A common technique used to generate tumor cells for cancer research could produce misleading and unrepresentative results, a new study shows. Because cancer cells taken directly from patients are scarce, researchers often culture these human-derived cells inside successive generations of mice. They then study such “passaged” cells to see how they behave and how the disease might respond to treatment.
The new findings highlight a serious limitation with this procedure: human cancer cells passaged in this way end up very different from the cells in the original tumor samples. Specifically, such serial transplantation leaves behind less than 1 percent of the cells present in the human disease. Studying such unrepresentative cell populations raises doubts about the extent to which the results can be applied back to people.
“Transplantation into mice is the number one preclinical model that people use. And that model itself induces some selection that people really hadn’t appreciated,” says Akil Merchant, a study coauthor and physician-scientist who works on leukemia at the Cedars Sinai Medical Center in Los Angeles. “Models are always limited. We know that, but now we have some insights into some of the very specific ways they are limited.”
The findings emerged from a study of cell heterogeneity in leukemia. “The heterogeneity of the cancer cells develops with time, so you are trying to treat a moving target,” explains Rong Lu, a biologist at the University of Southern California and coauthor on the paper. “We want to understand how these individual cells compete with each other and how they are different in terms of their growth, their metastasis, and also in terms of their response to chemotherapy treatment.”
Using a technique to read the gene expression of individual cells, and a DNA-based labelling system to “barcode” the cells, the study authors tracked the migration and proliferation of individual human cancer cells after they were transplanted into mice. This is important because cancer researchers use mice to mimic the human body to study the disease and to analyze the effects of treatments such as chemotherapy.
The study allowed the scientists to compare the cancer cells before and after they were transplanted into the mice, to follow cells as they spread around the animals’ bodies, and to track how the cells expanded their populations and evolved through successive generations. They also delivered chemotherapy drugs to the mice and watched what happened to the cells in response. The results were published November 11 in Nature Communications.
Tracking the cell types when the mice received the chemotherapy showed distinct patterns of gene expression in cells that responded to the treatment.
Different types of cancer cells tended to accumulate in different organs.
To the surprise of the researchers, they also found that different types of cancer cells tended to accumulate in different organs. Cells with high expression of the CMC2 gene, which is involved in cancer metastasis, colonized the ovaries, while cells with low expression of CMC2 were found mostly in the blood and spleen. Other types of cells, including those with high expression of the genes BTK, DNAJC, and LRIF1, accumulated in patches of the bone marrow. When the researchers deactivated these genes, the leukemia cells lost their ability to stick to other cells in the bone marrow, and migrated away.
“This tumor is in circulation. It’s liquid and it’s moving around, yet there was clearly spatial heterogeneity. That was a remarkable finding,” says Merchant.
That spatial pattern could signal further bad news about the reliability of working with cancer cells, he says. If a similar anatomical spread happens in leukemia patients, then taking a single sample from a single site will miss many cell types. In addition to introducing bias to research, the uneven distribution of cancer cells in bone marrow could mean that standard biopsies of this tissue may not accurately diagnose the disease.
The study produced some “really interesting findings,” says Long Nguyen, an oncologist at the University of Cambridge who was not involved with the research, such as showing that a different type of cells tended to dominate after the mice had received the chemotherapy compared with before the treatment.
One explanation for this is that the previously-dominant cells were killed by the drugs, presenting other cell types with the opportunity to grow to fill the niche. Some cancer experts think tumors contain undifferentiated stem cells that can go on to form these new types of cancer cells.
“I think the implication is that there are possibly cancer stem cells that remain dormant but are activated after chemotherapy. I think this has big implications for why there is relapse following chemotherapy treatments,” Nguyen says.