Esmail Zanjani, a researcher at the University of Nevada at Reno, has been able to grow mostly human livers in sheep. Zanjani achieved this by injecting into growing sheep fetuses either adult stem cells derived from bone marrow, or embryonic stem cells from one of the federally approved stem cell colonies.

The research was originally aimed at finding out if it would be possible to transplant stem cells into developing human fetuses to correct defects in utero. But Zanjani observed the human cells integrating into and then proliferating within a wide range of organs and tissues, including the pancreas, heart, skin and liver. Now, it looks like it will be possible to grow inside animals human cells, tissues and organs that might be suitable for transplants into people.

Let's say you need a new liver. Zanjani would take some of your bone marrow stem cells, and inject them into a fetal sheep at the proper moment. A few weeks later the lamb would be born with a liver made up chiefly of your cells. The lamb would be sacrificed and your new liver installed. Once installed, your immune system would eliminate the lamb's liver cells, leaving behind a brand new organ perfectly matched to your body. Providing sick people with life-saving transplants is certainly a morally worthy activity.

It is now common to place single human genes into plants and animals and even bacteria, to produce various therapeutic proteins, including insulin and human growth hormone. Few people now believe that putting a single human gene into another creature transforms that creature into a human being.

It turns out that many genes are like Animal-Kingdom cassettes—they can be mixed and matched across species. A gene crucial to building a fruit fly's eye will trigger eye development in a frog. Now that both the human and mouse genomes have been sequenced, researchers know that 99 percent of mouse genes have homologues in humans; even more amazingly, 96 percent are present in the same order on the genome. Of course, how those genes are expressed is very different, and mouse proteins, while similar, also differ in crucial ways.

Nevertheless, mixing human and animal genes and cells does pose some moral conundrums. First, consider the possibility of crossbreeding humans with other primates. There is some evidence that such mixing might succeed. Researcher J. Michael Bedford reported in 1977 that human sperm could penetrate the protective outer membranes of gibbon eggs. So far, from what we know, no one has attempted to create a human/chimpanzee hybrid. But would that be wrong?

Bioethicist Joseph Fletcher once suggested that it would be ethical to create parahumans, e.g., human/animal hybrids to do dangerous and demeaning jobs. Is Fletcher's proposal all that different from training dolphins to find underwater explosives, or using dogs to corral dangerous criminals?

Part of the problem arises from calling the creatures parahuman without defining which human characteristics might be added to them. Would giving an animal the ability to walk upright on two legs be morally problematic? Probably not. Would giving such creatures the ability to talk; that is, the capacity to understand and communicate with other language users, be morally problematic? That certainly raises the bar.

Human/animal crossbreeding is not the only way in which animals might be given the ability to talk. In 2002, researchers in Britain discovered that the FOXP2 gene in humans is required for articulate speech. While it is not the language gene, it is certainly one of the genes necessary for the ability to talk. The proteins produced by human FOXP2 gene differ by only two amino acids from the proteins produced by the FOXP2 gene in chimpanzees, gorillas, and orangutans. Would research that creates a transgenic chimpanzee with human FOXP2 genes elicit moral concern? Since FOXP2 orchestrates the actions of a variety of genes early in the development of the brains of human fetuses, and might have similar effects in chimpanzee fetuses, there may be grounds for ethical worries about such an experiment.

But what about just installing human brain cells directly in animals? Stanford University's Irving Weissman has injected human neural stem cells from aborted fetuses into the brains of fetal mice, where they have integrated and grown into human neurons and glia that intermingle with mouse brain cells, making up about 1 percent of the tissue in their brains. However, there is no evidence the chimeric mice began to contemplate the meaning of life. We need to give such chimeric mice no more or less moral consideration than we already give laboratory mice.

Weissman has said that he would like to inject human stem cells into the developing brains of fetal mice, with the goal of producing mouse brains composed chiefly of human brain cells. Such mice might be useful for testing drugs to cure or prevent various human brain diseases. Since the brains would have the architecture of mouse brains, it is unlikely that they would become biotech Stuart Littles and exhibit any characteristics that would cause us moral concerns.

But what about injecting human brain stem cells into the developing brains of fetal chimpanzees? That's clearly a bit closer to the line, but if human cells are simply integrated into the typical architecture of a chimpanzee's brain, then again, it would probably create no new ethical problems.

Beside the possibility of giving human characteristics to animals, injecting human stem cells into non-humans could create other moral concerns. For example, stem cells might transdifferentiate into gamete-producing cells, and integrate themselves into the ovaries and testes of mice, where they would produce completely human eggs and sperm. One could imagine such chimeric male and female mice mating and producing a completely human embryo. Of course, that embryo would be unable to develop in the uterus of a mouse, so the world would not have to deal with the birth of a child whose mom was a rodent. But again, what if this research were done with larger chimeric animals—say, cows—that could possibly carry a human baby to term?

But chimeric mice could also be used to help people overcome infertility. Bone marrow stem cells from an infertile woman or man might be injected into a fetal mouse, where they could be transdifferentiated into gamete-producing cells. Gametes might be harvested from the mice and used in IVF procedures to engender a child. Assuming it's medically safe, producing a child in this way would not be unethical.

Finally, it has to be asked: would eating a liver composed chiefly of human liver cells grown in a sheep be cannibalism? I say yes; don't do it. Save them for transplants.

As humanity's biotechnological prowess increases, we will confront again and again the question of what, if any, limits should be placed on research that mixes human and animal genes, cells and tissues. The main ethical concern about such research is not the creation of improved and useful animals, but the risk of producing what would be, in effect, diminished human beings.