Chromosomal instability (CIN) results in the accumulation of structural and numerical alterations across the genome. Most CIN-related alterations arise from endogenous causes, such as defects in DNA repair mechanisms. Current experimental models often exhibit multiple types of CIN simultaneously. For instance, widely used cancer cell lines harbour mutations in multiple genes and display highly unstable genetic backgrounds, making it difficult to study the genomic impact of a specific CIN-related mutational process. Therefore, to better understand the role of CIN in tumorigenesis, it is essential to develop new model systems that express a single CIN type in a normal genetic background. As a proof of concept, we focused on developing in vitro models of impaired homologous recombination (IHR), a therapeutically targetable type of CIN.

We have induced IHR in the diploid human-immortalized cell line hTERT RPE-1 Cas9 by knocking out BRCA1 or BRCA2 via CRISPR/Cas9 gene editing. Prior to BRCA1 or BRCA2 editing, we knocked out TP53 to ensure tolerance to CIN. Once we identified knockout clones, we performed single-cell shallow Whole Genome Sequencing (sWGS) after a limited amount of cell cycles post-editing (~80). We also assessed their sensitivity to olaparib and rucaparib, two PARP inhibitors (PARPi) routinely used to treat patients with IHR tumours.

We have observed that TP53-/- BRCA1-/- and TP53-/- BRCA2-/- cells acquire new copy number alterations compared to parental (TP53-/- BRCA1/2WT) cells, confirming that IHR-related CIN was induced via CRISPR/Cas9 gene editing in a TP53-/- background. These newly-acquired alterations are probabilistically mapped to CIN signatures linked to mitotic errors and IHR, suggesting a specific pattern of molecular defects induced by the loss of BRCA1 or BRCA2. Parental single cells (TP53-/-) display minimal genomic alterations associated with occasional chromosome missegregation. Importantly, the genomic scars left by inducing IHR were not detected at bulk sequencing level, despite the clear differences in genomic changes observed at single-cell sequencing levell, underscoring the need for single cell analyses. Indeed, TP53-/- BRCA1-/- and TP53-/- BRCA2-/- cells show significantly increased sensitivity to olaparib and rucaparib compared to parental cells.

These findings underscore the value of models harbouring isolated molecular defects in linking specific CIN patterns to their underlying causes and that our single-cell workflow accurately detects CIN-types in a premalignant scenario, which has not been possible to date given the limited resolution of standard bulk sequencing. Therefore, we are now increasing the collection of these models by knocking out 562 genes related to other DNA Damage Response pathways.