Mice and pigs are easy; rats are harder. Humans are harder still, but not as hard as monkeys. No one knows why some species are tougher to clone than others. But studies with mice are helping define those differences and should ultimately enable researchers to achieve greater success in obtaining human embryonic stem cells to treat diseases and replace organs.

In terms of therapeutic cloning (otherwise known as nuclear transfer), the most notable achievement so far comes from Woo Suk Hwang of Seoul National University. In February his lab reported on how it transferred the nuclei of several human donor cells into 242 eggs that had their nuclei removed. The batch produced 30 blastocysts (an early stage of embryonic development), of which only one yielded a self-renewing stem cell line.

Hwang's accomplishment highlights the enormous technical difficulty in working with human cells compared with mouse cells. The human egg is finicky: it is larger, stickier and more fragile than a mouse egg. "It is like a balloon instead of a tennis ball," says Hwang, who, instead of taking the standard approach of removing the nucleus of an egg by aspiration, gently squeezed it out of a tiny slit. He found that, compared with the situation in mice, human DNA reprogramming inside the egg takes twice as long, and embryonic stem cells divide half as fast. And unlike the case with mice, researchers must remove the inner cell mass of human embryonic stem cells from the outer (preplacental) cell layer of the blastocyst, or else the cells stop dividing, explains Kevin Eggan of the Society of Fellows at Harvard University.

Other obstacles arise from lack of experience. "We had to start from the bottom and open every door," Hwang recalls. For want of a better selection, he used a derivative of a culture media meant for growing bovine embryonic stem cells. Limited access to human embryonic stem cells as well as legal and ethical issues also slows progress, explains Douglas Melton of Harvard Medical School, who has made available 17 new embryonic stem cell lines he developed from donated embryos. The "starting material" may also make a difference, Melton suggests. "In humans, we have only worked with frozen embryos, but in mice we work with fresh ones," he says of U.S. researchers.

Other quandaries involve the optimal donor cell for nuclear transfer. The cell should be both accessible and efficient. Hwang used cumulous cells from the ovaries--not exactly easy to access and infeasible for males.

As for efficiency, research suggests that the more immature and undifferentiated a donor cell, the better. The most efficient--and the toughest to obtain--are embryonic stem cells, explains Rudolf Jaenisch, a leading expert in mouse cloning at the Whitehead Institute for Biomedical Research in Cambridge, Mass. Next come the relatively immature but rare and inaccessible somatic stem cells, followed by the more accessible but mature differentiated cells that make up tissues. At the most difficult and inefficient extreme lie the exquisitely specialized immune B and T cells and then the nondividing, highly specialized olfactory sensory neurons, which until recently were considered essentially unclonable. In February, however, Jaenisch, Eggan and Richard Axel of Columbia University successfully cloned mice from olfactory neurons, a feat hailed as a technical tour de force.

Now researchers need to find cells that are more efficient than neurons and more practical than embryonic stem cells. Eggan says experiments with mice indicate that nuclear transfer works best when an egg is ripe for fertilization and ready to divide, and it divides more readily if the transferred nucleus comes from a cell that divides frequently, like an immune cell. Embryonic stem cells, however, more readily develop a self-renewing cell line if the donor cell was quiescent, like a neuron or somatic stem cell. The trick is to find a donor cell that starts the egg dividing and that yields perpetual stem cells.

Melton believes such technical difficulties are just a sign that these are the early days of nuclear transfer. "In a few years, this will all seem easy," he predicts.