In the late twentieth century, there was a rise in a type of cancer called mesothelioma, which is caused by exposure to asbestos used in building materials. Mesothelioma often arises decades after exposure, accounting for tens of thousands of deaths annually worldwide1. Even with the treatments currently available, it is inevitably fatal. There is therefore an urgent need to develop more effective therapies for this type of cancer. Writing in Nature, Wu et al.2 report that mutations in a cell-signalling pathway that commonly occur in mesothelioma create a tumour vulnerability that might be targeted to treat this disease.
Mesothelioma most often originates in the lining of the lungs, in cells that form the pleural membrane. Mutations frequently found in mesothelioma cells often inactivate proteins, called tumour suppressors, that function in anticancer pathways. One of the most common such inactivated proteins is called merlin (encoded by the NF2 gene), which functions in the highly evolutionarily conserved Hippo signalling pathway. This pathway was originally identified in the fruit fly Drosophila melanogaster3,4, and it comprises a signalling cascade that controls cell proliferation and organ size. If merlin or another protein in this pathway, such as LATS2, is inactivated, downstream proteins called YAP and TAZ can boost the expression of genes that promote tumour formation. Certain cancers can even become ‘addicted’ to YAP-mediated transcription for their survival5.
However, if merlin, LATS2 and another protein called LATS1 are functional, YAP and TAZ undergo phosphorylation (a phosphate group is attached to them), which modifies the proteins and blocks their function by preventing them from entering the nucleus to drive gene expression6. Mutations in the genes encoding merlin and LATS2 are positively selected during tumour development7, consistent with their normal roles as tumour-suppressor proteins in mesothelioma.
Wu and colleagues studied the gene-expression profiles of human cancer cells grown in vitro, and report that YAP and TAZ drive the expression of proteins, such as ACSL4, that are needed for a type of cell death called ferroptosis. The authors also uncovered a connection between the ability of cells to suppress ferroptosis and the cell–cell contact that depends on the protein E-cadherin. The authors report that high expression of E-cadherin in human mesothelioma cells grown in vitro is associated with resistance to ferroptosis. E-cadherin activates the Hippo pathway, and the authors went on to explore the relationship between this pathway and ferroptosis.
Cell death that occurs through ferroptosis depends on a reaction between cellular iron and hydrogen peroxide8. During ferroptosis, a polyunsaturated fatty acid — a type of lipid found in the cell membrane — undergoes a modification called peroxidation, which causes an increase in the level of molecules termed reactive oxygen species. Ferroptosis is often linked to depletion of the amino acid cysteine, which is imported into cells by the protein SLC7A11. Cysteine provides a building block for the production of glutathione, a molecule involved in a pathway that can combat ferroptosis.
The drug sorafenib is approved for clinical use. It can induce ferroptosis by inhibiting SLC7A11. The authors demonstrate that sorafenib treatment of cultured human mesothelioma cells that have mutations in the gene encoding merlin causes the cells to undergo ferroptosis. They report that this sensitivity to ferroptosis depends on YAP- and TAZ-mediated gene expression (Fig. 1).
Two independent clinical trials9,10 found that sorafenib caused tumour shrinkage or stabilization in people with mesothelioma. However, neither trial evaluated the mutations present in the patients’ tumours, and it is tempting to speculate that the tumours of people who responded particularly well had mutations that inhibited the Hippo signalling pathway and that thereby boosted YAP- and TAZ-mediated gene expression.
Might other mutations beyond those in the Hippo pathway also regulate ferroptosis in mesothelioma? The most commonly mutated gene in this cancer11 encodes the tumour-suppressor protein BAP1. This enzyme affects gene expression, and can cause a reduction in the expression of SLC7A11, which, in turn, leads to ferroptosis12. If the gene that encodes BAP1 is mutated, ferroptosis does not occur12. Therefore, the presence of wild-type BAP1 might help to enhance ferroptosis, along with any boost to ferroptosis provided by the use of SLC7A11 inhibitors. It is not known whether drugs that induce ferroptosis, such as sorafenib, would be effective in cells in which mutations inactivate BAP1.
Other approaches to targeting mesothelioma in which the Hippo pathway is inactivated are being explored. For example, in animal studies, loss of merlin expression is associated with cancer-cell vulnerability to inhibition of a protein called focal adhesion kinase13. However, no clinical benefit was found with this approach in a clinical trial14. Direct targeting of the interaction between YAP and TEAD, a protein to which YAP binds when it drives gene expression, is another strategy being pursued to block cancer-promoting gene expression15. Finally, YAP and TAZ recruit the protein BRD4 to drive the expression of specific genes, and use of a small-molecule inhibitor to target BRD4 can disrupt YAP- and TAZ-mediated gene expression16. This class of small-molecule inhibitor is entering early clinical trials. All of these approaches aim to block YAP- and TAZ-mediated gene expression. However, if the anticancer strategy being used aimed to trigger ferroptosis in mesothelioma cells, then YAP- and TAZ-mediated gene expression would be required.
Identifying a tumour that has an inactivated Hippo signalling pathway as a means of a developing personalized cancer therapy — the ultimate goal — poses some challenges for mesothelioma. Focusing only on tumours that have lost merlin function would probably miss mesotheliomas in which Hippo signalling is inhibited by inactivation of other proteins, such as LATS1 and LATS2. A previous study17 of the Hippo pathway in various cancers has revealed that 22 genes are commonly transcribed by YAP and TAZ, and this transcriptional profile might offer a way to identify ferroptosis-sensitive tumours. Furthermore, because this profile was found17 in several types of tumour, triggering ferroptosis might be worth exploring for cancers other than mesothelioma.
Wu and colleagues’ report highlights a strategy that could offer a way of developing a personally tailored anticancer therapy. However, therapies targeted to mutations in an individual’s mesothelioma are still in their infancy. Clinical trials that take this approach, for example the mesothelioma stratified therapy trial in which I am involved (see go.nature.com/2o19lah), might help to make progress in such endeavours, and provide improved treatments at a time of unmet clinical need.
Nature 572, 314-315 (2019)
Competing Financial Interests
D.F. has received consultation fees from Inventiva.