Harvard Medical School researchers are following the path of cancer cell death using time lapse imaging and TRAIL, a naturally occurring protein that triggers cell death. This research could provide insights into future treatment strategies.

The study, published earlier this month in Nature, explored why some individual, genetically identical cells survive after the administration of TRAIL TRAIL is currently undergoing clinical trials as a potential cancer treatment drug.  The acronym is shorthand for a complicated description of a substance that induces cell death.

When TRAIL is administered, it attaches to the cell, which triggers a cascade of reactions in the cell activating more proteins and receptors. Pores in the cell membranes then open up and leak toxic proteins that had previously been locked up safely until TRAIL came along to activate the reaction.

“Because of that pore threshold, it’s like a light switch. You’re either off and alive or on and dead,” said Sabrina Spencer, a visiting research fellow in systems biology at Harvard Medical School.

The scientists tracked individual cell division and TRAIL-induced cell death and found that different protein levels may play a role in determining cell death.

“What had been previously observed but swept under the rug was that, when you administer drugs like TRAIL, there is always fraction of cells that don’t die,” Spencer said. This is why cancer therapeutics are never given only once, she added.

Spencer and her colleagues administered TRAIL to cells that survived the first round of battle, cells that were initially resistant. When the cells first divided, which is part of their natural behavior (like skin reproducing itself to heal a skinned knee), they reacted with the same resistance to TRAIL.

However, after the cells had some additional time to divide and differentiate, much like twin sisters tend to grow up and look slightly different, the cells no longer reacted the same to TRAIL.

“This was a key hint showing that differences in protein levels play an important role in deciding cell fate,” said Suzanne Gaudet, co-author of the study and assistant professor in the department of genetics at Harvard and the department of cancer biology at the Dana-Farber Cancer Institute in Boston.

Showing that protein levels may be behind this “fractional kill” of tumors in some cases provides an explanation that could lead to predictive cancer treatments in the future.

Marcus E. Peter, a professor at the Ben May Department for Cancer Research at the University of Chicago, was not involved in the study but said that there have been only two ways until now to explain the different response of genetically identical cells.

The Darwinian selection theory states that there is population of resistant cells and a population of sensitive cells. When killing the sensitive cells, the resistant cells remain. The other model that was considered specifically in the case of TRAIL, was that the resistance is actually caused by the response to the administration of TRAIL.

What this new study adds to the picture is a third possibility, Peter said.

“In addition to the selection, in addition to the response to the drug, there is a stochastic situation just by chance that cells have variant expressions of certain proteins, some of which actually render the cells resistant to the drug,” he said. “Obviously you can only find this by doing this single cell analysis.”

The authors have acknowledged that it will take a few years before they actually find out the impact of their results but the results offer a hopeful new direction.

“From our perspective, that’s definitely one direction we want to take,” Gaudet said, “to develop models that are going to be predictive and see if that can be extended to tumor samples, and develop enough knowledge in certain types of cancer cells, so we could see how well that will extend to tumor samples.”

It is not understood yet why certain tumors are resistant and others are not, Peter said.

“You can just view TRAIL as a drug, and the response of many cancers to drugs is always the same: the single biggest problem is that patients relapse and tumors develop resistance,” he said. But he added that the results of this study could have a tremendous impact on the way we design experiments to test new drugs.

It is possible that this could be used to pretest the efficiency of a treatment on a tumor in the future by measuring the reactions of a small sample of cells, according to Spencer and Gaudet.

“If you take the cells out and treat them with the drug the patient might get, you could then decide what to treat the patient with,” Spencer said.

However a limitation of these findings, Peter said, is that it is still difficult to tell what the contribution of these three models is for each drug, each cancer, and each cell type. There are multiple cell types that make up a tumor, different genetically defined cells carrying different mutations.

“These findings may apply to one group of cells, but what the impact of that is in the context of an animal tumor model or a real cancer over time in a patient, that’s hard to say right now,” Peter said.