July/August 2007

Guide Dogs

Continued from page 3

By Katherine Bourzac, SM '04

Most of the genome is merely dead space between genes, where the sequence of bases is not important for the functions of life. Mutations that occur in such stretches occasionally produce functional sequences, but they're usually neither helpful nor harmful. That means they accumulate relatively quickly: exerting no influence on an individual's chances of reproducing, they are not subject to natural selection, so they are passed on at a much greater rate than changes in functional areas, which are often detrimental.

"The more evolutionary time there is between two mammalian [species], the more that unimportant things will change," says Lindblad-Toh. But a string of DNA whose sequence is conserved unchanged across species probably has an important function. Once researchers pinpoint those sequences in the areas between genes, they can test them individually to determine their functions.

Comparing the human genome with that of a distantly related organism such as yeast, however, is difficult: to extend the Rosetta stone metaphor, the two texts carry too many different messages. Comparing genomes that carry many of the same messages, like those of the human and the dog, is more productive. "To really understand the human genome, we're focusing on mammals," says Lindblad-Toh. Broad scientists, some of whom participated in the Human Genome Project before the institute was founded in 2003, have been involved in sequencing the chimpanzee, mouse, dog, and horse, among other animals (see "The Broad's Menagerie," p. M17). Work on the guinea pig, elephant, rabbit, little brown bat, bush baby, and ground squirrel is also under way at the Broad.

Researchers initially teased out the functional 5 percent of the human genome by comparing it with that of only one other animal, the mouse. Adding the dog genome made for a powerful triad. Now researchers are focusing on what the functional non-gene elements do. Evidence is growing that they regulate the genome. Without regulatory elements, a gene would be like an unread book gathering dust in the corner of a library storage room: just an inert string of letters.

Knowing what each gene does is very important, but it is not enough. "Some diseases are clearly due to one protein not being there," says Lindblad-Toh; enzyme deficiencies are an example. "But with common diseases like cancer or diabetes, it could just be a function of how much or how little protein you make, and if you make it at the right time or not." It seems likely that those factors are controlled by regulatory elements.

"My strong belief is that a lot of common diseases are caused by regulatory mutations," says Lindblad-Toh. Early Broad Institute research on the dog genome appears to support her hypothesis. Broad researchers are looking for mutations associated with several traits and diseases, including white coat color, thyroid problems, bone cancer, heart problems, and shar-pei fever. For several of these, there is preliminary evidence that the responsible mutations lie outside genes--in regulatory elements.

Zeroing in on where these elements are in the genome is much easier than figuring out what they do and how. One type of regulator, called an enhancer, starts the process that enables genes to pass along their information. Enhancers can be within, near, or very far from the genes they regulate. Researchers hypothesize that proteins bind to an enhancer and to an area immediately preceding a gene; they also bind to each other, pulling the DNA around in a loop and allowing transcription of the gene to begin. But such mechanisms are just beginning to be understood.