Dr. Yu，可能她是第一作者。

Human embryonic stem cells
Me too, too
Nov 22nd 2007
From The Economist print edition

How to make human embryonic stem cells without destroying human embryos

SCIENCE moves fast. On November 14th Nature, one of the world's leading scientific journals, published a paper about the creation of embryonic stem cells using a technique called somatic-cell nuclear transfer (basically, taking the nucleus from a body cell and putting it in an unfertilised egg). This made the news because the researchers had performed their trick in monkeys. The result was thus the first primate embryos to have been cloned, as earlier reports of human cloning turned out to have been fraudulent.

There is, however, a second way of making an embryonic stem cell that has the genes of an existing individual. This is to take a body cell and order it to turn into a stem cell using a set of molecular instructions. A group of researchers at Kyoto University, led by Shinya Yamanaka, did this with mice last year. A commentary published alongside the Nature paper, by Ian Wilmut, the man who produced the first cloned sheep (using somatic-cell nuclear transfer to do so), referred to Dr Yamanaka's technique but said, “There is so far no sign that this approach could be effective in human cells.”

Well, there is now. Two papers published this week, one in Cell by Dr Yamanaka and one in Science by Junying Yu of the University of Wisconsin-Madison, have shown how to do the trick in human cells. Dr Wilmut is so impressed with their data that he has said he is now going to concentrate his efforts on this alternative technique.

People are interested in embryonic stem cells because they have the potential to grow into any other type of body cell. That raises the possibility of replacing worn-out tissues and organs with new ones that have the same genes as the patient and would thus be in no danger of being rejected by his immune system. This idea, known as regenerative medicine, is still some way off, but in the more immediate future the ability to grow pure samples of particular body tissues would be of enormous value in developing drugs. Genetically bespoke stem cells thus have the potential to be big business.

Non-destructive testing
The problem is that making them using somatic-cell nuclear transfer involves breaking up viable embryos, since the cells in question are found inside early-stage embryos, called blastocysts. That offends some people's moral sensibilities. Indeed, the same objection also applies to stem cells that are extracted from embryos produced without nuclear transfer. (Usually these are surplus to requirements for in vitro fertilisation.) Both problems would go away if cells that behave like embryonic stem cells could be made without destroying embryos. And that is where Dr Yamanaka and Dr Yu may be able to help out. For, instead of starting with human embryos, they started with human body cells. Dr Yamanaka used skin cells from the face, whereas Dr Yu plumped for ones from the foreskin.

The trick is to persuade such cells to forget what they have become and remember what they (or, rather, their ancestors) once were in the embryo. In other words, specialised cells are ordered to change back into stem cells. This can be done by tinkering with their genes.

What makes a cell “this” rather than “that” is the way its genes are expressed. Though most cells in any individual have the same set of genes, not all of these genes are active in any given cell. The activity of genes is controlled by proteins called transcription factors that are, themselves, the products of genes. What Dr Yamanaka discovered in his work on mice was a group of transcription factors that switched on the pattern of gene expression which says, “I am an embryonic stem cell.”

Surprisingly, there were only four of them, known as Oct3/4, Sox2, c-Myc and Klf4. At least, that was the case for mice, as he found by tinkering with various combinations of transcription factors until he discovered a set that would do the trick. And mice and men being pretty similar at the genetic level, it is not too surprising that the same set works in people.

Indeed, it seems to work well. Dr Yamanaka's team has, for instance, been able to persuade its stem cells to turn into nerve cells and heart cells. That is encouraging for regenerative-medicine enthusiasts. The new cells also produced all the molecules known to be characteristic of stem cells and, when grown in bulk, organised themselves into three-layered structures characteristic of embryos. (Each layer is responsible for different sorts of tissue and organ.)

To be fair, all this was also true of the cells described in the Nature paper, which was written by Shoukhrat Mitalipov, of the Oregon Stem Cell Centre in Beaverton, and his colleagues. But his cells came from macaques, not humans, and his technique involves destroying the embryo.

Dr Yamanaka, then, has shown what is possible, but Dr Yu may have trumped him. That is because one of Dr Yamanaka's transcription factors, c-Myc, sometimes has the unfortunate side-effect of causing cancer. By going back to basics, Dr Yu found a different combination (Oct4, Sox2, Nanog and Lin28) that does not suffer from this problem. Cells transmuted by this combination also have the characteristics of embryonic stem cells.

The main thing to be said against the methods used by Dr Yamanaka and Dr Yu is the way that they persuade their experimental cells to churn out the appropriate transcription factors. They do it by infecting them with retroviruses that have had the four relevant genes spliced into their genomes. Retroviruses work by adding their genes into the chromosomes of their hosts, which then merrily churn out bits of new viruses. As Dr Wilmut points out, that would make transplanting tissues created this way into people too risky to contemplate. But a modified approach that does not involve viruses is likely to be the ultimate way of making human stem cells. And no embryo need be destroyed.