Abstract 3122: Single-cell DNA replication dynamics for genomically unstable cancer cells

Abstract

Abstract Studying DNA replication dynamics of genomically unstable cancers is important for understanding how genome diversity generating processes shape subclonal evolutionary trajectories. The emergence of single-cell whole genome sequencing (scWGS) has enabled novel interrogation of how copy number aberration (CNA) patterns shape tumor evolution and, separately, of single-cell replication timing (scRT) profiles in developing tissues or diploid cell lines; however, there is a lack of computational methods to study scRT dynamics in samples with high CNA rates. Exploring such an intersection aims to understand genomic and epigenomic co-evolution as replication timing (RT) is linked with epigenetic modifications and nuclear organization. Additionally, it aims to quantify replication stress and relative proliferation rates of tumor subclones. We thus developed a method for probabilistic estimation of replication timing (PERT) which infers scRT profiles for cancer cells with subclonal and cell-specific CNAs. PERT uses a Bayesian framework and informative priors to jointly infer replication and copy number states of S-phase cells. We benchmarked PERT’s accuracy using 200 simulated datasets with varying cell-specific CNA rates and subclonal structures and its ability to improve cell cycle predictions through permutation experiments on cell cycle sorted cell lines. PERT enabled analysis of &gt10,000 S-phase cells across a cohort consisting of 6 wild-type and 6 genetically engineered cell lines, 13 high grade serous ovarian cancers, and 6 triple negative breast cancers (TNBC). Of this cohort, 7 were time-series datasets with 3 of such being TNBC patient-derived xenografts containing on and off cisplatin timepoints. PERT’s output enabled us to calculate genome-wide scRT variability and detect sample and subclonal RT shifts. We found that cells with significantly earlier or later replication of chromosome X had unbalanced ratios of active (Xa) to inactive (Xi) alleles, in line with previous reports that Xi replicates much later than Xa due to its repressive epigenetic marks and location within the nucleus. We found that intrasample scRT variability increased as a function of the number of subclones in a sample, consistent with CNAs altering RT, whereas intraclonal scRT variability increased in settings consistent with replication stress. Finally, we developed an analysis framework for time-series data to examine whether subclones enriched for S-phase cells undergo expansion if left untreated or contract if treated with cisplatin compared to cells depleted for S-phase cells. Altogether, this work illuminates how aberrant replication can both generate and result from CNAs while also improving prediction of replication stress, proliferation rates and chemosensitivity at subclonal resolution. This work is broadly important for understanding how to better target the evolvability of genomically unstable cancers. Citation Format: Adam C. Weiner, Andrew McPherson, Sohrab P. Shah. Single-cell DNA replication dynamics for genomically unstable cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3122.