How cancer hijacks cell death: a new view of metastasis

Cancer metastasis is the main reason patients succumb to their disease. Yet, understanding how tumours initiate metastasis has been a challenge because of the partially random nature of invasion and the difficulty to directly follow it.

Most studies suggest that metastasis stems from cells detaching from a primary tumour as they accumulate more mutations. However, studies from my lab on epithelial cell turnover suggest a very different model where cells invade independently of the primary tumour via a dysregulated cell death process.

“There is a lack of understanding as to how cell division and cell death are linked – largely because, generally, the two processes are studied in separate labs.”

Most solid tumours arise from epithelia and, like blood cells – the source of liquid tumours – epithelial cells turnover at some of the highest rates in the body. Easy to see then why it is considered likely that cells undergoing division and death more rapidly are more prone to developing cancer. To prevent carcinomas (epithelial cancers), it is critical that the number of epithelial cells dividing match those that are dying. There is, however, a lack of understanding as to how cell division and cell death are linked – largely because, generally, the two processes are studied in separate labs.

My lab has been attempting to bridge this gap and has discovered that these processes are linked mechanically in epithelia. If there are too few cells, the stretch that they experience rapidly activates division; too many and crowding causes them to die by a process known as cell extrusion.

We initially discovered epithelial cell extrusion seamlessly removes dying cells without creating any gaps in the epithelial cellular sheet. This, of course, allows the epithelium to maintain an intact barrier. Later, however, we found that the process also works on living cells. In fact, we found most epithelial cells die from being extruded whilst alive when they are surplus to requirements. Typically, live cells extrude in response to crowding to maintain constant cells numbers. Interestingly, it is known aggressive carcinomas lack signalling factors vital for extrusion, which goes some way to support the idea this process is critical for maintaining correct cell numbers.

Extrusion and cancer

While pancreatic and some lung and colon cancers directly downregulate an important extrusion receptor, others hijack the trafficking of its ligand. In all cases where extrusion is disrupted, three things happen instead: cells form chemotherapeutic-resistant masses; the epithelium becomes leaky and likely more prone to inflammation; and finally, cells extrude aberrantly in the wrong direction.

During normal cell turnover, cells extrude apically out of the epithelium into dead space to later die. However, when extrusion signalling is disrupted, cells instead extrude basally, back into the tissue the epithelia encases. This basal cell extrusion (BCE) could then enable cells to invade throughout the body and metastasize.

We have found that restoring normal, apical extrusion dramatically reduces both tumours and their metastases in a mouse model for pancreatic cancer, suggesting defective extrusion promotes metastasis. However, these models did not allow us to see if BCE might drive metastasis initiation. We needed a new model. As such we developed an embryonic zebrafish with transparent skin as a model for simple epithelia where we can directly follow all cellular movements.

Transparent fish

By expressing a driver mutation of metastatic cancer which also disrupts extrusion – in this case the KRAS mutation – with a GFP tag throughout the zebrafish skin, it was possible to induce carcinoma and demonstrate that cells with the mutation can invade by basal extrusion. Our ability to visualise the behaviour of transformed cells within the zebrafish skin also revealed other surprising findings.

“This finding lies in contrast to the prevailing model that cells metastasize from a primary mass.”

Firstly, our observations suggested cancerous cells can invade at sites remote from where they form primary masses. While masses of cells accumulated at regions where cells normally would have extruded, cells invaded by aberrant basal extrusion at sites where cells typically divide. While we had been focusing on the masses, individual cells had been invading at distant sites.

This finding lies in contrast to the prevailing model that cells metastasize from a primary mass. It suggests that metastases could develop in parallel with primary tumours, rather than resulting from them, which may change the way we diagnose and treat cancers.

Although cells with KRAS mutations can invade, migrate, divide, and enter the bloodstream, most die after a few days. However, if they also contain a common collaborating mutation in a cell death gene – p53 – they survive and form distant metastases throughout the body. These results suggest that we might be able to screen and ward off circulating transformed cells with the potential to cause metastatic disease using blood tests and, potentially, use less toxic treatments than current chemotherapies.

Another unexpected finding from zebrafish was that the mechanics of BCE caused cells to partially de-differentiate their epithelial characteristics and adopt new plasticity. As cells invaded by BCE we discovered they simultaneously pinched off their entire apical surface, containing epithelial contacts and other features which, in effect, cause the cells to act collectively as a sheet. Once they invaded and shed their old epithelial identities, they could migrate independently. This also causes a sort of rebirth, or de-differentiation, as these cells later adopt different markers and behaviours – like neurons or mesenchymal cells.

It is this ability of BCE to knock epithelial cells back to a form of multipotency which could also play a role in therapy resistance. As the newly multipotent cells go on to develop into different cell types, it’s conceivable drug resistance will follow.

Because embryonic zebrafish epidermis may be intrinsically more plastic than aged organ epithelia where carcinomas develop, it is not clear if the de-differentiation associated with BCE would promote new plasticity in human cancers. However, we do know that some types of particularly metastatic cancers are highly multipotent, suggesting this mechanism of epithelial-to-mesenchymal transition may be relevant in human cancers.

A conserved mechanism?

Our findings in zebrafish have led to a new view on how cancer cells might disseminate and even become resistant to therapies. Yet, to establish how this new invasion mechanism might impact human cancer, we need to learn if it is conserved in different organisms and organs.

Epithelia that coat different organs will have different architectures and mechanics, which could alter extrusion. Additionally, most carcinomas develop in older individuals, with more senescent cells and different mechanical properties. While we have identified a few oncogenic mutations that hijack extrusion signalling and lead to BCE, we now need to consider if oncogenic mutations do so. Determining if BCE promotes invasion in different species and epithelial types, and identifying signals and conditions that promote BCE will give us a better insight into whether this mechanism is at play in human cancers.

Identifying the signals and mechanics that drive BCE could provide clues for preventing invasion and, possibly even for targeting cells that have invaded by this mechanism. We have already determined that the commonly used anti-malarial drug, chloroquine, can divert basally extruding cells back apically to die.

If BCE does significantly contribute to metastasis, we can build on these findings to develop new approaches for preventing metastatic cancers.

Author
Jody Rosenblatt is Professor of Cell Biology within the The Randall Centre for Cell & Molecular Biophysics at Kings College. She holds a Cancer Research UK Programme Award.