When cancer spreads from its primary site to become established at a secondary location, this process, termed metastasis, is often fatal. The timing of metastasis initiation varies depending on the size, stage and differentiation status of the tumour at the primary site1. Metastasis probably requires cancer cells to undergo changes, including those that promote the acquisition of invasive properties. Loss of expression of the protein E-cadherin enables cells to migrate, but how E-cadherin fulfils its role as a central regulator of metastasis isn’t fully understood2. Writing in Nature, Padmanaban et al.3 report that E-cadherin contributes to an aspect of metastasis that differs from the protein’s previously known effect on cell invasion and migration.
E-cadherin is present on the membrane of epithelial cells, which form a barrier layer on surfaces of the body. When a process called epithelial-to-mesenchymal transition (EMT) is triggered, loss of E-cadherin occurs, and epithelial cells acquire the characteristics of mesenchymal cells, which are highly mobile. EMT is usually activated by specific stimuli4, such as exposure to the signalling protein TGF-β. The process occurs in embryonic development5, and aids the repositioning of normal epithelial cells within organs during healing6. However, EMT can be hijacked for the development and spread of cancer7–11.
Adhesion between epithelial cells is mediated by E-cadherin, leading to the view that such adhesion suppresses metastasis12. Yet, counter-intuitively, some pieces of the puzzle do not fit this model. There is compelling clinical evidence that metastatic cancer cells commonly express E-cadherin and molecules associated with epithelial-cell fate13. For example, E-cadherin is found in metastatic cells in a type of breast cancer called invasive ductal breast carcinoma14,15. One idea to reconcile this discrepancy is if metastatic cancer cells that have undergone EMT and reached a secondary site then undergo a reversal process called mesenchymal-to-epithelial transition16. The presence of E-cadherin in cells at a secondary site does not then necessarily indicate that the protein helped metastatic cells to become established there.
Padmanaban and colleagues investigated E-cadherin’s role in a broad spectrum of mouse models of different types of invasive ductal breast carcinoma, and also analysed human cancer cells that were introduced into the mouse models. The authors engineered mouse or human cancer cells so that E-cadherin expression could be lowered or blocked and its expression tracked by monitoring fluorescent proteins. Cancer cells that expressed E-cadherin displayed less migratory behaviour in vitro than did those that did not express it, consistent with previous findings.
When the authors studied the behaviour of human cancer cells transplanted into mice, they found that, unexpectedly, cells expressing E-cadherin were more common than those lacking E-cadherin in primary tumours. This was also the case in tumour cells that had escaped the primary site, termed circulating tumour cells (CTCs), and in tumours that had metastasized. And when tumour cells were implanted in the animals’ mammary gland or injected into the bloodstream, those that expressed E-cadherin became established at a secondary site, whereas those lacking E-cadherin rarely did so (Fig. 1). This is surprising, because E-cadherin has not previously been shown to have a role in aiding the survival of metastatic cancer cells.
The authors found that, compared with E-cadherin-expressing cancer cells, cells lacking E-cadherin had higher levels of expression of genes associated with a type of cell death called apoptosis, and of genes that function in stress-related pathways. TGF-β and molecules known as reactive oxygen species (ROS) contributed to the upregulation of these pathways. Padmanaban et al. carried out in vitro and in vivo experiments in which they used inhibitors to target TGF-β, ROS or components required for apoptosis, and found that this treatment counteracted the effects of E-cadherin loss in cancer cells. Pathways involving TGF-β and ROS are known to be important in triggering apoptosis in invasive ductal breast carcinoma cells that have low levels of E-cadherin expression as they start to undergo EMT14,17.
Padmanaban and colleagues have uncovered a way in which E-cadherin functions in a context-dependent manner to promote tumour progression and metastasis by helping metastatic cells to overcome cellular stress mediated by TGF-β and ROS. E-cadherin loss compromised the metastatic potential of the cells by affecting cell survival and thereby impairing tumour establishment and cell proliferation at a secondary site. Thus, with regard to the metastasis of invasive ductal breast carcinoma, the pro-survival contribution of E-cadherin outweighs the advantage of E-cadherin loss boosting invasiveness.
A future research direction worth pursuing would be to determine whether there are any differences in the expression of the gene that encodes E-cadherin in cells from primary tumours, CTCs and metastatic sites. During EMT, cells are thought to go through distinct states18, but the cell-fate transitions that occur in tumours undergoing EMT are still unknown, and might vary depending on the tumour type. It is therefore unclear whether invasive ductal breast carcinoma cells that express E-cadherin, even at low levels, are in an EMT state or are a specific cellular lineage that is not undergoing EMT. Single-cell RNA sequencing could shed light on this by revealing whether there are distinct cell populations (clones) in primary tumours or metastatic cells that do not show signs of transitioning through EMT states.
Collective dissemination, in which different types of cell migrate together in a cluster, boosts a tumour’s ability to colonize distant sites19. If such tumour co-dependencies occur between stress-resistant cells that have high levels of E-cadherin and invasive cells that have low E-cadherin levels, collective dissemination might aid metastasis after a person’s tumour has been treated by, for example, chemotherapy, causing cellular stress. Examining the patterns of such tumour evolution, as well as analysing the mechanisms of therapy failure and pathways of cancer-cell growth, particularly of treatment-resistant cancer cells, will be crucial for the development of new clinical targets, treatments and therapeutic windows of opportunity. Whether E-cadherin expression has a key role in the survival of different types of tumour or in different forms of metastasis should also be explored.
The general factors that affect tumour growth provide clues to why cellular adaptations occur during metastasis. In-depth analyses are nevertheless needed, because the genetic background of a given type of tumour might affect the requirements for metastasis to occur. Padmanaban et al. have shown how E-cadherin is needed for metastasis of invasive ductal carcinoma, but other types of cancer might use alternative mechanisms to manage stress, perhaps generating different tumour vulnerabilities that could be exploited therapeutically.
It would be better to prevent metastasis than to have to treat cells that have metastasized. Understanding how E-cadherin expression is stabilized might reveal a vulnerability of cancer cells on a path towards metastasis. Developing tailored treatments to tackle or prevent metastasis should be a goal of cancer research, and we are headed in the right direction to make progress on this front.
Nature 573, 353-354 (2019)