The extraordinary progress scientists at the Center for Genome Research have made toward creating maps of the chromosomes of mice and humans and a description of even headier work begun March 1 were the focus of a talk Eric S. Lander, director of the CGR and associate professor of biology, gave to the MIT Corporation earlier this month.
In October 1990 the CGR, which is based at MIT and the Whitehead Institute for Biomedical Research and is a consortium of five institutions including MIT and Whitehead, was inaugurated with a grant from the National Center for Human Genome Research (NCHGR). One of the major goals of the CGR was to develop a genetic map of the mouse in five years. The map was completed last year-three years ahead of schedule.So the researchers applied for and received an early renewal of the grant effective March 1. Now the goal is to create a more detailed map for the mouse-and for humans.
FROM $500 to $47 MILLION
"The first couple of years of the project laid the basis of technology and methodology to do this [proceed toward the new goal]," Professor Lander said. In fact, the technology has advanced so quickly that "the estimate made in 1992 of how much money it would take to do [the more detailed maps of mice and humans] was $500 million. We proposed a year later to do it for $52 million," Professor Lander said. (The CGR received $47 million.)Maps of the mouse and human genomes, or all the genes that make up an organism, are a basic scientific need, Professor Lander said. These maps provide the infrastructure scientists need to locate specific genes responsible for disease or of interest for other reasons. Without this infrastructure finding such genes is much more difficult and takes more time. It's also expensive. For example, in 1989 scientists isolated the gene for Cystic Fibrosis. That breakthrough, however, cost $20 million. "There are 4,000 different genetic traits in the human," Professor Lander said, "and we'd like to know the genes for each of these. But certainly MIT can't afford to spend $20 million on each; even the nation can't afford to spend $20 million on each."
BUILDING AN INFRASTRUCTURE
The ultimate goal of the Human Genome Project, then, is to streamline this process by creating three successively more detailed maps of human genetic material. The first, a genetic map, defines the location of DNA "sign posts," or markers, up and down the chromosomes. Geneticists express the DNA that makes up our genes with various combinations of four letters-A, T C, and G, Professor Lander explained. "You own a genome of three billion letters. And the genome you get from mom and the genome you get from dad are virtually identical. They are 99.9 percent identical," Professor Lander said. In fact, he continued, "Any two of us in this room have 99.9 percent identity in our DNA. "But that one tenth of a percent difference means that there are three million little, mostly irrelevant spelling differences. And when we have one of those spelling differences.that's a genetic trait."And those spelling differences serve as markers on a genetic map that can help scientists determine the rough location of a particular gene. The second kind of map, a physical map, will allow scientists to pinpoint the exact location of specific genes. The physical map, when completed, will consist of "test tubes labeled one to N consisting of overlapping pieces of DNA" for all the chromosomes of the organism being mapped. "So if I want to find a gene in a particular region of, say, chromosome 7, I can go down to tube number 672, pull it off the shelf, and know the gene will be in that region," Professor Lander said.The final, and most complicated, map is the sequence of individual nucleotide bases that make up each gene.
MIT MAP OF THE MOUSE
An initial goal of the CGR was to develop a genetic map for the mouse. That was completed last year, three years ahead of schedule. This map, which includes more than a thousand markers, "is now distributed around the world," Professor Lander said. "More than 300 labs are now using what is called the MIT map of the mouse. And 90 percent of the markers on that map have MIT in their name."The MIT map took about a year and a half of painstaking work to create. "How did we get students to want to do this?" Professor Lander asked. "They quickly doped out that if you invest a little bit in the infrastructure, if the first chapter of your thesis is building new tools, you'll be the first to get to apply them to new problems."Case in point: "Once this map was built, William Dietrich, a graduate student in my lab, was able to go off and map a major modifying gene controlling cancer [in the mouse]. It lives on the bottom of chromosome four."Having put a year and a half into building the map, he was able to do that with only five days' effort. That's what having the infrastructure around meant."
MAPPING THE Y CHROMOSOME
In another major advance at the CGR, last year scientists in the laboratory of Associate Professor David C. Page created a complete genetic map of the human Y chromosome. "I'm particularly proud of this because it was, for those of you who think of the Human Genome Project as `big science,' a three-person project," Professor Lander said. "Two postdocs and a technician accomplished that work and did so in 18 months. I think they really set a standard."
With the early renewal of the original NCHGR grant, the CGR is working toward a physical map of both the mouse and human genomes.The project involves scientists from biology, computer science, operations research, and Lincoln Laboratory. Such interdisciplinary ties are crucial to the success of the project, and one reason why MIT was given a grant to start the CGR in the first place."As you can imagine, building the tools to make these maps has much similarity to what we do every day in our [biology] labs," Professor Lander said. "But it also has a certain component of streamlining and of efficiency that is different from what we normally do in molecular biology labs."And so the planners of the Human Genome Project correctly recognized that this would require not just molecular biology, but expertise in automation and engineering, and information processing and computer science."It's no surprise, then, that at least to us, it seemed that MIT was a natural place to do this sort of work. Because to have collaborations, not just within biology or chemistry, but broadly across the whole range of engineering, was something we felt we were better suited to do, for example, than most medical schools."These collaborations should prove useful to the next major goal of the Human Genome Project: sequencing all that genetic material. "Noone in the world knows how to do the sequencing at anything like even $3 billion," Professor Lander said. However, he continued, "I think the next five years will be a revolution in getting those costs down by a factor of 20." To that end, even now the MIT scientists are working with Lincoln Lab "on finding ways to apply Lincoln's micro-scaling abilities to trying to chip off a factor of 20 from the cost of sequencing. The new applications, the new frontiers here are in the sequencing."
Professor Lander concluded: "The overall project is tremendously fun because everybody doing it is young and excited. We've made sure not to tell our people what they can't do, what isn't possible. And they have surprised us again and again by being able to do much more than we thought."The Center for Genome Research consists of a consortium of five institutions. In addition to MIT and the Whitehead Institute, they are Princeton University; the Jackson Laboratory in Bar Harbor, Maine; and the Centre d'Etude de Polymorphisme Humaine in Paris, France.
A version of this article appeared in the March 17, 1993 issue of MIT Tech Talk (Volume 37, Number 26).