations accumulate and lead quickly to themalignancy.The researchers had identified such a mismatchrepair gene, mutL, in bacteria and yeast cells.Because long stretches of DNA were identical inboth the prokaryotic and eukaryotic organisms,they surmised that this homology could carryover into human DNA repair genes.Turning to TIGRPursuing this presumption, Hopkins' Vogelsteinand Kinzler last December got in touch with TheInstitute for Genomic Research (TIGR). This isVenter's non-profit undertaking for large-scalesequencing and sorting of gene fragments.There, Venter had developed a method forrapidly identifying human genes.Instead of isolating a genomic DNA genesequence, then seeking its function, TIGR startswith random bits of messenger RNA, from whichthey sequence the complementary DNA, whichis an exact gene fragment or expressedsequence tag (EST).Venter searched his supercomputer data base,now numbering upward of 100,000 cDNAfragments, and in minutes came up with threecandidate gene sequences. One of them,hMLH1, mapped to human chromosome 3'sregion 21, where studies of families with ahistory of HNPCC had suggested that a colon-cancer gene might resideCreighton University in Omaha and theUniversity of Helsinki, Finland, analyzed bloodsamples from 10 HNPCC extended families andfound germline deletion mutations of the hMLH1gene in nine of them. "This work," Venter said,"represents an exciting step in the understandingof cancer and provides a terrific example of thepower of the new approach for human diseasegene discovery."Dana Farber's Kolodner found his humanhomolog of the bacterial mismatch repair gene inLiskay's laboratory. Liskay had parlayed thebacterial MutL gene into a mouse equivalent,which he used to tease out two incompletehuman fragments. Liskay and Kolodner usedthese to probe for the full human copy, whichthey pinpointed in three HNPCC families, andreached the finish line neck and neck withVogelstein's group.Clinical Pay-Off in OffingThis doubly confirmed initial gene discovery willmake it possible to screen family members forpresence of the mutated hMLH1 and measuretheir odds of inheriting the disease. Peoplewithout the mutation would be spareduncertainty, and the cost of frequent discomfitingcheck-ups. Those with the defect could bediagnosed in time to remove most tumors whiletheir disease is still in a curable stage."Genetic diagnosis combined with effectivesurveillance should reduce cancer deaths inthese families by over 90 percent," saidVogelstein. He and others think that theremaining 10 percent of HNPCC-prone peoplemight harbor one of the other two partial genesthat TIGR had turned up, namely hPMS1 andhPMS2.To pin down these other candidate genes,Vogelstein will continue to collaborate with TIGRand William Haseltine, chairman and chiefexecutive officer of Human Genome SciencesInc. (HGS), the commercial arm of Venter'sinstitute. HGS's mission, Haseltine said, is "todevelop and market unique drugs and diagnosticproducts, based on its leadership position in thediscovery and expression of novel genes fromhuman and microbial origin."Haseltine announced that HGS "has filed forpatents to use the genes for diagnosis andtreatment of cancer and other diseases."Science staff writer Robert Service, in acommentary on the Vogelstein paper, reportedthat Francis Collins, who directs the NationalInstitutes for Health National Center for HumanGenome Research, worries that cDNA databases in the hands of private organizations"could keep much of the information out of thepublic domain." Collins added, "I'd be a littlemore relieved if all this work was going to bemade available to the scientific community."Venter told Service that "TIGR's work on a largenumber of sequences will be made public byOctober." HGS obliged Vogelstein to sign anagreement giving his company first rights incommercializing any products developed as aresult of their collaboration.031794 Cancer

-- David N. Leff Science Editor

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