The cancer-stem-cell theory holds that tumours are formed by a very small population of self-renewing cells. The idea is controversial, but developmental biologist Wei-Qiang Gao hasn't allowed that to hamper his lab's quest to find such cells in the prostate. He and his colleagues have achieved an important first step — identifying the normal adult stem cells responsible for generating prostate tissue in mice.

According to the popular cancer-stem-cell hypothesis, tumours are derived from previously normal stem cells that have accumulated mutations. And because, like normal stem cells, these cancer stem cells are long-lived and slow growing, they are not killed off by treatments that target rapidly dividing, non-stem cancer cells. Much has been written about the model, says Gao, a senior scientist at Genentech in South San Francisco, California, but conclusive evidence that such cells exist has been lacking.

“Our main interest is in identifying human-prostate-cancer stem cells,” explains Gao. But before he and his co-workers could do that, they needed to identify and characterize the normal prostate stem cells (PSCs) from which cancer stem cells — if they exist — might be derived.

Although three cell-surface proteins common to candidate PSCs had already been identified, they weren't specific enough to isolate only potential PSCs. Gao's group identified a fourth marker, CD117, that is specific to these cells by looking at a region of the prostate thought to be the stem-cell niche. Combined, the four markers form a PSC 'barcode' that can be used to sort out cells from prostate tissue. When these cells were transplanted into immunocompromised mice that would not reject them, the cells could generate prostate tissue that had the proper morphology and secreted prostate-specific proteins (see page 804).

But for a cell to be defined as a tissue stem cell, it must be able to give rise to all of the cell types within that tissue on its own. So the researchers repeated the transplant experiments with single cells from the barcode group. Individually, these cells were able to generate prostates.

“Scientifically, it is important to show whether such a stem cell exists within an organ. We show that conclusively,” says Gao. But, he adds, “regenerating the prostate might have limited application for patients”, considering that it is not a vital organ, it is not crucial to reproduction, and it can cause problems later in life — by, for example, becoming enlarged or cancerous.

Gao says that the project was driven largely by postdoc Kevin Leong. “Scientists at Genentech have the freedom to pursue basic-research questions because management believes it will eventually lead to innovation and exciting products,” says Gao, who has worked for the company for almost 16 years. In this case, the work resulted in identifying CD117 as a likely marker for human PSCs as well — which might be of help in the effort to differentiate and target prostate cancer stem cells.

Gao says that even if cancer stem cells turn out not to exist, cells will be identified that perhaps go by another name — such as 'cancer-initiating cells', 'chemotherapy-resistant cells' or 'more metastatic cells'. “Whether or not the cancer-stem-cell model is correct, tumour masses seem to be heterogeneous and traditional therapies will probably not be able to kill all tumour cells,” says Gao. “A therapy that specifically targets these subpopulations of cells is needed for more effective cancer treatment.”