Inhibitors of ATR, a central kinase controlling DNA replication origin firing and cellular checkpoint activity, are currently in multiple clinical trials, yet mechanisms underpinning sensitivity and robust patient stratification biomarkers are lacking. In our new study we carefully characterized the mechanism underlying sensitivity to the important clinical target, the ATR kinase. Using an unbiased, multimodal approach in multiple cancer cell lines, AML patient samples and wider databases, we discovered an overlooked role for deregulated DNA replication origin firing in ATR inhibitor sensitivity.

We used a non-invasive live microscope to monitor cell confluence in a panel of breast cancer cell lines to identify sensitive and resistant cell lines to ATR inhibition. We performed immunofluorescence microscopy to investigate rates of chromosome instability and replication stress. We then performed proteomics, phospho-proteomics and gene expression analysis to discover determinants of ATRi sensitivity. In parallel, we applied a newly developed machine learning coupled with long-read sequencing method to investigate DNA replication dynamics. Then, we used publicly available transcriptomic data from Cancer Cell Line Encyclopaedia and published Acute Myeloid Leukaemia patient data to investigate gene expression signature enrichment in larger datasets and including more cancer types.

We discovered that all cell lines had increased levels of replication stress and chromosome instability upon ATRi, regardless of sensitivity. Interestingly, sensitive cell lines had higher expression and activity of DNA replication origin firing factors at basal. Moreover, in response to ATR inhibition, they massively increased origin firing, leading to the activation of the DNA damage response pathway and cell death. ATRi sensitivity was partly rescued upon co-treatment with XL-413, a CDC7 inhibitor that decreases origin firing. High expression of DNA replication initiation factors correlated with ATRi sensitivity across multiple cancer types, and in acute myeloid leukaemia patient samples.

Our study reveals that in sensitive cancer cell lines or patient samples, excessive origin firing is a detrimental response that leads to cell death. Moreover, this vulnerable state can be detected in sensitive samples by a higher level of origin firing-related factors, providing potential new and specific biomarkers for ATR inhibitor sensitivity that are compatible with clinical application.