The Persistence of Memory ... in Reprogrammed Cells

Like a Texan who keeps his drawl after moving to California, adult cells reprogrammed to resemble embryonic cells retain some signatures of the tissue from which they came. That message, delivered in two studies published today, is both good news and bad news for researchers who hope to use so-called induced pluripotent stem cells (iPS cells) to study diseases and perhaps some day treat patients.

Techniques that reprogram cells have revolutionized the stem cell field. By turning on a several genes in adult cells, scientists can transform skin or blood cells into stem cells that can become every cell type in the body—without the ethical and practical complications of using embryos or oocytes. These iPS cells are already making it possible to study diseases in new ways, and they raise the hope of someday using a patient's own cells to treat disease. But as scientists study and use iPS cells, it has become increasingly clear that reprogrammed cells are not exactly like traditional stem cells derived from embryos.

Two new studies document one reason why: Adult cells maintain a memory of their original tissue even after being reprogrammed into iPS cells. George Daley of Children's Hospital Boston and his colleagues discovered this when they were studying the ability of different stem cell types to make blood. Despite multiple tries, they could not get iPS cells made from fibroblasts (connective tissue cells found in the skin) to make blood cells. When they made iPS cells from blood cells, however, the reprogrammed cells made plenty of blood. Further experiments showed that each iPS cell line had a different pattern of DNA methylation, a molecular coating of the DNA that turns genes on or off in the cell. The researchers report online today in Nature that iPS cells retain the methylation pattern typical of the cell type they came from. Even different types of blood cells make slightly different iPS cells, the researchers found.

In a related paper published online today in Nature Biotechnology, Konrad Hochedlinger of the Harvard Stem Cell Institute in Cambridge and his colleagues compared the gene expression patterns in mouse iPS cells derived from white blood cells, muscle precursor cells, immune system cells called B cells, and fibroblasts taken from tail tips. (They were careful to compare cells taken from genetically identical mice.) They found that they could identify the tissue origin of each iPS cell line by simply looking at the cells' gene expression—at least in freshly made iPS cells. But these differences disappeared over time. After 3 months, the gene-expression patterns in the various iPS cells were indistinguishable, the team found. It isn't clear exactly why the differences fade, Hochedlinger says, but it may be that the expression of embryonic genes is strengthened as the cells grow in culture, gradually overwriting the cells' old gene-expression patterns.

The observations add weight to the theory that transforming an adult cell's DNA into an embryonic state is a gradual reprogramming process, Hochedlinger says. And it suggests that stem cells derived from embryos should remain the primary reference for iPS cells when researchers want to compare how cells from diseased patients behave, says Nissim Benvenisty of Hebrew University of Jerusalem, who has studied differences between ES cells and iPS cells derived from carriers of fragile X syndrome. At the same time, the cell-memory effect "probably gives another dimension to [iPS cells'] usefulness," Benvenisty says. For example, he says, iPS cells made from fragile X syndrome patients can help scientists understand how the faulty methylation in that disease affects brain cell development.