Combination is key in triple negative breast cancer

In 2019 SPECIFICANCER was funded by Cancer Research UK and The Mark Foundation for Cancer Research, through Cancer Grand Challenges, to tackle the tissue specificity challenge. The team set out to dissect the mechanisms underlying the tissue-specific manifestation of cancer-causing mutations, in the hopes of developing therapeutic interventions specifically tailored to tumour type. And now the team have done exactly that, finding a combination of inhibitors, targeting both epigenetic and oncogenic pathways, drives differentiation to enable hijacking of tissue-specific cell death pathways in TNBC.

Towards addressing the tissue specificity challenge, Karen Cichowski’s and Kristian Helin’s labs focused on understanding why the major epigenetic regulator Polycomb Repressive Complex 2 (PRC2) is oncogenic in some tissues but tumour suppressive in others. The team reasoned that as TNBC is a particularly undifferentiated tumour type, with most tumours being in a basal like state, that perhaps if they were able to reprogram cell state, the tumours might become sensitive to more standard targeted agents. Amy Schade, first author of the Nature paper, joined Karen’s lab as a post doc in 2019. She explains the rationale behind the team’s thinking, ‘We know that while the PI3K/AKT pathway is activated in the majority of TNBC's, inhibitors of PI3K/AKT are ineffective. But they are effective in hormone receptor positive (HR+) luminal breast cancer and in fact they're approved in combination with an oestrogen receptor antagonist.’

These concepts hold true in other tumour types, with the SPECIFICANCER team already publishing findings from colorectal cancer in Cancer Discovery earlier this year. However, in the colon EZH2 inhibitors cooperate with RAS pathway inhibitors, as opposed to AKT inhibitors, to suppress Wnt signalling and drive differentiation from a stem-like to an epithelial state before inducing cell death. In other words, the PRC2 complex cooperatively functions with the critical tissue-specific oncogenes in different tumour types to regulate pathways that are important for that particular tissue, and changes in cell state can be exploited across tissues to enable cell death. The work in colon was carried out by Patrick Loi and Amy in Karen’s lab in parallel to the work in TNBC, allowing the co-first authors to see patterns emerging in the different tissues. Amy explains, ‘Although there are clearly different responses in different tissues, and different driving oncogenes, there appears to be a paradigm as to how EZH2 inhibitors contribute to therapeutic responses. EZH2 inhibitors relieve the repression of critical genes, but you need the additional stimulus of the inhibitor of the oncogenic kinase to fully activate transcription and the therapeutic response. And in both cases we found that this converged on the pro-apoptotic protein BMF, which we think is serving as a cell state and/or stress sensor that triggers cell death in response to these drug combinations.’

Karen highlights that in the colon the combination of epigenetic and oncogenic inhibitors may also be co-opting a normal tissue-specific mechanism of cell death, ‘It's actually thought that as cells differentiate and then migrate up and out of the intestinal crypts, that towards the villus tips, they start expressing BMF. So we think a normal mechanism of cell death is again been hijacked.’ 

In colorectal cancer, a complicated and plastic tumour type, the input of Owen Sansom and Kevin Haigis from SPECIFICANCER was critical, with Kristian Helin’s lab being instrumental for all of the EZH2 projects. The team has more work in other tissue types to come, which reinforces the paradigms governing tissue specificity that the team have identified. Karen emphasises ‘We think EZH2 is acting as an epigenetic insulator. In other words, the PRC2 complex instructs tissue-specific oncogenes by cooperatively regulating cell state, and also reinforces the effects of specific oncogenes in different tissues by cooperatively preventing apoptosis.’ 

Collaborating with the lab of Simon Knott at the Cedars-Sinai Medical Center, through a DOD-funded Transformative Breast Cancer Consortium Award, the team also developed a classifier to predict which TNBC patients would be responsive to combined EZH2 and AKT inhibition by using machine learning. The classifier is a transcriptional signature based on the baseline epigenetic state of the tumour cells. The team hope to incorporate this into clinical trials, with the approach currently requiring RNA sequencing of patient tumour biopsies. The team also identified that non-responsive TNBC cells could be sensitised by forced expression of GATA3, the key driver of differentiation, in combination with STING agonists, triggering IL6 expression and subsequent upregulation of BMF and cell death. This finding suggests that the conversion to a sensitive state is possible, which will be an area of future research.

In addition to detailing how EZH2 interplays with tissue specific oncogenic pathways, Amy also led work from the SPECIFICANCER team showing how EZH2 cooperates with other epigenetic programs, further demonstrating how EZH2 exhibits tissue-specific effects. Published last year in PLOS BIOLOGY, Amy demonstrated the promise of combined EZH2 and histone deacetylase (HDAC) inhibition in castration-resistant prostate cancers, an incurable form of the disease. In this case, combined inhibition allows the sequential demethylation and acetylation of histone H3, to allow derepression and induction of targets. In prostate, it is upregulation of ATF3, a transcription factor responsive to a wide range of cellular stresses, that is critical in mediating cancer cell death.