BackgroundHomologous recombination (HR) repair deficiency arising from defects in BRCA1 or BRCA2 is associated with characteristic patterns of somatic mutations. In this genetic study, we ask whether inactivating mutations in further genes of the HR pathway or the DNA damage checkpoint also give rise to somatic mutation patterns that can be used for treatment prediction.ResultsUsing whole genome sequencing of an isogenic knockout cell line panel, we find a universal HR deficiency-specific base substitution signature that is similar to COSMIC signature 3. In contrast, we detect different deletion phenotypes corresponding to specific HR mutants. The inactivation of BRCA2 or PALB2 leads to larger deletions, typically with microhomology, when compared to the disruption of BRCA1, RAD51 paralogs, or RAD54. Comparison with the deletion spectrum of Cas9 cut sites suggests that most spontaneously arising genomic deletions are not the consequence of double-strand breaks. Surprisingly, the inactivation of checkpoint kinases ATM and CHK2 has no mutagenic consequences. Analysis of tumor exomes with biallelic inactivating mutations in the investigated genes confirms the validity of the cell line models. We present a comprehensive analysis of sensitivity of the investigated mutants to 13 therapeutic agents for the purpose of correlating genomic mutagenic phenotypes with drug sensitivity.ConclusionOur results suggest that no single genomic mutational class shows perfect correlation with sensitivity to common treatments, but the contribution of COSMIC signature 3 to base substitutions, or a combined measure of different features, may be reasonably good at predicting platinum and PARP inhibitor sensitivity.

Calcification of various tissues is a significant health issue associated with aging, cancer and autoimmune diseases. There are both environmental and genetic factors behind this phenomenon and understanding them is essential for the development of efficient therapeutic approaches. Pseudoxanthoma elasticum (PXE) is a rare genetic disease, a prototype for calcification disorders, resulting from the dysfunction of ABCC6, a transport protein found in the membranes of cells. It is identified by excess calcification in a variety of tissues (e.g., eyes, skin, arteries) and currently it has no cure, known treatments target the symptoms only. Preclinical studies of PXE have been successful in mice, proving the usefulness of animal models for the study of the disease. Here, we present a new zebrafish (Danio rerio) model for PXE. By resolving some ambiguous assemblies in the zebrafish genome, we show that there are two functional and one non-functional paralogs for ABCC6 in zebrafish (abcc6a, abcc6b.1, and abcc6b.2, respectively). We created single and double mutants for the functional paralogs and characterized their calcification defects with a combination of techniques. Zebrafish deficient in abcc6a show defects in their vertebral calcification and also display ectopic calcification foci in their soft tissues. Our results also suggest that the impairment of abcc6b.1 does not affect this biological process.