PO-332 Genomic landscapes, neoantigen profiles and biological impact of MLH1 inactivation in cancer cells

Abstract

IntroductionAlterations in DNA repair pathways are thought to fuel tumour progression. Mismatch Repair (MMR) deficient cancers show peculiar biological features such as an indolent progression and a resolute therapeutic response to checkpoint inhibitors. The genomic and biological bases of the peculiar clinical features are poorly understood. Further progress in this area is limited by the paucity of models to study the impact of MMR genes inactivation at the genomic and biological levels. To address this issue we developed a bioinformatic workflow to monitor the neoantigen repertoire induced by inactivation of the Mlh1 gene (a key player of the MMR machinery), in murine cell lines.Material and methodsWe inactivated Mlh1 throughout the CRISPR-Cas9 technology in CT26 (colon cancer), PDAC (pancreatic cancer) and TSA (breast cancer) murine cell lines. We performed whole exome sequencing (WES) at different time points and then we quantified the amount of mutations (SNVs and indels). We generated a pipeline that characterises the neoantigen repertoire, starting from annotated alterations and the HLA of specific murine strain. In parallel, we inoculated MMR-proficient and -deficient cells in immuno-compromised and -competent mice and monitored their growth.Results and discussionsIn all pre-clinical models analysed we found a massive increment in the number of non-synonymous alterations (up to 100% increase respect to basal population) after Mlh1 inactivation. Notably, analysis of MMR deficient mouse cells at different time points showed a renewal of mutational profile and consequently an accumulation of predicted neoantigens. We further characterised the SNVs and frameshifts acquired by Mlh1-knockout cells. In agreement with data in human tumours, the fraction of predicted neoantigens derived from frameshifts was higher than the SNV-derived neoantigens. When injected in immuno-compromised mice the Mlh1-knockout cells and their wild type counterpart showed comparable growth. On the contrary, MMR-deficient cells but not their control counterpart grew poorly in immuno-competent mice and responded promptly to treatment with checkpoint inhibitors.ConclusionWe find that Mlh1 gene inactivation drives dynamic neoantigen profiles, which can be monitored with an ad hoc bioinformatic pipeline. These analyses provide mechanistic support to understand why MMR deficient cells engage the immune system of the host, foster immune surveillance and tumour control.