The origins of cancer stem cells have long been elusive and controversial. Although some researchers think that cancer stem cells come from normal stem cells gone awry, others suspect that the cells arise from differentiated cells that have been reprogrammed. In support of the reprogramming theory, a tissue-culture study suggests that mutant mouse embryonic fibroblasts that lack retinoblastoma proteins are reprogrammed when they are cultured in suspension and come to resemble cancer stem cells.

The retinoblastoma 1 (RB1) pathway is crucial for detecting cell-cell contact and regulating cell cycle arrest. Consequently, whereas cultured mouse embryonic fibroblasts normally grow in a monolayer, mouse embryonic fibroblasts lacking RB1 grow into mounds of cells that eventually detach and form colonies of free-floating spheres. The RB1 pathway is also frequently lost in cancers. These and other observations set Douglas Dean of the University of Louisville Health Sciences Center in Kentucky, and his colleagues to wondering whether the spheres had a role in reprogramming differentiated cells into cancer stem cells.

Reporting in Cell Stem Cell, Dean and his colleagues show that when RB1−/− cells are forced to grow in spheres for two weeks, a new cellular morphology emerges1. Moreover, some cells from two-week-old spheres, but not from younger spheres or from the original monolayer, persistently express embryonic stem cell genes, including Oct4 and Nanog, as well as cancer stem cell markers. “It's like a switch has been flipped,” says Dean. Thus, he argues that the spheres might enable the reprogramming of cells into cancer stem cells. “There may be a technique down the road for producing an inducible pluripotent stem cell from a fibroblast using this technique,” he adds excitedly.

But Thea Tlsty, a pathologist at University of California, San Francisco, is not convinced for two reasons. The first is that in the two weeks it took for the new morphology and characteristics to arise, the culture setup might have selected stem cell-like cells that were already present in the population rather than enabling de novo reprogramming. Although Dean took precautions to minimize this possibility, he acknowledges that it has not yet been definitively ruled out.

The second reason is that Tlsty is uncertain about the claim that cells from aged spheres even have the characteristics of cancer stem cells. When Dean transplanted sphere-reared cells into mice, he observed tumour formation, supporting the notion that these cells have cancer stem cell capabilities. Yet this evidence isn't conclusive, explains Tlsty. “A cancer stem cell phenotype is usually accompanied by an invasive phenotype, and while I saw a growth in vivo, I did not see an examination of invasive phenotypes.”

“To clear up my confusion, I would want them to start with a single differentiated somatic cell,” says Tlsty. “And I would want to see some evidence of invasion potential.”

Dean readily notes another limitation of his study. “This is a cell culture–based set of studies,” he says. “What we're doing is trying to put forward a hypothesis that we and others can test in real tumours.”

Related articles

John Dick: careful assays for cancer stem cells

Cancer stem cells, becoming common

Cancer stem cell sightings and slightings