Abstract
CDH1 is a well characterised tumour suppressor gene, and CDH1 dysfunction can result in abnormal cell differentiation, proliferation, and tumour invasiveness. Germline inactivating variants of CDH1 are causative of hereditary diffuse gastric cancer (HDGC), a cancer syndrome characterised by severely increased risk of developing both diffuse-type gastric cancer (DGC) and lobular breast cancer, with lifetime risks of 70% and 40% respectively. Current treatment options for HDGC are limited, and endoscopic surveillance often fails to detect early-stage DGC. As a result, prophylactic total gastrectomy is the recommended treatment for carriers of pathogenic CDH1 variants. However, this procedure causes serious long-term morbidity, and warrants the development of a chemoprevention strategy for HDGC.
Because loss of function mutations are difficult to target therapeutically, due to the absence of an obvious druggable target, we have taken a synthetic lethal approach to identify vulnerabilities in CDH1-null cells. Synthetic lethality is a gene-gene relationship in which loss of function in either gene does not affect cell viability, but simultaneous loss of both genes results in cell death. Synthetic lethal interactions can be perceived as cell vulnerabilities that occur following the inactivation of one gene, which can be exploited by targeting synthetic lethal genes to induce cell death. In the context of HDGC, synthetic lethal partners of CDH1 can be considered actionable targets for cancer chemoprevention.
In this study, I performed drug screening in an isogenic pair of non-tumourigenic epithelial breast cell lines (MCF10A), with and without CDH1 expression. Relative to wildtype MCF10A cells, MCF10A CDH1-/- cells exhibited elevated sensitivity to inhibitors of sphingolipid metabolism, endocytosis, and autophagy.
To further validate candidate synthetic lethal compounds, two distinct gastric organoid systems were established from mice harbouring an inducible Cdh1 knockout system. One model was derived from neonatal murine gastric tissue cultured under air-liquid interface conditions, and the other was derived from adult murine gastric stem cells, with each system presenting unique advantages and disadvantages. The adult stem cell-derived organoid model of HDGC was optimised for drug screening with candidate synthetic lethal compounds. These compounds comprised the sphingosine kinase 1 inhibitor PF-543, the endocytosis inhibitors chlorpromazine and PP1, and the autophagy inhibitor chloroquine. Treatment with all four candidate compounds induced synthetic lethality in gastric organoids, validating the results observed in isogenic MCF10A cells.
Collectively, these results suggest that membrane organisation and vesicle trafficking represent a previously unrecognised druggable vulnerability in CDH1-null cells, and may lead to the development of new therapies for HDGC.