Such "humanized" animals could have countless uses. They would almost certainly provide better ways to test a new drug's efficacy and toxicity, for example, than the ordinary mice typically used today.
But few scientists are eager to do that experiment. The risk, they say, is that some human cells will find their way to the developing testes or ovaries, where they might grow into human sperm and eggs. If two such chimeras -- say, mice -- were to mate, a human embryo might form, trapped in a mouse.
Not everyone agrees that this would be a terrible result.
"What would be so dreadful?" asked Ann McLaren, a renowned developmental biologist at the University of Cambridge in England. After all, she said, no human embryo could develop successfully in a mouse womb. It would simply die, she told the academy. No harm done.
But others disagree -- if only out of fear of a public backlash.
"Certainly you'd get a negative response from people to have a human embryo trying to grow in the wrong place," said Cynthia B. Cohen, a senior research fellow at Georgetown University's Kennedy Institute of Ethics and a member of Canada's Stem Cell Oversight Committee, which supported a ban on such experiments there.
But what about experiments in which scientists add human stem cells not to an animal embryo but to an animal fetus, which has already made its eggs and sperm? Then the only question is how human a creature one dares to make.
In one ongoing set of experiments, Jeffrey L. Platt at the Mayo Clinic in Rochester, Minn., has created human-pig chimeras by adding human-blood-forming stem cells to pig fetuses. The resulting pigs have both pig and human blood in their vessels. And it's not just pig blood cells being swept along with human blood cells; some of the cells themselves have merged, creating hybrids.
It is important to have learned that human and pig cells can fuse, Platt said, because he and others have been considering transplanting modified pig organs into people and have been wondering if that might pose a risk of pig viruses getting into patient's cells. Now scientists know the risk is real, he said, because the viruses may gain access when the two cells fuse.
In other experiments led by Esmail Zanjani, chairman of animal biotechnology at the University of Nevada at Reno, scientists have been adding human stem cells to sheep fetuses. The team now has sheep whose livers are up to 80 percent human -- and make all the compounds human livers make.
Zanjani's goal is to make the humanized livers available to people who need transplants. The sheep portions will be rejected by the immune system, he predicted, while the human part will take root.
"I don't see why anyone would raise objections to our work," Zanjani said in an interview.
Perhaps the most ambitious efforts to make use of chimeras come from Irving Weissman, director of Stanford University's Institute of Cancer/Stem Cell Biology and Medicine. Weissman helped make the first mouse with a nearly complete human immune system -- an animal that has proved invaluable for tests of new drugs against the AIDS virus, which does not infect conventional mice.
More recently his team injected human neural stem cells into mouse fetuses, creating mice whose brains are about 1 percent human. By dissecting the mice at various stages, the researchers were able to see how the added brain cells moved about as they multiplied and made connections with mouse cells.
Already, he said, they have learned things they "never would have learned had there been a bioethical ban."
Now he wants to add human brain stem cells that have the defects that cause Parkinson's disease, Lou Gehrig's disease and other brain ailments -- and study how those cells make connections.
Scientists suspect that these diseases, though they manifest themselves in adulthood, begin when something goes wrong early in development. If those errors can be found, researchers would have a much better chance of designing useful drugs, Weissman said. And those drugs could be tested in the chimeras in ways not possible in patients.
Now Weissman says he is thinking about making chimeric mice whose brains are 100 percent human. He proposes keeping tabs on the mice as they develop. If the brains look as if they are taking on a distinctly human architecture -- a development that could hint at a glimmer of humanness -- they could be killed, he said. If they look as if they are organizing themselves in a mouse brain architecture, they could be used for research.
So far this is just a "thought experiment," Weissman said, but he asked the university's ethics group for an opinion anyway.
"Everyone said the mice would be useful," he said. "But no one was sure if it should be done."