Heterogeneity and Plasticity in Metastatic Breast Cancer

Breast cancer is the most common cancer amongst women. As for most solid tumours, breast cancer-related deaths are the result of metastasis, a succession of complex steps leading to the dissemination of cancer cells and colonization of distant organs. The heterogeneity and plasticity of cancer cells facilitate the selection of clones that withstand the harsh conditions encountered during their journey as circulating tumour cells and invading a foreign microenvironment. While the main steps of the metastatic cascade have been described, the precise cellular and molecular mechanisms underlying each of its steps remain poorly understood.
In my PhD studies, I aimed to identify metastasis drivers in breast cancer using in vivo models of metastasis and RNA-Seq. This overarching goal materialised into two different projects.
The first part of my work leveraged bulk RNA-sequencing of matched primary tumours and metastases in different sites to identify molecular processes potentially involved in metastasis. Our initial findings identified glucocorticoid signalling as promoting lung metastasis. NR3C1, also known as the glucocorticoid receptor (GR), mediates the effect of stress hormones. We showed that upon stimulation, GR controls processes involved in steps of the metastatic cascade. Mechanistically this control translates into an up-regulation of the kinase ROR1, a WNT5A receptor, modulating the Wnt and Hippo pathways. We showed how GR and ROR1 ablation decrease metastasis. Our studies highlight the importance of glucocorticoid signalling in triple-negative breast cancer and call for caution when using activators of GR such as Dexamethasone to combat the side effects of chemotherapy.
In the second part of my work, we compared matched primary tumours and lung metastases of different breast cancer PDX models using single-cell RNA-seq. This technique allowed us to assess potential mechanisms driving metastases at the level of single cells, probing biological processes occurring in specific subpopulations of cancer cells. Our data highlight the significant inter and intra-patient heterogeneity of breast cancer. We found that cell populations in primary tumours and lung macrometastases are transcriptionally similar, but that three main phenotypes are making up lesions in both sites. Each of these phenotypes appears to be engaged in the epithelial-to-mesenchymal transition (EMT), yet at varying levels, as indicated by canonical EMT and proliferation markers. We show that the top differentially expressed genes differentiating each of these phenotypes are controlled via partial EMT transcription factors (TFs). This finding argues for the importance of partial EMT in the metastatic process and sheds further light on the processes underlying the metastatic cascade.