Of Mice and Men

In the summer of 1980, Melvin Bosma, an immunologist at the Fox Chase Cancer Center in Philadelphia, began to examine four mice that were brought to his attention because of the peculiar results of their blood tests. The mice had none of the usual antibodies in their blood. In fact, they seemed to have no immune reactions at all.

The four mice turned out to be littermates, which suggested that their lack of immunity might be a genetic trait passed down from mother and father. Bosma soon realized that he and his coresearcher and wife, Gayle Bosma, had discovered an exceedingly rare and useful spontaneous mutation.

Named "scid" for severe combined immunodeficiency, the mutant mice cannot make T cells or B cells, white blood cells that fight off infections and foreign implanted tissues. It is one of the best animal models ever found for studying the basic biology of the immune system as well as several diseases.

The scid mouse "was a gift of nature," Melvin Bosma says. The Bosmas belong to a generation of modern scientists who use various strains of mice to study fundamental life processes. Mice are small, handy, and remarkably fecund: three months after her own birth, a female mouse can produce a dozen new babies. Mice live only two to three years, allowing researchers to follow disease processes from beginning to end in a relatively short time. And their genes are remarkably similar to those of humans, despite an evolutionary distance of 75 million years between the two species.

As a result, researchers have come to realize that they can use mouse genes as a blueprint for finding and studying human genes, including disease genes. They can also use mouse models to test drugs, devise novel therapies, and study the physiology and biochemistry of genetic diseases in ways not possible in humans.

Most of the mutant mice strains under study arose spontaneously, and many of them are housed at the Jackson Laboratory in Bar Harbor, Maine, which maintains over 450 strains of mice with a wide variety of genetic afflictions and apt names. The "stargazer" mutant, for example, has a neurological disorder that forces it to throw back its head and look at the sky. "Twitcher," "shiverer," and "quaking" mice have damaged nerve fibers and abnormal gaits. "Dwarf' and "little" are lacking growth hormones. At least 20 new mutants are identified each year. Using various breeding tricks, the scientists are able to perpetuate the mutant strains in stable colonies.

Some of these mice have versions of common human diseases, such as diabetes, whose two forms - juvenile (Type I) and late-onset (Type II) afflict millions of people worldwide. According to diabetes expert Edward Leiter of the Jackson Laboratory, these mice give researchers the opportunity to tease apart the genetic and environmental factors that underlie the disease. When fed high-fat diets, for example, some strains of mice become glucose-intolerant (as in Type-II diabetes), while others do not. By comparing the two strains of mice, scientists may be able to find the genes responsible for the different responses to diet.

The discovery of the scid mouse has allowed scientists to range farther afield and, for example, produce the first animal model of AIDS. It was generally believed that mice could not be infected with AIDS because many human viruses, such as HIV, which causes AIDS, attack only humans and higher primates. But a few years ago, Michael McCune, who was treating AIDS patients at San Francisco General Hospital, had an inspiration: Since scid mice do not have a working immune system, they cannot reject human tissue, he reasoned; then why not implant human immune-system tissues into them, to see if that would allow the AIDS virus to infect the mice? It was a wild idea, he admits. But to the utter astonishment of all his colleagues at Stanford University, where the experiment was carried out, it worked.

McCune's laboratory infected his scid-human mice with the AIDS virus. The mice were treated with the drug AZT at different intervals after infection to test the advantages of early treatment. Hundreds of other antiviral compounds are also being screened, either alone or in combination, to see if they will stop the fatal march of AIDS.

In similar ways, a variety of fetal and adult human tissues can be implanted into scid mice to study cancer, brain disorders, and other conditions, as well as to develop gene therapy. McCune and other scientists have actually grown human lung, pancreas, intestinal, and brain tissues in these animals.

The goal of research by Charles Baum of Systemix, a Palo Alto biotechnology firm, and HHMI investigator Irving Weissman of Stanford, is to isolate the human blood stem cell - a master cell that gives rise to all types of human blood and immune cells, which scientists will need in large quantities to make gene therapy become practical.

Despite such progress, researchers still lack animal models for most human genetic diseases. But this is changing rapidly as scientists learn they no longer have to wait for Mother Nature to make their mutants - they can create them to order. By inserting foreign genes into animal embryos, they can produce "transgenic" animals, whose cells follow the instructions of the interloper genes as well as those of their ancestral genes. The result is an explosion of new information on how genes work in specific cells and how they go about promoting health and disease in both mice and humans.

In 1982, Richard Palmiter, an HHMI investigator at the University of Washington in Seattle, Ralph Brinster of the University of Pennsylvania, and their colleagues injected a modified rat growth-hormone gene into a fertilized mouse egg. The gene normally produces small quantities of growth hormone in the pituitary gland, but the researchers wanted to see what would happen if an animal had unusually high levels of the hormone. Before injecting the gene into the mouse egg, therefore, they attached it to promoters, regions of DNA that control which tissue expresses a gene. The promoters would redirect the gene's expression to other cell types, such as liver cells, where the gene would be freed from its normal controls and would produce large quantities of hormone. Then they implanted the egg into a mouse foster mother. The mother gave birth to normal-sized pups that grew at an unusually rapid pace to become giant mice, nearly twice the size of their littermates. The picture of one of these super-mice was splashed across newspapers and magazines throughout the world.

This experiment paved the way for the first attempt to cure a genetic disorder - dwarfism - in transgenic animals by gene therapy. The mice "patients" were undersized because they lacked sufficient growth hormone. By inserting a modified growth-hormone gene into them, the researchers and Robert Hammer, who was then at the University of Pennsylvania, corrected the genetic defect. The correction was so good that the mice grew slightly larger than normal.

Since that seminal experiment, geneticists have been striving to find permanent cures for a variety of genetic diseases. But first they have to understand the basic biology of the diseases.

A few years ago, five research groups - including the team of Palmiter, Brinster, Richard Behringer, then an HHMI associate in Brinster's lab, Tim Townes of the University of Alabama at Birmingham, and their associates - succeeded in creating various mouse models of sickle-cell disease. A painful and often lethal genetic disease, sickle-cell disease affects about 2 million people around the world, mostly blacks. One in twelve American blacks carries the abnormal gene. More than 50,000 Americans suffer from the disease, which develops in children who inherit the gene from both parents.

Sickle-cell disease was the first genetic disease to be understood at the molecular level. It results from the mutation of a single base in the DNA that codes for hemoglobin, the oxygen-carrying protein in red blood cells - yet there is still no effective treatment for it. Scientists have long been eager for an animal model on which to test new therapies.

To produce such a model, Richard Behringer injected a few hundred copies of the abnormal gene into fertilized mouse eggs. The pups born of these eggs were examined to see which, by chance, had incorporated the human gene into one or more chromosomes. Those with the gene were mated in the hope that the human gene had lodged in their egg or sperm and that the trait would be passed on to the next generation. By breeding these mice with a strain of mice that produced low amounts of mouse hemoglobin, the researchers have created mice with red blood cells that sickle. However, these mice still have some of their normal mouse hemoglobin, Palmiter has pointed out. Further experiments may yield mice with more abnormal human hemoglobin - mice that may be more like sickle-cell patients. Researchers could then test anti-sickling drugs in the mice and explore methods of gene therapy.

In other laboratories, cancers of the eye, breast, lymph tissues, pancreas, and other organs have been induced in mice by cancer genes and combinations of cancer genes. This work is fueling one of the most important revolutions in twentieth-century medicine - the ultimate understanding of cancer as a genetic disease.

It is now believed that cancer is caused by genetic mutations - most often, by a series of mutations, some of which may be inherited. Certain normal genes involved in cell growth, development, and differentiation can be converted into cancer-causing "oncogenes" by mutation. Other genes that normally prevent the uncontrolled growth of cells - "suppressor" genes - can also produce cancer if they are knocked out by genetic mutations. Single mutations are generally not sufficient to cause cancer, but they produce changes that predispose cells to malignant growth. Additional mutations in other genes, caused by damage from the environment, continue the cells' malignant transformation. Thus, cancer is a multi-step process involving the interaction between genes and their environment.

To test the hypothesis, researchers have put a variety of oncogenes into mice, using promoters to direct the genes to specific tissues. In this way, HHMI investigator Philip Leder and his co-workers at Harvard created and patented "onco-mice" - animals that reliably develop breast and lymph cancers. Onco-mice are being used worldwide to test drugs and therapies against those two forms of cancer.

Aside from cancer, cardiovascular disease is the biggest killer in developed countries. Several researchers are using animal models in an effort to uncover the genes responsible for high levels of fats and cholesterol in the blood, to develop tests for identifying people with a genetic predisposition to heart disease, and to explore new therapies.

Michael Brown, Joseph Goldstein, Robert Hammer, and their associates at the University of Texas recently created transgenic mice that never develop high levels of cholesterol in their blood, no matter what they eat, because they carry a human gene for receptor proteins that latch onto LDL, the carrier of "bad" cholesterol, and remove it from the bloodstream.

In the HHMI unit at the University of Michigan, James Wilson is working with mutant rabbits that have a defective gene for making LDL receptors - as do humans with familial hypercholesterolemia, a disease that leads to premature heart attacks. To cure these rabbits, Wilson and his colleagues took liver cells from them, inserted normal genes for LDL receptors into the cells, then reimplanted the treated cells into rabbits with the defect. The animals' cholesterol levels fell for up to a week after the transplant, though they subsequently returned to their previous high.

Most researchers make transgenic animals as Palmiter and Brinster do, by injecting foreign genes into fertilized eggs and then implanting the eggs in a surrogate mother. Others link the foreign genes to a virus and inject the viral combination into an embryo at a later stage of development; the virus then integrates itself into the animal's chromosomes, carrying the foreign genes along with it. Either way, the researchers face a major problem: They cannot specify where in the animal's DNA the foreign gene will become integrated. If a gene is taken up at the wrong spot, it may disrupt a native gene or even cause a lethal mutation; if taken up in the right spot, it may cure a disease. Furthermore one animal might integrate hundreds of copies of the gene, while another animal, under the same experimental conditions, might integrate but a single copy.

Another problem is that researchers can only add genes to the animal's own genome: They cannot get rid of an animal's unwanted or defective gene. The animal's own gene, equivalent to the inserted foreign gene, still functions in some form and could influence the experiment, as appears to have happened in mouse models of sickle-cell disease.

Recently a new technique has overcome these limitations and opened up a whole world of possibilities for making almost any wished-for animal model. It is called homologous recombination, or more loosely, gene targeting, and it is awesomely precise.

In homologous recombination, a desired gene finds an identical, or homologous, sequence of DNA in the animal's genome and swaps places with it. Several teams of researchers, including that of Oliver Smithies, who is now at the University of North Carolina, have devised ingenious ways to achieve this precise exchange.

In 1987, Mario Capecchi and Kirk Thomas of the HHMI unit at the University of Utah showed that such strategies could be particularly effective when used with mouse embryonic stem (ES) cells. These cells, which were first isolated from very early mouse embryos by Martin Evans at Cambridge University and Gail Martin at the University of California, San Francisco, in the early 1980s, are unspecialized precursors of other cells. Each one is capable of giving rise to an entire animal. When Capecchi introduced new genes into ES cells, most of the genes integrated into the chromosomes randomly. But in 1 out of 100 or 1,000 times, depending on the gene, the new gene found its exact counterpart in the mouse genome and integrated itself there, Capecchi says. It was as if the foreign gene had cruised up and down the mouse chromosomes and found its home address.

The challenge, then, was to find those cells in which the genes had landed on their home targets. If they could be identified, a powerful new genetic tool would be available.

Capecchi and his associates have pioneered a double-barreled system for selecting these rare cells. They start out by adding two other genes to the gene that is seeking its homologous site in an ES cell. One of the added genes (the herpes virus thymidine kinase, or "tk" gene) makes a cell vulnerable to the antibiotic gancyclovir. It is placed just outside the homologous sequence, so when the inserted DNA hits its precise target the tk gene is discarded because its sequence does not match up. The cells in which homologous recombination has occurred, therefore, remain invulnerable to gancyclovir, but those in which the DNA integrated randomly still have the tk gene and are killed by the antibiotic. The second added gene, which confers resistance to the antibiotic neomycin, is placed right in the middle of the homologous sequence. When the cells are bathed with neomycin, those that failed to integrate the homologous gene are weeded out. The cells that survive both treatments have the homologous gene right on target. These cells can then be used to recreate a mouse. Whole generations of mice can be raised from these cells, permanently carrying the inserted gene in exactly the right spot.

Homologous recombination allows scientists to carry out a new type of research with transgenic mice: knockout experiments. By introducing defective mouse genes into ES cells, researchers knock out the native gene, and the resulting mouse expresses the defective gene.

The technique should prove important for studying human diseases, such as cystic fibrosis, which until now have had no animal models. But there's a rub, says Capecchi. To do knockout experiments, researchers must already have cloned the human disease gene and its mouse homologue. So far, only a few hundred such genes have been deciphered.

Nevertheless, scientists are reveling in their new freedom to manipulate genomes in many ways. "With gene targeting, you can knock out a gene, replace it, or change it more subtly," says Capecchi. "For example, you can put a gene under the control of a switch so that if you inject a certain drug, the gene is turned on or turned off. " Almost any disease process can now be studied in animals in this way.

Mice remain the favorite subjects for such studies - not only because they are convenient to work with, but because recent advances in mapping both the mouse and human genomes have made comparisons between the two especially fruitful. Scientists who work with transgenic mice confess astonishment and glee at the rapid progress in the field. The ability to add or replace mouse genes at will is leading to a medical revolution, they say, and new experiments are limited only by the human imagination.

Go to next story: The YACs are Here!

Go to Macrophage-tropic HIV - critical for AIDS pathogenesis?

Return to About Biotech directory

Issues Index

About Biotech Index