Oncology Abstracts

Breast cancer is a heterogeneous disease that can be classified into a number of molecular subtypes that predict prognosis and influence clinical treatment. Cellular heterogeneity is also evident within breast cancers and plays a key role in their development, evolution and metastatic progression. How clinical heterogeneity relates to cellular heterogeneity is poorly understood. We have approached this question using single-cell RNA-Seq on 1000s of individual cells from well-established in vitro and in vivo models, as well as clinical samples of Estrogen Receptor positive (ER+) breast cancer. Supervised and unsupervised approaches have identified cellular populations with transcriptional signatures of diverse cancer associated phenotypes, including proliferation, hypoxia and treatment resistance. In particular, distinct sub-populations of cells with a heterogeneous mix of molecular subtypes and signatures suggesting innate resistance to endocrine therapies have been identified. Gene regulatory networks were then used to identify transcription factor regulons that are active in individual cells, leading us to identify potential transcriptional drivers (such as: KLF5 and E2F7) of the putative endocrine resistant cells. This approach has been extended into a number of clinical ER+ breast cancers, highlighting a complex ecosystem of tumour-associated cells and identified a heterogeneous mix of epithelial cells expressing transcriptional markers of both luminal and basal cells. This is a somewhat confounding finding in ER+ breast cancers and highlights the potential power of single-cell approaches to identify specific cellular populations that could contribute to malignancy or relapse following treatment. Overall, our results suggest a high degree of cellular heterogeneity within breast cancers that can be functionally dissected into sub-populations with transcriptional phenotypes of potential clinical relevance. In particular, the identification of cells associated with treatment resistance hints at ways in which single-cell genomics could be used to predict and track variable treatment response and resistance during breast cancer treatment.