Today, Phase I of the international effort to map and sequence all100,000 or so genes in the human genome reached a double climax.
Two papers in today's Nature mark the milestone back to back. Onebears the title: "A comprehensive genetic map of the human genomebased on 5,264 microsatellites."
An article telling the other half of the story discloses: "Acomprehensive genetic map of the mouse genome."
This final genetic linkage map of Mus musculus, the common labmouse, is authored by molecular biologist Eric Lander, who heads theWhitehead/Massachusetts Institute of Technology (MIT) Center forGenome Research, in Cambridge, Mass.
Lander's murine-mapping paper has 20 co-authors, of which the firstis molecular geneticist William Dietrich, who led the five-yearproject.
His opposite number is French molecular geneticist Gabor Gyapay,who managed the human mapping endeavor at Genethon, a non-profit, Paris-based laboratory, largely financed by France's MuscularDystrophy "telethon" fund-raisers. (See BioWorld Today, June 30,1994, p. 1.) Gyapay's group reported to Jean Weissenbach,Genethon's scientific director, and senior author of the Nature paper.
"Some 25 to 28 persons who worked on this project [but only 12 co-authors] took some five years to construct the human genetic-linkagemap," Gyapay told BioWorld Today.
During that half-decade, he recounted, two earlier papers publishedinterim versions of the map, in Nature in 1991 and 1994.
To explain its importance, he cited the example of fragile-Xsyndrome. "To localize the fragile-X syndrome gene took some 10years. If we re-do this experience now," he said, "with this powerfulgenetic-linkage tool, it would take one-and-a-half to two years, nomore."
Gyapay added, "As a second point, without these human maps wecouldn't even think of localizing the genes responsible for polygenicdiseases, let's say, schizophrenia or insulin-dependent diabetes. Nowwe can think of them."
MIT's Lander, who collaborates closely with Genethon's mappers,explained that "It is crucial to map both the mouse and the humangenomes, because much of human disease research is done inlaboratory models, and the mouse is the best available model."
Even in the murine map's pre-completion version, he added, "it hasenabled the analysis of previously intractable genetic traits, [notably]asthma, hypertension, colon cancer, diabetes and epilepsy."
Mus musculus and Homo sapiens have about the same number ofgenes, but the mouse packages its genome into 20 pairedchromosomes, against 23 for the human.
The final mouse map contains 7,377 loci (genetic marker sites),averaging one every 400,000 base pairs. The human map comprises5,264 markers, more or less 700,000 base pairs apart.
Lander compares these milestone-like markers to bookmarks, spacedthroughout the genome. "Once we discover that a disease gene we'relooking for occurs, say, in the final pages of chapter 10, then itbecomes much easier to search through nearby pages and find thegene."
Lander's paper in Nature notes that "A full description of the markers_ including primer sequences, locus sequence, genotypes in eachcross, and allele sizes in the characterized strains _ would requireover 500 pages of this journal."
The markers laid down in both the murine and human maps consist ofwhat used to be called "junk DNA" _ now, microsatellites. In thesetwo cases, they consist of apparently pointlessly repeating base pairs_ CACACACACA and so on. ("C" stands for cytosine; "A" foradenine. See BioWorld Today , March 8, 1996, p. 2.)
"There are 80,000 to 100,000 CA tandem repeats in every humangenome," Gyapay observed. "They spread more or less uniformly allover the genome, and are highly polymorphic."
That is, individuals can inherit varying lengths or numbers of thestuttering CA sequences from maternal and paternal parents. "Thesepolymorphisms of length occur in every somatic [non-germline]cell," Gyapay went on. "Let's say you inherited a CA repeat at acertain locus, 10 from mother, 15 from father. Because the lengthsare different, you can follow their inheritance pattern from generationto generation, and observe their crossing over. This is the basis of thegenetic map."
That basis, he said, is grounded in statistical probability. "You canjudge the quality of a genetic-linkage map by its density _ howmany markers it localizes. You look at the probability that a givenmarker is placed in a given order, then look at the alternativepossibility. In our map, the difference between the two is odds of onein two thousand, so at the moment, it is an incomparable precisiontool for the localization of genes."
To explore and map these genetic variants in human DNA, Genethonanalyzed stored blood samples from 134 healthy individuals in threegenerations of eight large families.
Five years ago, when he and his co-workers launched this genetic-mapping endeavor, Gyapay recalled, "The number of genes localizedfrom the appearance of our first paper was about 10. Now it's morethan 200."
This increase, he pointed out, adds more than numbers. It opens up anew dimension: "Now, polygenic, multifunctional diseases can belocalized or treated by these methods."
So far, between 1992 and 1996, he added, using this map, "theinternational scientific community has localized 223 genes implicatedin more than 200 diseases. A number of orphan diseases have thusbeen able to emerge from their oblivion. For some of them _ cysticfibrosis, Hurler's syndrome, immune deficiencies _ therapeuticstrategies are being designed, which could not have been imagined 10years ago."
Gyapay concluded, "One of the latest discoveries made using themap was identification of the genetic defect responsible for a seriousneurological disorder, Friedreich's ataxia. (See BioWorld Today,March 8, 1996, p. 2.)
Phase II And Beyond
So much for Phase I of the international genome-mapping program.What will Phase II do?
It will build on Phase I's genetic map "as a preliminary scaffold forconstructing a genome-wide physical map," said the mouse map'sfirst author, William Dietrich.
"Now that you have your disease gene genetically mapped to an areathat would probably correspond to several million base pairs ofDNA," he said, "how do you find the gene then?"
His answer: "Having the physical map, which is integrated with thegenetic map, you can then say, `I know that my disease gene maps tothis part of chromosome 5, say, and I know which yeast artificialchromosomes, or various fragments of DNA that I know about, cloneto that region. And I can look specifically at those specific pieces ofDNA that cause the disease.'"
"Physically mapping these organized sets of isolated and mappedDNA pieces representing all the DNA in the human genome,"predicted Francis Collins, who directs the National Institutes ofHealth's National Center for Human Genome Research, "will likelybe completed in the next two years."
Then on to complete Phase III _ sequencing the 3 billion base pairsof the human genome.
"There is a growing confidence," Collins concluded, "that this mostdifficult of the tasks . . . will also be successfully accomplished _perhaps even a bit ahead of the target date of 2005."
Genethon and MIT are both making their complete genetic-linkagemaps and data available to the scientific community via the Internet.
* The Human map: (http://www.genethon.fr).
* The Mouse map: (http://www-genome.wi.mit.edu).
A reprint of Genethon's map in Nature will also be distributed toparticipants in the Human Genome Organization meeting inHeidelberg, Germany, on March 22, 1996. n
-- David N. Leff Science Editor
(c) 1997 American Health Consultants. All rights reserved.