Abstract and Introduction
The search for agents that target and kill cancer stem cells is on. In August, researchers reported in Cell that high-throughput screening could be used to identify drugs that target these cells, notorious for their resistance to existing treatments and their putative ability to generate new tumors.
Killing cancer stem cells may be the key to preventing cancer's recurrence, say researchers who have confidence in the cancer stem cell model of carcinogenesis. The model holds that only a specific subset of cancer cells can give rise to new cells or metastases, and some say this explains why cancers can seem to disappear after treatment and then recur with a vengeance. However, there is a debate about how relevant such cells are to cancer treatment and even whether they exist (see related article, News, J. Natl. Cancer Inst. 2009;101:546–7). Those who believe that the model oversimplifies the complexity of human cancers are wary of this study's results.
The successful use of high-throughput screening in finding the agents depended on developing a stable, in vitro stem cell culture, said Tamer Onder, Ph.D., who co-led the work with fellow researcher Piyush Gupta, Ph.D. of the Massachusetts Institute of Technology, Cambridge, Mass. "It's hard to maintain the cancer stem cell state in culture—that's why it's been hard to find therapies that target them," said Onder, now at Children's Hospital Boston. "Stem cells isolated from patients either do not grow in culture, or when they do, they lose their characteristics."
Onder said the key to the culture was to induce the cells to undergo an epithelial–mesenchymal transition (EMT), which the team accomplished by inhibiting the cells' expression of E-cadherin. (Although the link is not completely understood, induction of the EMT activates the same transcription factors that give cancer cells the motility, ability for self-renewal, and resistance to apoptosis that marks cancer stem cells.) The resulting cells displayed three characteristics of cancer stem cells: They could form mammospheres in culture, had two known genetic markers of stem cells (high expression of CD44 and low expression of CD24), and could generate new tumors when injected into mice at a much lower dilution than could the control cells.
About 16,000 compounds were then screened for their ability to kill the stem cells at a greater rate than the control cancer cells. The screen turned up 32 such compounds. The researchers winnowed the results down to the most promising and focused on one called salinomycin. Cancer stem cells treated with salinomycin were much less able to form new tumors when injected into mice. Further, mice previously injected with other human breast cancer cells and treated with salinomycin had tumors that were smaller and contained fewer cancer stem cells than mice treated with a control drug (paclitaxel).
"We were happy to see that the tumors shrank and had fewer cancer stem cells," said Onder. "This was the proof of principle for us."