Reprogramming cells make ES-like colonies Credit: Jaenisch lab

On a Friday night in Philadelphia, at the annual meeting of the International Society for Stem Cell Research (ISSCR), there was standing room only in the lecture hall. Researchers had crowded in hoping to learn bench tricks from Kyoto University's Shinya Yamanaka and others who have used genetic engineering to successfully reprogram differentiated cells to a state virtually identical to embryonic stem (ES) cells.

A collection of scientists who have reprogrammed cells to pluripotency walked through their recipes, and then ISSCR president and Harvard stem cell biologist George Daley asked a simple question: what are the minimum criteria to determine that a cell has been induced to pluripotency?

Since human induced pluripotent stem (iPS) cells were reported late last year, they have shifted the scientific landscape. While careful to acknowledge that there are some questions that can be answered only with embryos, prominent stem cell scientists, including Ian Wilmut (who led the first team to clone a mammal) and James Thomson (who led the first team to derive human ES cells) have predicted that researchers will gradually move away from ES cells towards iPS cells because creating the latter does not require eggs, embryos or great manual dexterity. And while stem cell scientists universally acknowledge the need for further study of various routes to reprogramming and fuller comparisons between iPS and ES cells, a walk through the poster session at the meeting showed which direction scientists were leaning. When I asked attendees for observations, more than one told me that anything on iPS cells was packed.

There is money around too. The iPS session I attended was sponsored by a little-known biotech company called iZumi Bio. California-based venture capitalist firm Kleiner Perkins Caufield & Byers has backed the new company, based in San Francisco, California, which has begun a collaboration on iPS cells with the Gladstone Institutes, a non-profit research institute in San Francisco. Deepak Srivastava, director of the Gladstone Institute of Cardiovascular Disease, has joined iZumi Bio's board. Previously, he successfully recruited reprogramming pioneer Yamanaka, into a specially created position at the Gladstone that allows Yamanaka to split time between the institute and Kyoto University. (See Reprogramming expert makes his lab in two countries)

Although iZumi Bio may be the first company to launch because of the promise of iPS cells, the research field is crowded. Besides Yamanaka's lab, more than a half-dozen others have reported reprogramming human cells. And non-academic labs that have done so might not be publicizing their work. With more and more labs jumping into the field, how can scientists know that other scientists (and journal editors) will trust results from the cells they have made? In short, Daley wanted to know, how can scientists know they have made iPS cells?

What criteria?

I think teratoma is a poor man's way to test ES cells, but ES cells don't have oncogenes in them. Shahin Rafii

For mice, there is an empirical answer: mix the supposedly reprogrammed cells with mice embryos, and see if they undergo the complex processes necessary to form sperm and egg in the resultant mice. For humans, the question is much more open, but one criterion is whether the cells form teratomas when injected into mice without immune systems. Ideally, these strange clumps will contain cells representing all the major tissue types (endoderm, mesoderm, ectoderm, and germ tissue). However, these assays can take months, and may not capture different degrees of reprogramming.

Daley, who has also reprogrammed human cells, asked what other tests should be performed. Could cells be tested for markers or function? What criteria shoud apply to human and mouse cells?

Junying Yu of the University of Wisconsin-Madison described expression of key antigens and a gene expression profile; Yamanaka added that the retroviral genes still necessary to make the cells needed to be silenced. Kathrin Plath of the University of California, Los Angeles, seemed to speak for the group when describing what the criteria should be: “I don't know,” she said, “but more than a teratoma.” Alex Meissner of the Broad Institute in Cambridge, Massachusetts, thought genome-wide screens examining epigenetic status would be useful; however, such sophisticated technology would not be available to many of the labs working to make iPS cells. And panellists also described partially reprogrammed states in which cells express pluripotency markers without turning off pathways used in differentiation.

Shahin Rafii, a physician-scientist at Weill Cornell Medical College in New York is studying whether human spermatagonial stem cells can be reprogrammed. He says that whether or not the cells can form teratomas may be unrelated to the purposes that scientists want to use the cells for, and won't reveal whether iPS cells harbor nasty surprises. “I think teratoma is a poor man's way to test ES cells, but ES cells don't have oncogenes in them,” he says. In fact, less aggressive but still pluripotent cells might not be able to form teratomas in mice. A better test, he says, is whether the prospective iPS cells form functional tissue that does not develop tumours.

After the session Daley said that the histopathology of naturally occurring teratomas can predict their clinical behaviour (the more differentiated, the less problematic), and that perhaps a similar grading system could be used to predict the behavior of reprogrammed cells. He hopes the ISSCR will pull together a working group to address the issue later this year.