UGI Internal Seminar - Daniel Jacobson, UCL Genetics Institute

Abstract:

Homologous recombination is a largely error-free method of DNA repair that relies on genes such as BRCA1 and BRCA2. Defects in these genes can lead to homologous recombination deficiency (HRD) which confers sensitivity to PARP inhibitors and drives widespread chromosomal instability, triggering cGAS/STING-mediated inflammation. This inflammation is often treated with immunosuppressors such as infliximab and tocilizumab, raising questions about their combined effect with DNA damaging agents. The crosstalk between these processes complicates the search for safe and effective combination therapies, which calls for innovative computational approaches to predict optimised combination treatments and their mechanisms of action. We developed a mutation-based classifier of HRD for exome sequenced breast cancer, which we then applied to >900 primary breast cancer samples from The Cancer Genome Atlas. Using these classifications and matched bulk RNA-seq data, we applied multinomial elastic net regression to develop and validate a 228-gene transcriptional signature of heterogeneous HRD. This signature correlates with PARP inhibitor sensitivity in breast cancer cell lines, and in patients from the I-SPY2 trial treated with olaparib and durvalumab. Subsequently, the signature was employed to elucidate tumour-wide HRD prevalence and heterogeneity across two cohorts of single cell-sequenced breast cancers, and we demonstrate that HRD cells display a global reduction of responsiveness to interactions with the tumour microenvironment, typified by decreased TNF-alpha-mediated communication. Finally, we developed an executable model describing the crosstalk between the DNA damage and cGAS/STING-mediated inflammatory responses, which we used to screen for effective and safe combination treatments. We predict that DNA damaging agents synergise with tocilizumab, and that infliximab may marginally diminish treatment efficacy. Overall, we reveal insights into the complex relationship between the DNA damage and inflammatory responses, and provide a resource for which novel therapeutic combinations can be screened and demonstrate its clinical significance in combining cancer treatment with immunosuppression and inflammatory targeting.