Making the Paper

Nature Reports Stem Cells
Published online: 27 September 2007 | doi:10.1038/stemcells.2007.94

Embryonic stem cells make their own niche

Monya Baker¹

Niches-in-dishes may yield clues to differentiation

Mick Bhatia

Mickie Bhatia wanted a straightforward answer to a practical question—how to save money on laboratory supplies. Instead, he uncovered a new way to think about how stem cells self-renew and how they can be guided into differentiation¹.

Scientists who grow human embryonic stem (hES) cells need to keep cultures well supplied with a suite of signaling proteins, including fibroblast growth factor (FGF). If levels of FGF are boosted 20-fold, however, the other proteins can be largely eliminated. Bhatia, head of McMaster University's Stem Cell and Cancer Research Institute, thought that figuring out why might reveal a recipe for less expensive media. Using proteomics, his group identified insulin growth factor (IGF) as an important supporting factor for ES cells. The role of IGF in hES cells had not been previously characterized, so the obvious next step, he thought, was to find the pathways through which IGF and FGF interact within a cell. Perhaps then he could find some common downstream factor that could be triggered to make cells grow better.

A literature search revealed few connections between the growth factors, so Bhatia's lab started blocking receptors for the growth factors to see what would happen. They saw unexpectedly disparate results. Blocking IGF receptors decreased the number of cells in culture. Blocking FGF receptors had no effect on the total number of cells but increased the percentage of cells that became the fibroblast-like cells known to arise spontaneously in hES cell cultures. Adding more FGF while blocking IGF receptors couldn't reverse the effect (total cell numbers still decreased); nor could adding more IGF when FGF receptors were blocked (more fibroblast-like cells still appeared).

After months of frustration, Bhatia's team began to realize that the growth factors were acting on distinct populations of cells. "Our brains didn't really go in that direction until our data shoved us this way," he recalls. Suddenly, puzzling observations resolved into predictable patterns. IGF and FGF each act on different cells, and then the two kinds of cells act on each other. By borrowing cell-sorting techniques usually used with blood and neural stem cells, they could separate the ES cells according to which receptor was on their surfaces; only cells expressing the IGF receptor were able to generate new colonies of hES cells. The cells responding to FGF, it seemed, acted on the fibroblast-like cells, which were in turn supplying the ES cells with the signals necessary to proliferate and self-renew. When the fibroblast cells were in short supply, the ES cells generated more of them.

Bhatia's team has plated individual ES cells and observed them through time-lapse photography. "Preliminary experiments suggest that hESCs produce fibroblasts through asymmetric divisions almost immediately," he says. "We've never seen them do without them." Work by Shinya Yamanaka and others has shown that mouse fibroblasts can be genetically engineered to behave like ES cells, so Bhatia is looking carefully to see how other fibroblasts differ from the ones made directly by ES cells in culture, and particularly whether ES cell–derived fibroblasts ever revert to ES cells.

These niches-in-dishes might explain some of the difficulties getting cells to differentiate from ES cells, says Bhatia. "We can generate millions of blood cells, but none of them are able to reconstitute the blood system of a mouse." Bhatia suspects that's because cells in culture never generate their in vivo progenitor, hematopoietic stem cells. "As a cell is differentiating in vivo it is constantly coming into contact with other cell types," he says. "In vitro they tend to be more pure, and I think that may be a problem." Bhatia fears that when ES cells are differentiated in culture, the cells go through only a "caricature" of the natural process. Simply shutting down the self-renewal system does not automatically open the door for cells to move down the desired lineage.

But perhaps more interesting is the possibility that, like ES cells, tissue-specific stem cells may generate more-specialized cells that support self-renewal in vitro. Some 20 other laboratories have contacted him to say that they've seen similar results in their stem cell systems, he says, and he's already started collaborations with several. Perhaps, like differentiating cells, science thrives best when scientists stay in contact with many others.

Author affiliation

1. Monya Baker is news editor of Nature Reports Stem Cells