The death of a wedge issue

I hope this time I'm finally right about this. I've been hopeful that some strategy of developing stem cells would allow us to bypass the absurd ethical restrictions from those who think one type of destruction of an embryo is worse than another. Particularly promising were spermatogonial stem cells, but they could only be made from men (and the procedure might have been unpopular), and placental/amniotic stem cells, which were limited by the ability to passage them without differentiation, and supply (not everybody freezes back their placentas).

The ideal stem cell would have the following properties.
1. It would be immortal until differentiated - meaning that you could make as many as you want from a single cell
2. It would be totipotent - that means it could make any cell in the body - this can be tested by injection into blastocysts to make chimeric animals or by in vitro differentiation in EBS
3. It would be genetically matched to an individual - that would allow tissues derived from the stem cell to be compatible with a recipient.

Adult stem cells just never were able to meet all three of these requirements. Usually, they would excellent ability to differentiate into what they ordinarily make, but they couldn't transdifferentiate - that is make a cell it wouldn't ordinarily make in the body. Blood stem cells could make endless amounts of blood, but it was unclear if they could effectively make anything else. Mesenchymal stem cells could make things like cartilage and bone really well, but appeared limited in making non-mesenchymal cells, like neurons. And many tissues don't appear to have an adult stem cell population, or, their isolation would not be possible without killing or injuring the donor.

Several news reports from today have been discussing this new advance (Alex Palazzo was hinting about this last week in his coverage of this paper). Here is the new paper (subscription not required) here, and here's Nature's coverage:

Last year, Yamanaka introduced a system that uses mouse fibroblasts, a common cell type that can easily be harvested from skin, instead of eggs4. Four genes, which code for four specific proteins known as transcription factors, are transferred into the cells using retroviruses. The proteins trigger the expression of other genes that lead the cells to become pluripotent, meaning that they could potentially become any of the body's cells. Yamanaka calls them induced pluripotent stem cells (iPS cells). "It's easy. There's no trick, no magic," says Yamanaka.

The results were met with amazement, along with a good dose of scepticism. Four factors seemed too simple. And although the cells had some characteristics of embryonic cells -- they formed colonies, could propagate continuously and could form cancerous growths called teratomas -- they lacked others. Introduction of iPS cells into a developing embryo, for example, did not produce a 'chimaera' -- a mouse carrying a mix of DNA from both the original embryo and the iPS cells throughout its body. "I was not comfortable with the term 'pluripotent' last year," says Hans Schöler, a stem-cell specialist at the Max Planck Institute for Molecular Biomedicine in Münster who is not involved with any of the three articles.

This week, Yamanaka presents a second generation of iPS cells1, which pass all these tests. In addition, a group led by Rudolf Jaenisch2 at the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts, and a collaborative effort3 between Konrad Hochedlinger of the Harvard Stem Cell Institute and Kathrin Plath of the University of California, Los Angeles, used the same four factors and got strikingly similar results.

The improvement over last year's results was simple. The four transcription factors used by Yamanaka reprogramme cells inconsistently and inefficiently, so that less than 0.1% of the million cells in a simple skin biopsy will be fully reprogrammed. The difficulty is isolating those in which reprogramming has been successful. Researchers do this by inserting a gene for antibiotic resistance that is activated only when proteins characteristic of stem cells are expressed. The cells can then be doused with antibiotics, killing off the failures.

The protein Yamanaka used as a marker for stem cells last year was not terribly good at identifying reprogrammed cells. This time, all three groups used two other protein markers -- Nanog and Oct4 -- to great effect. All three groups were able to produce chimaeric mice using iPS cells isolated in this way; and the mice passed iPS DNA on to their offspring.

Jaenisch also used a special embryo to produce fetuses whose cells were derived entirely from iPS cells. "Only the best embryonic stem cells can do this," he says.

Adult stem cell hypers shouldn't claim victory yet. These have major promise but they haven't killed the wedge issue yet (something I really am hoping for). For one, they haven't been able to jump from mice to humans:

But applying the method to human cells has yet to be successful. "We are working very hard -- day and night," says Yamanaka. It will probably require more transcription factors, he adds.

If it works, researchers could produce iPS cells from patients with conditions such as Parkinson's disease or diabetes and observe the molecular changes in the cells as they develop. This 'disease in a dish' would offer the chance to see how different environmental factors contribute to the condition, and to test the ability of drugs to check disease progression.

The second major problem is that two cancer risks are created by these cells. The first is that the retroviruses used to transform the adult cells into ES cells randomly inserts into the genome, causing a cancer risk. The second is that in order to get these genes to be expressed inappropriately, you have to use constitutive promoters to drive expression - in other words, the genes keep on getting expressed even after the cells are re-differentiating - which may be causing cancerous transformations in these cells.

But the iPS cells aren't perfect, and could not be used safely to make genetically matched cells for transplant in, for example, spinal-cord injuries. Yamanaka found that one of the factors seems to contribute to cancer in 20% of his chimaeric mice. He thinks this can be fixed, but the retroviruses used may themselves also cause mutations and cancer. "This is really dangerous. We would never transplant these into a patient," says Jaenisch. In his view, research into embryonic stem cells made by cloning remains "absolutely essential".

So this is not a total victory for ES stem cell research, but it's very hopeful. Ideally they would be able to create this transformation in adult cells by just injecting the proteins these genes make - but the critical issue then is identifying the rare cells that gets transformed.

Alternately, the promoters could be changed from constitutive to drug-activated, so only with administration of tetracycline the genes will get expressed. That way, once they're differentiating, the genes can be shut off, avoiding the cancer transformation.

They also will need to deal with the problem of random retroviral insertion into the genome as retroviruses can cause cancer all on their own. This could be bypassed with an integrase system (which may work ideally) that allows for insertion into a distinct and safe chromosomal location, or possibly a different viral system could be used - like adenovirus - that doesn't lead to genomic insertion.

I'll need to make time to fully read the paper and I'll post again with a full review of this article. I'll also have some fun going through the adult stem cell hypers who will inevitably start taking credit for something they had nothing to do with (and is still far from replacing embryo-derived ES cells).