Fernandez-Sanroman, A. | bioRxiv Science

bioRxiv2024

A comprehensive genetic catalog of human double-strand break repair

The analysis of DNA sequence outcomes provides molecular insights into double-strand break (DSB) repair mechanisms. By employing parallel in-pool profiling of Cas9-induced indels within a genome-wide knockout library, we present a comprehensive catalog detailing how virtually every human gene influences the DSB repair process. This REPAIRome resource is validated through the identification of novel mechanisms, pathways and factors involved in DSB repair, including unexpected opposing roles for XLF and PAXX in DNA end processing, a molecular explanation for Cas9-induced multi-nucleotide insertions, the identification of HLTF as a DSB-repair factor, the involvement of the SAGA complex in microhomology-mediated end joining, and importantly, an indel mutational signature linked to VHL loss, renal carcinoma and hypoxia. Collectively, these results exemplify the potential of REPAIRome to drive future discoveries in DSB repair, CRISPR-Cas gene editing and the etiology of cancer mutational signatures.

molecular biology↗

bioRxiv2023

Predicting response to cytotoxic chemotherapy

Cytotoxic chemotherapies have been a crucial part of cancer treatment for over 40 years. While their primary target is cancer cells, they can also harm normal cells, resulting in dose-limiting toxicity. Most chemotherapies were approved before the advent of precision biomarkers, as such, many patients experience severe toxic side effects without any benefit. To address this challenge, we have developed three precision biomarkers to predict response to platins, taxanes, and anthracyclines. Based on chromosomal instability (CIN) signatures, these biomarkers can be computed from a single genomic test. For platins and taxanes, we used CIN signatures related to impaired homologous recombination, while for anthracyclines, we discovered a CIN signature representing micronuclei induction which predicts resistance. In a clinical study involving 41 high-grade serous ovarian cancers, patients predicted to be sensitive by these biomarkers showed significantly prolonged progression-free survival. To further validate the effectiveness of the taxane and anthracycline predictors, we conducted a retrospective randomised control study involving 182 ovarian and 219 breast cancer patients. Patients predicted as resistant showed increased risk of time to treatment failure compared to standard of care, hazard ratios of 1.73 (95%CI=0.98-3.07) for taxane in ovarian, 3.67 (95%CI=2.12-6.34) for taxane in breast, and 1.93 (95%CI=1.22-3.04) for doxorubicin in ovarian. We also found that liquid biopsies can be used to make these predictions in up to 30% of ovarian cancer patients. Our findings highlight the clinical value of CIN signatures in predicting treatment response to various chemotherapies across multiple different types of cancer. The ability to quantify multiple CIN signature biomarkers using a single genomic test offers a unified approach to guide treatment decisions for cytotoxic chemotherapies. Ultimately, this has the potential to transform the current one-size-fits-all chemotherapy approach into a more precise and tailored form of medicine.

cancer biology↗