A new technique can simultaneously assay a hundred molecular markers on single cells, allowing a team of researchers to track the developmental history of individual cancer cells.
The results, presented at the annual meeting of the American Association for Cancer Research on 4 April, could be used to bring order to the heterogeneous “mess” of different cells within a tumour, says Gary Nolan, who studies immune system signaling at Stanford University in California. “’Heterogeneity’ is the new buzz word in cancer,” he says. “But it’s useless if you’re not going to try and bring some order to that heterogeneity.”
Understanding that mess may be critical for understanding a tumour’s response to chemotherapy. Even tiny populations of drug-resistant cells are sometimes all it takes to reseed a tumour after therapy seemed to have eradicated it.
One way to differentiate among cells in a tumour is to track the proteins they express. Normally, researchers use a technique called fluorescence-activated cell sorting to separate and count cells. The proteins are labeled with fluorescent tags, and a cell sorter can distinguish up to about 17 different tags in a single run.
But Nolan felt he needed to go much higher than 17 if his lab was to tackle complex cell signaling networks.
So he teamed up with University of Toronto mass spectrometrist Scott Tanner who had developed a new method for sorting cells. Instead of fluorescent tags, Nolan’s team labels protein-binding compounds with stable metal isotopes. Samples are shot through a 7500K argon plasma flame, and ions — particularly the heavy metals used as labels — are sucked into a ‘time-of-flight’ mass spectrometer, which identifies the ions based on the time it takes for them to reach a detector.
The machines are made by a company called DVS Sciences. Nolan, who now serves as an adviser for the company, says he bought “the first two machines off the rack”. Thus far, his team has used the instruments to measure about 34 different markers on a cell, but he says the technique can handle up to 100.
For now, Nolan is using the method to create a family tree of the cells within the tumour. His team traces the lineage of individual cells as different proteins are gradually switched on and off during maturation from a cancer stem cell to a mature tumour cell. The researchers then look at how treatment with drugs or immune system proteins perturbs these lineages. “We can see that there are structured lineages within the cancers, and each of those lineages reacts differently to drugs,” says Nolan.
That information can be powerful for drug developers. “It doesn’t necessarily help me to know that I’ve got two drugs, each of which kills 50% of the tumour cells,” he explains. The second drug might never be approved because it performs no better than the first. “But if my measure is that I’m acting on the half of the tumour that the first drug missed, that’s important.”
The machines have facilitated the kind of single-cell network analysis Nolan advocated in a recent commentary in Nature Chemical Biology. The article, entitled ‘What’s wrong with drug screening today’, was not particularly diplomatic, Nolan admits. “I’ve had cancer six times,” he says. “I’m in no mood for being diplomatic. People are dying and we’ve got work to do.”