Geneticists can now pinpoint a single gene in a mouse, knock out that gene, and figure out what it did by seeing what the mouse can’t do.
Fifteen years ago molecular geneticist Mario Capecchi had an idea he thought was great but others thought was impossible. People were very skeptical, he recalls. In fact, our first grant proposal was refused, the comment being that it just wouldn’t work. But we carried on anyway. It was a gamble. Capecchi’s impossible dream was to find a way to disable a single gene in a mouse--just one of a hundred thousand--and see what happened.
Today that dream is a reality, and knockout mice are the hottest rodents in genetics. More than 100 different strains are squeaking about in laboratory cages around the world; Capecchi himself has 30 and counting in his University of Utah lab. It takes a year’s hard labor to breed each mouse, but to geneticists the time is well spent. That’s because the knockout approach gives them the hope of figuring out how 100,000 mouse genes add up to a mouse--and in broad strokes, to a human, since humans and mice share about 90 percent of the same genes.
By destroying genes one at a time, geneticists can deduce what each gene does by observing what the mouse can no longer do. Capecchi’s team, for instance, is showing how each of a class of genes called homeobox genes plays a critical role in embryonic development: different ones seem to be responsible for the proper growth of different body parts, and when any one is knocked out, the mouse is likely to die at birth or soon after. Other researchers are focusing on the genetic basis of behavior. When two groups of neurobiologists recently knocked out genes encoding enzymes called kinases, they found that the mice--otherwise apparently normal-- couldn’t pass a standard lab-mouse test: forced to swim around in murky water, the animals couldn’t find the platform that allowed them to escape. The knocked-out enzymes are apparently essential for that type of learning.
Knockout mice also make great animal models for human diseases. If a particular genetic defect is known to cause a disease in humans, and if the same gene is knocked out in a mouse, the mechanism of the disease and the potential effectiveness of drugs and gene therapies can all be studied in the mouse model. Last year, for instance, three research groups reported knocking out the gene that, when damaged, causes cystic fibrosis. The knockout mice displayed some of the symptoms of the human knockout: clogged lungs, digestive disorders, and early death.
Knockout technology depends on the tendency of genes to recognize similar genes in a cell and to trade pieces of DNA back and forth. To make a knockout mouse, you simply pick a strand of DNA that contains your favorite mouse gene (say, the one you’ve devoted the last ten years of your life to studying and which you’re sure must be superimportant). In the test tube, you rip the middle out of the gene using DNA-cutting enzymes. Then you stick the two outer bits of the DNA strand back together again and-- with the help of a jolt of electric current--drive this sandwich-minus- filling into a mouse cell in a dish.
The gutted gene makes its way into the cell nucleus, where the mouse’s own set of genetic blueprints is stored as DNA double helices. If the experiment is done in thousands of cells, chances are that in at least one, the gutted gene will find itself alongside the intact version of the gene. When that happens, the bad gene can bind to the good gene and replace it (by a mechanism that is not really understood). Then, as the cell divides, the bad gene gets passed on to the daughter cells.
Geneticists have a neat trick for telling which of the thousands of cells they started with end up having a gene knocked out: they insert another gene, one that confers resistance to some drug, in the middle of the bad replacement gene. When all the cell cultures are exposed to the drug, only the cells with the knockout gene survive. These knockout cells are then poked into a mouse embryo. A mouse embryo is what they were extracted from in the first place, at a young enough age for them still to have the potential to become any kind of mouse cell at all. and that’s what they do once injected back into an embryo: they grow into various bits of mouse.
A second neat trick allows researchers to make sure the knockout cells have taken: the cells and the embryo they’re injected into come from mice with different fur colors. So the fully formed mouse, which contains a mix of knockout and normal cells, has splotchy fur, like a calico cat. Most important, it has sperm cells or egg cells that contain the knockout gene and will pass it on to the mouse’s offspring. After a couple of generations of old-fashioned selective breeding, researchers can produce a mouse that has solid-color fur and that has the knockout gene--and no intact copies of the normal gene--in all its cells.
All this takes about a year. (Developing all these tricks took Capecchi and various other researchers more than a decade.) During that time the researchers performing the experiment are on tenterhooks, waiting to see what their gene destruction has accomplished.
In some recent experiments, however, the surprise has been what didn’t happen when a gene was knocked out. For instance, mice deprived of a gene called myoD, deemed crucial for muscle formation, turned out just fine in every way. Similarly, the deletion of a gene called p53, defects in which have been implicated in most cancers, was expected to be lethal before birth. But the mice were born healthy enough (although they later had higher cancer rates than normal).
Such surprises suggest to some researchers that the body doesn’t rely on one lone gene for each task in the cell--that it apparently keeps several around in case one goes bad. It’s also possible, though, that each gene really is important, and that the negative results are just cases in which researchers haven’t looked in the right place for the gene’s function. There are many functions a gene could have--enabling the animal to learn fast, say, or to tolerate extreme heat or cold--that would be crucial for survival in the wild but hard to pick out in the plush, pampered life of the lab. I’d be surprised if a gene is maintained over millions of years without having a function, says Capecchi. Look hard enough at these mice, and you’ll always find something wrong.