Tracking the Transition
A triple-transgenic mouse model allows researchers to trace the lineage of EMT tumor cells, and reveals the transition’s surprising lack of significance in metastasis
At Weill Cornell Medical College in New York City, Dingcheng Gao’s research group studies cell and developmental biology. Recently, he and his colleagues published a paper outlining their research into EMT (1). They identified a key difficulty with our understanding to date: namely, that there’s no way to track transient and reversible EMT phenotypes in living organisms. Without that ability, we can’t find out whether or not cells are indeed undergoing EMT to initiate metastasis, then undergoing MET to return to an epithelial phenotype.
So Gao and his colleagues generated a triple-transgenic mouse model known as MMTV-PyMT/Rosa26-RFP-GFP/Fsp1-cre, or tri-PyMT. The mouse has three special attributes: an oncogene (PyMT or, in some cases, Neu) driven by the MMTV promoter; a recombinase (Cre) driven by the mesenchymal-specific Fsp1 promoter; and two fluorescent proteins, red and green, each under separate control. The fluorescent proteins combine to form an irreversible color switch system – so once a cell has undergone EMT (and acquired green fluorescence), it’s incapable of reverting to red fluorescence. This means that it’s easy to see which cells have made the transition from epithelial to mesenchymal, even after they have transitioned back to epithelial characteristics.
“We wanted to find direct evidence in vivo to prove the EMT/MET hypothesis in metastasis formation,” explains Gao. “Therefore, we established the EMT lineage tracing model using a permanent fluorescent marker switch to trace the reversible EMT process.” But the team were in for a surprise. The cancer cells of the mice, which developed primary breast tumors followed by spontaneous lung metastases, didn’t show the expected results. In fact, they showed exactly the opposite: none of the secondary lesions changed color following the natural progression of lung metastasis. The lack of color switching indicates that Fsp1, the mesenchymal promoter designed to permit green fluorescence, was never activated – and thus, that the metastatic cells may never have undergone EMT. Furthermore, inhibiting EMT with the use of the microRNA miR-200 prevented red-to-green color switching – but had no effect on the ability of tumor cells to metastasize.
“Cancer cells are capable of metastasizing through other mechanisms, such as collective invasion and random dissemination,” says Gao. He cites a recent report by Cheung et al. in which the authors traced the lineage of metastatic tumors and showed that seeding by cell clusters, rather than by single cells, can result in polyclonal metastases (2). Collective invasion is typical of carcinomas like those often found in the breast or lung, and challenges the belief that metastases arise from single “escaped” tumor cells that undergo EMT. But if EMT isn’t the key player in cancer dissemination, then what is?
“We’ve observed that EMT is a relatively rare event in primary tumors,” says Gao. “Even though EMT tumor cells gain some anti-apoptosis properties that may help them survive in circulation, these advantages are accompanied by a downside – a decreased ability to proliferate. In general, metastasis is a very inefficient process for tumor cells. In our experiments, the rare cells that had undergone EMT were easily outnumbered by the epithelial cells, not just in the primary tumor, but also in the circulation and metastatic lesions.” So if EMT is costly for tumor cells and most metastatic cells show no evidence of needing to undergo it despite its potential survival advantage, what is its purpose in the tumor?
The second part of the Cornell paper offers an answer. Evidence from previous studies has suggested a link between EMT and chemoresistance – most notably in residual breast cancer, where the remaining tumor cells display mesenchymal characteristics (2). Gao and his team decided to investigate this link by treating their tri-PyMT mouse models with cyclophosphamide. Even during the initial treatment phase, green fluorescent (mesenchymal) cells were less proliferative – but also less apoptotic – than epithelial cells, indicating lower susceptibility to chemotherapy. But in metastatic lung tumors, the effect stood out even more. The mesenchymal cells outnumbered the epithelial population by almost three to one, and made notable contributions to five of the 17 total metastatic lesions (in contrast to untreated mice, where no lesions contained a significant mesenchymal cell population). “Post-EMT tumor cells showed a greater ability to survive chemo treatment,” Gao summarizes. “This won them a better chance to develop into metastatic lesions.”
“Our results suggest that tumor cells that undergo EMT are more resistant to chemotherapy than non-EMT cells. More importantly, we have observed a significant contribution of these EMT tumor cells to metastasis formation under chemotherapy conditions. Therefore, targeting EMT tumor cells may provide novel therapeutic approaches to overcome chemoresistant metastasis.” Gao thinks this is a vital piece of knowledge in the clinic. “Given that most patients with advanced-stage tumors are treated with chemotherapy, it’s important to evaluate the EMT status of their tumors. Patients whose cells have undergone the transition would benefit from EMT-targeting therapy approaches.” Not every cancer recurrence is based on EMT, though, and we now know that most metastases may not be rooted in that transition, either. As a result, pathologists need to closely examine tumor characteristics on a case-by-case basis to determine how best to treat and monitor each patient.
Of course, there’s much still to be learned about the nature of metastasis. “One immediately attractive question,” says Gao, “is whether the metastatic epithelial tumor cells differ in other characteristics from the majority of cells in the primary tumor. Characteristics like CK14 expression, multiple clonality, and other potential mechanisms in metastasis need to be further investigated.” For his part, he and his laboratory are currently focused on developing novel strategies for targeting EMT tumor cells, with the hope of one day finding a way to overcome cancer chemoresistance.
- KR Fischer et al., “Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance”, Nature, 527, 472–476 (2015). PMID: 26560033.
- KJ Cheung et al., "Polyclonal breast cancer metastases arise from collective dissemination of keratin 14-expressing tumor cell clusters", Proc Natl Acad Sci USA, 113, E854–E863 (2016). PMID: 26831077.
- CJ Creighton et al., “Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features”, Proc Natl Acad Sci USA, 106, 13820–13825 (2009). PMID: 19666588.
Dingcheng Gao is Assistant Professor of Cell and Developmental Biology in Cardiothoracic Surgery at Weill Cornell Medical College, New York, USA.
