How Well Do Mice Model Humans?

STRIKING RESEMBLANCE: James Croom, who studies Down syndrome mice at North Carolina State University, says the animals are providing valuable information useful to humans. When a page-one article in the May 3, 1998, Sunday New York Times portrayed angiogenesis inhibitors that fight cancer in mice as being possible just around the corner for humans, criticism for raising false hopes erupted. Merely 10 weeks later, however, when researchers from the University of Hawaii reported cloning the first mouse from an adult cell's nucleus, many scientists regarded the work as a signal that cloning humans had come a giant step closer. How can the mouse be deemed too unlike humans in the first circumstance to provide useful information, yet disturbingly similar in the second? The fact is, researchers say, mice provide an excellent model of human physiology, but that is just what they are--a model. "People have a basic misunderstanding of the mouse. They get upset that the exact pathophysiology might be different, but a mouse is not a total mimic. A mouse is a discovery tool, a device, to understand the molecular pathways that underlie disease processes," says Kenneth Paigen, director of the Jackson Laboratory, a nonprofit mammalian research facility in Bar Harbor, Maine. "You can't do a lot of types of experiments on humans. You can't order people to mate, and you can't make a human knockout. We have to turn to models," he adds. But Mus musculus is a lab favorite for practical reasons, too. "Mice are cheap, small, and not terribly cognizant of their environment, so you can handle them without much perturbing the experiment," relates Thomas Wilkie, an assistant professor of pharmacology at the University of Texas Southwestern Medical Center. Mouse mutants were valuable in biomedical research long before transgenics and knockouts came on the scene. "Over the past several decades, many inbred strains with highly specific characteristics have been established that have been important in genetics, cancer, and immunology research," says Steven Barthold, director of the center for comparative medicine and the mouse biology program at the University of California, Davis. The Jackson Laboratory, which opened in 1929 to breed mice to study cancer, provides many such inbred lines. Sometimes mouse work weds the old with the new. Sally Camper, an associate professor of human genetics at the University of Michigan Medical School, uses a mouse mutant called shaker-2 to study human deafness. The shaker mouse tosses its head, walks in circles, lacks a startle reflex, and is profoundly deaf, due to an X-ray-induced deletion of an ancestor's gene in 1928. Using transgenic technology, Camper's group introduced pieces of mouse genome to see which one corrected the deafness (F. J. Probst et al., Science, 280:1444-7, and A. Wang et al., Science, 280:1447-50, May 29, 1998). Mice were the basis of monoclonal antibody (MAb) technology, developed in 1975 (C. Milstein, Scientific American, 243:66-74, 1980). A mouse's B lymphocyte fused with a mouse's cancer cell mass-produces a single, or monoclonal, type of antibody. Tagged MAbs could then be used to detect or treat a range of disorders. But the technology fell short of "magic bullet" expectations. "We originally thought that if we made mouse antibodies, we could throw them into a patient and, shazam, there'd be a cure. Well, it didn't work," recalls Donald Drakeman, president and chief executive officer of Medarex Inc., a biopharmaceutical company in Annandale, N.J. The problem with MAb technology was that the human immune system rejected mouse-derived antibodies. So, researchers replaced most of the mouse's antibody genes with human DNA. A few such "humanized" antibodies have reached the clinic, including the recently approved Herceptin, which treats a specific type of breast cancer. But Medarex has removed the mouse antibody genes and replaced them with the human versions, creating "human mice" that produce completely human antibodies. Says Drakeman, "Antibodies have been around forever, and people expected a big breakthrough technology in the 1980s. But monoclonal antibody technology is really going to fulfill its promise in the first decade of the new millennium." Medarex's antibodies are headed toward clinical trials to treat autoimmune disorders. KEEP PERSPECTIVE: Jerry Shay stresses that although the benefits of mice far outweigh the negatives, there are important differences between the mouse and human species. Mice have eclectic uses. They are used to study such varied disorders as AIDS, multiple sclerosis, and seizures; to probe intracellular effects of neurotoxins in bioweapons research; and to dampen transplant rejection. In basic research, mice have revealed how the vertebrate embryo develops and how changes in the prenatal environment affect learning. Some mice that are given human genes acquire new skills. One strain can see farther than normal into the red part of the spectrum, thanks to a human photopigment gene, and another manufactures human growth hormone. Mice are particularly valuable in identifying human genes, because genome organization corresponds so closely between the two species. "Mammalian genomes are all derived from the same original mammalian chromosome set. Take that set, cut it into about 150 pieces, and shuffle the pieces. One way produces a human, another way, a mouse. Within the fragments, the gene order is conserved," says Paigen. Databases match human and mouse genes. Many anatomical structures are similar between the species, too. For example, Karen Avraham, a professor of human genetics at Tel Aviv University, has used mice to help identify and analyze several genes that cause deafness in humans (O. Vahava et al., Science, 279:1950-3, March 20, 1998). "It is much easier to dissect the biological function of a gene in a mouse model [than in a human], especially since the structure and function of the inner ear are so similar between the two mammals," she says. Mouse variants arise in several ways. Some mutations are spontaneous, such as the seizure-stricken "stargazer," with disrupted calcium channels. It appeared in a stock at the Jackson Laboratory in 1990. Radiation or chemicals also are used to induce mutations. Other manipulations add or delete genes: a transgenic mouse receives a gene, and a knockout mouse loses one. A combination transgenic/knockout animal might have a human gene added and its mouse counterpart jettisoned. In 1993, the Jackson Laboratory started the Mouse Mutant Resource to distribute induced, transgenic, and knockout mice from research labs. "We set up outside committees that meet once a month to consider new admissions. Once a strain is selected, it is imported to the lab, where the mice are hysterectomy-derived, or 'rederived.' Once we are sure that the offspring are clean, we put them into a colony. We also cryopreserve sperm or embryos once the mice are rederived, to protect against accidental loss," explains director John Sharp. Some mouse models strikingly resemble their human counterparts. A mouse with an extra chunk of mouse chromosome 16 shares physical features with a person who has Down syndrome, which is caused by an extra copy of human chromosome 21, notes James Croom, a professor of nutrition and physiology at North Carolina State University. "Their ears are set back on the head, the snout is flatter, and they seem to have a dowager's hump," he says. Croom studies nutrient levels in the bloodstreams of the Down syndrome mice, which is providing information on the metabolic derangements that are part of the condition in humans. A mouse with sickle cell disease is startlingly like a human with the disease, because human globin genes replace its own. In sickle cell disease, two abnormal beta chains of the hemoglobin tetramer cause the molecules to polymerize under low oxygen conditions, which bends the cells containing them into a sickle shape. The sickled cells block circulation, causing strokes, organ death, and great pain. The two alpha globin chains are normal. Attempts to replace only the beta globin genes with human versions were unsuccessful, so two research teams set about substituting the entire tetramer. SUCCESSFUL BREEDING: Inbred strains of mice have been important to researchers studying genetics, cancer, and immunology, says Steven Barthold of U.C.-Davis. Tom Ryan, a postdoctoral research associate in the laboratory of Tim Townes, a professor of biochemical and molecular genetics at the University of Alabama, knocked out the mouse beta globin genes. Meanwhile, postdoc Chris Paszty, working with Edward Rubin at Lawrence Berkeley National Laboratory, knocked out the mouse alpha genes. The two groups discovered each other and joined forces to create a mouse with fully human hemoglobin, including the fetal version, and with the sickle type of beta subunits (T. M. Ryan et al., Science, 278:873-6, 1997; C. Paszty et al., Science, 278:876-8, 1997). The sickle cell mice, which have the same symptoms as humans, are in great demand. "We are using the mice to test gene therapy strategies. Many groups have requested the mice to test drugs that may directly inhibit the polymerization of sickle hemoglobin, enhance gamma [fetal] globin gene expression, or alter the physiology of sickled red blood cells," says Townes. Mouse models can sometimes reveal unknown aspects of human pathology, or explain old findings using new technology. For example, a mouse transgenic for human growth hormone, created by Steven Goldstein, professor of surgery at the University of Michigan Medical School, grows bigger bones, but the strength declines as the rodent passes puberty. This information may help physicians manage adolescents who received growth hormone as children to treat short stature. A transgenic mouse is furthering understanding of Huntington's disease (HD). In this disorder, an abnormally long repeat of the DNA triplet CAG sticks extra glutamates in the encoded protein, huntingtin, which somehow causes it to accumulate and trigger cell death in the striatum. In humans, HD causes uncontrollable movements and personality changes beginning in middle age. A mouse transgenic for HD, developed by Gillian Bates in the division of medicine and molecular genetics at Guy's Hospital in London, has symptoms similar to those of human sufferers, including the movements, seizures, and decreased brain size, weight, and muscle mass. Bates's team saw unusual clumps in parts of mouse brains and used a MAb to identify the material as huntingtin (S. W. Davies et al., Cell, 90:537-48 and E. Scherzinger et al., Cell, 90:549-58, 1997). These deposits, it turned out, correspond to similar ones that Columbia University researchers had discovered in 1979 in neuron nuclei in the brains of people who had died of HD, but that they couldn't identify further because they lacked the MAb tool. "This is a case where pathology in the mouse led to discovery of the pathology in humans," says Jackson Laboratory's Sharp. Sometimes a subtle difference in a biochemical pathway between mouse and human has a huge impact on health. This is the case for Lesch-Nyhan syndrome, an enzyme deficiency. "In humans, it causes mental retardation and severe behavioral deficits that cause self-mutilation. But a mouse knockout has no phenotype," explains Geoffrey Duyk, chief scientific officer of Exelixis Pharmaceuticals Inc., a private genomics company in South San Francisco, Calif. But he adds that when a different enzyme in the same pathway in the mouse is knocked out, the same symptoms result. Retinoblastoma, an eye cancer, is another disorder with subtle distinctions between the species. "A heterozygote in humans produces retinal tumors in children and, as they grow up, osteosarcoma. In mice, it produces tumors of the pituitary and other endocrine glands. The same gene produces a tumor syndrome, but it is different," says Duyk. The mouse's life cycle and lifestyle may make it a less-than-perfect model for cancer in humans. "Mice are like people, but are 3,000 times smaller, grow after birth about 100 times faster, and age after maturity about 30 times faster," says David Harrison, senior staff scientist at the Jackson Laboratory. The smaller size of mice may enable drugs to travel beyond primary tumors, preventing metastases. In humans, such treatments would have more localized and limited effects, adds Texas Southwestern's Wilkie. The Jackson Laboratory: www.jax.org/ Davis Human/Mouse Homology Map: www3.ncbi.nlm.nih.gov/Homology/ Natural selection may explain why mice develop cancer easily. "In the wild, a mouse is short-lived, and few survive their first winter. A mouse's purpose is to eat, breed, and keep itself out of the food chain. A long-lived species such as the human, which takes a long period of time to nurture offspring to perpetuate the species, probably needs a mechanism to protect against developing cancer. Maybe mice don't need to evolve such a protective mechanism," says Jerry Shay, a professor of neuroscience at the University of Texas Southwestern Medical Center. If we didn't have this protection, he adds, more people would get cancer before middle age. Press reports that are too quick to leap from mouse to human, causing the public to doubt the value of mice when clinical success is not imminent, do a disservice to this extremely useful model organism. Says Duyk, "The mouse allows a close approximation of the physiology of humans, and is the most flexible organism to manipulate. In the pecking order, the mouse is the most important animal for studying the human." Shay stresses perspective in his summation: "I don't want to blast the mouse as a model--the positives far outweigh the negatives--but it is important to recognize species differences, too." Ricki Lewis is a contributing editor to The Scientist , and author of several college life science textbooks for McGraw-Hill Publishers.

How Well Do Mice Model Humans?

Interested in reading more?

Become a Member of