President Bush's Executive Order 13435 renames the “Human Embryonic Stem Cell Registry” the “Human Pluripotent Stem Cell Registry”

Within a month, the US National Institutes of Health (NIH) hopes to start adding to the registry that lists the human embryonic stem cell (ES cell) lines eligible for US federal research funding. The registry currently contains only the human ES cell lines already in existence in August 2001, when President George W. Bush declared that no federal funds could be used for subsequently created lines. But of the dozens of human ES cell lines established since then, none will be added to the registry (with the possible exception of a few created by an unconventional technique that removes individual cells from embryos without destroying them). Instead, the word 'pluripotent' will replace the word 'embryonic' in the name of the NIH Human Embryonic Stem Cell Registry, and the list will begin to include cell lines derived from non-embryonic sources.

The impetus for the change comes from the White House in the form of a executive order that touts the potential of non-embryonic stem cells, and accompanied Bush's veto of popular legislation to lift restrictions on federal funding for research on human ES cells1. Researchers who derive and assess potentially qualifying lines will be given higher priority for new NIH grants and will be eligible for supplemental funds for existing grants. Before that happens, however, the NIH Stem Cell Task Force must set criteria for pluripotency in human cells. Politics has, essentially, mandated that an answer be found to a fundamental scientific question.

Asked about registering lines already clearly eligible for federal funding, scientists interviewed for this article generally reacted with a mixture of confusion, annoyance and indifference. One called the plan a “distraction that won't open any doors”, and then asked not to be identified discussing politics. Some worried that political pressure on the NIH would hamper its ability to set a compelling definition. “I look forward to the day when the [registry] website is simply shut down, as its mere existence is a constant reminder of a public policy that does not serve the public good,” says stem-cell pioneer James Thomson at the University of Wisconsin-Madison. Of the two dozen or so human ES cell lines eligible for US federal funding, his are the most widely used.

“The term pluripotent has been used for every type of stem cell,” says Anthony Atala at Wake Forest University Baptist Medical Center, who recently identified stem cells in amniotic fluid that can differentiate into cell types representing bone, endothelial, fat, liver, muscle, and neuronal lineages2. Originally, 'pluripotent' meant a cell could give rise to cells representing the three germ layers found in early embryos. To assess this property, scientists can inject mouse or human cells under the skin of an immune-compromised mouse and see whether they form a benign tumour known as a teratoma. But as the field advanced, says Atala, more requirements were added to the term.

Defining an approach

The International Society for Stem Cell Research (ISSCR) has defined the term 'pluripotent' as: “The state of a single cell that is capable of differentiating into all tissues of an organism, but not alone capable of sustaining full organismal development.” The need for a definition, says ISSCR's president George Daley, is because experiments involving human pluripotent cells should be reviewed by ethical oversight committees, not because pluripotent cells derived without destroying embryos can be expected to replace ES cells.

How to evaluate human cells for pluripotency is unclear. Assessing pluripotency is, however, routine in mice. The most rigorous test is called tetraploid complementation and exploits the fact that pluripotent cells cannot make placenta. Putative pluripotent cells are mixed with a special tetraploid embryo that forms placental tissue but not body cells. The chimaeric embryos are implanted into surrogate mothers, and if pups are born, the mixed-in cells were pluripotent. But even some mouse ES cells fail this test, says Rudolf Jaenisch at the Whitehead Institute in Cambridge, Massachusetts, who pioneered the technique. A less rigorous test mixes cells with a normal mouse embryo and then tracks where those cells' descendants end up in the body. In particular, candidate pluripotent cells should be able to undergo meiosis to give rise to the germ line (that is, sperm and eggs). Attempts to generate humans to test cells' pluripotency is neither feasible nor ethical. However, mouse cells capable of forming teratomas routinely fail these additional tests.

The NIH's criteria for human pluripotency will probably include teratoma formation alongside microarray assays for transcription factors and other gene activity associated with pluripotency, says Story Landis, head of the NIH Stem Cell Task Force. The human ES cells already on the registry probably would not need to be reassessed, but cells derived from non-embryonic sources such as amniotic fluid or reprogrammed adult cells could be officially deemed pluripotent and added to the registry once the NIH sets its criteria. At the same time, the NIH also plans to resolve whether some ambiguous cases, such as cell lines derived from 'dead' or biopsied embryos, are eligible for federal funding.

“Being on the registry seems to be important to a lot of people,” Landis says, even when no one doubts the cells are eligible for federal funding. For example, representatives of groups that store umbilical cord blood have made enquiries, although their materials do not qualify as cell lines. Landis declines to speculate on their motivation, but researchers naturally want to boost the prestige and commercial value of their cell lines, and getting listed on the NIH registry would be one way to do that.

Moving from mouse to man

Calls for more research into non-embryonic sources of pluripotent stem cells were fuelled by a breakthrough reported this summer, when three sets of investigators reported that they had been able to convert mouse skin cells into cells that behaved almost identically to ES cells and could also form gametes in chimaeric mice3,4,5.

The researchers used retroviruses to insert extra copies of four genes associated with pluripotency into cultured mouse skin cells. In a process that took weeks and happened only in a handful of cells, the inserted genes somehow reprogrammed gene expression to a state very similar to that in mouse ES cells. Jaenisch led one of these groups; he believes that the epigenetic machinery within the cells reverts to a state resembling that found in the inner cell mass of embryos, from which ES cells are derived.

No live-born mouse pups produced through tetraploid complementation using reprogrammed cells have been reported yet, and though Jaenisch attributes this to an insufficient number of attempts, an imperfect resetting of the epigenetic state might also be a reason. In the chimaeric mice generated from these reprogrammed cells, cells from the host embryo may help reset the epigenetic state as the germ line develops. But for reprogrammed cells to form mice through tetraploid complementation, “they have to be in this state immediately at the time of injection into the tetraploid host embryo,” Jaenisch explains.

Even if the epigenetic state for pluripotency can be defined and manipulated in mice, those advances may not apply to humans. Mechanisms for silencing genes may be “deeper” in longer-lived organisms, says Roger Pedersen of Cambridge University in the UK. Moreover, pluripotency is maintained differently in mouse and human cells. Similar players (the transcription factors Sox2, Oct4, and Nanog) are involved, but they seem to regulate and be regulated by different genes. This summer, Pedersen was in one of two groups that succeeded in making stem-cell lines from mouse postimplantation embryos previously deemed too old to produce such cells.6,7,8 These so-called 'epiblast stem cells' formed teratomas reliably, but when mixed with normal mouse embryos they contributed to chimaeras in less than 1% of attempts, and even then the epiblast stem cells contributed to few tissues. Intriguingly, the regulation of pluripotency genes in these cells seemed closer to that observed in human cells.

The markers don't define embryonic stem cells. Embryonic stem cells define the markers. Roger Pedersen, Cambridge University

An assessment of 59 human ES cell lines published this year9 found that they are broadly similar in terms of cell-surface markers, patterns of gene expression, and genomic imprinting. Gold-standard markers for pluripotency exist, says Pedersen, who was part of the large consortium conducting the study. But he warns that simply knowing that a marker connotes pluripotency in ES cells does not mean it can confer pluripotency in other kinds of cells. “The markers don't define embryonic stem cells,” he says. “Embryonic stem cells define the markers.”

Within the inner cell mass of the blastocyst, well before germ layers form, embryonic cells show incredible heterogeneity, says Susan Fisher of the University of California, San Francisco. Assuming that those differences are not artefacts of culture, she says, that could explain why ES cell lines show different properties, such as how well they proliferate in culture or how they are inclined to differentiate. Rather than setting a definition of pluripotency that can be applied to new cell lines, she says, new cell lines should help guide a definition.

A practical definition

“At this stage I think the best we can do is to show that stem-cell lines deemed to be pluripotent are capable of generating cells typical of each of the three germ layers needed for normal human development,” says Glyn Stacey, head of the UK Stem Cell Bank at the National Institute for Biological Standards and Control in Hertfordshire, UK. Unfortunately, teratoma assays assess cells on morphology and histology rather than on function, and while in vitro assays can test some aspects of cell function (such as neural activity or insulin secretion), they cannot do so in the context of a working organ.

Of all body cells, gametes are believed to go through the most extensive epigenetic modifications. Perhaps getting cells to form sperm and egg (as well as the three germ layers) in a dish could show that human cells are pluripotent, says Janet Rossant, who studies mouse embryology at the Hospital for Sick Children in Toronto. The problem is that scientists have so far only been able to do this with mouse, not human, ES cells. “It's very hard to define operationally what a pluripotent cell would be,” says Rossant., “We don't have a gold-standard assay for pluripotency in human cells.”

It may not be possible to test human ES cells for pluripotency directly, says Jeanne Loring, a professor of developmental biology at the Scripps Research Institute, but at least you know that the inner cell mass in other embryos can, under the right conditions, develop into a person. “There just aren't any generally acceptable criteria that are stringent enough for cells from non-embryonic sources,” she says.

Some scientists believe that limited, controlled experiments that mix potentially pluripotent human cells with a mouse embryo could be revealing, but other experts interviewed thought such experiments were ethically troubling and scientifically inconclusive. One attempt to mix human ES cells with a mouse blastocyst has shown that the cells could engraft but not persist10. Pluripotent human cells might fail to contribute to the mouse embryo because of species differences in cell growth rates and surface proteins, rather than lack of pluripotency. In any case, Landis says that experiments with chimaeric embryos are not being considered as part of the NIH criteria.

Some scientists working on non-embryonic sources of pluripotent stem cells would like to see more nuanced definitions of pluripotency. What irks them is that the word 'multipotent' can paint too narrow a picture of a cell's flexibility. Shahin Rafii at Weill Cornell Medical College in New York, who recently derived teratoma-forming stem cells from mouse testis11, believes that focusing on the strictest assessment of pluripotency in mice obscures what the human equivalents of such cells might be able to do. Along with his colleague Daylon James, lead author on the mouse–human chimaera paper10, Rafii thinks it might be useful to distinguish between a strict definition of 'embryonic pluripotency', which could be used to address issues of development, and 'therapeutic pluripotency', which indicates that a cell line can make all of the major tissue types. For those working on drug-screening technologies and cell therapies, he says, the lure of pluripotency is to get functional tissue. “Everyone wants a neat package,” says Rafii, “but all these cells have unique traits.”

Landis worries that multiple meanings of pluripotency would “degrade the definition”. Besides, the presidential order calls for lines deemed pluripotent to be listed on the same registry as embryonic lines. What most concerns her is putting cell lines on the registry that are in fact not pluripotent. This would create the impression that they are equivalent to human embryonic-derived stem cell lines, she says. Although the difficulty of pushing cells down a particular lineage may vary, Landis says, every cell line added to the registry should have the same potential to differentiate as the human ES cell lines already there.

But even researchers working with potentially pluripotent stem cells want their cells to be studied in addition to, not instead of, ES cells. Jaenisch is working on reprogramming human cells. One day, perhaps, these patient-derived cells might be just as good for studying disease or developing cell therapies. To get to that point, he says, scientists need to figure out the epigenetic 'ground state' of human ES cells. Older human ES cell lines won't work for such studies, Jaenisch says. Instead, newer lines should be derived under better conditions.

Even if that happens, there's more to embryology than generating a panoply of specialized cell types, says ISSCR's Daley. “There's an overlapping and distinct set of questions when you are studying embyronic stem cells versus induced pluripotent stem cells,” he says. “It's never going to be the case that induced pluripotent stem cells will allow you to ask all the same questions.”

Note to readers: When I said I was trying to figure out how pluripotency could be assessed in human cells, several of the experts contacted for this story had the same reaction: “Good luck with that one.” In a short, general article, nuance can be sanded away and important topics neglected altogether. Please send me your comments on what I've missed. I'd like to put them together as a follow-up. My email is m dot baker at naturesf dot com. MB