Abstract PR018: Adaptive local fitness landscapes for aneuploid karyotypes

Abstract

Abstract Aneuploidy, which occurs in most solid tumors, dramatically alters the expression of many genes which in turn has broad phenotypic effects. Recent evidence suggests that specific aneuploid karyotypes confer a selective advantage. However, various questions remain unanswered regarding the magnitude of these selective advantages, as well as the repeatability and predictability of karyotype evolution. We developed ALFA-K (Adaptive Local Fitness landscapes for Aneuploid Karotypes) – the first approach to infer karyotype fitness landscapes in a context-dependent manner. Based on the dynamics of just a few subclones derived from longitudinal single cell sequencing data, our method can extrapolate the fitness of thousands of karyotypes. We validate the predictive power of our method on data from several P53 deficient cell lines which exhibited substantial subclonal evolution across multiple passages in vivo or in vitro. Based on topological analyses of the inferred fitness landscapes we investigated how mis-segregation rate itself shapes the stringency of karyotype evolution and whether karyotype evolution is subject to diminishing returns epistasis, sign epistasis or reciprocal sign epistasis. We adapted metrics from geo-statistics to study the fitness landscape of a P53 deficient 184-hTERT diploid breast epithelial cell line that underwent whole genome doubling (WGD) and found support for the hypothesis that WGD benefits tumor cells by enhancing their robustness against chromosomal instability. The stark phenotypic consequences of karyotype changes emphasize the value of predicting the evolutionary trajectory of karyotype composition when exposing cells to a new environment. Our work contributes to accumulating evidence that karyotype evolution is predictable. Citation Format: Richard J. Beck, Noemi Andor. Adaptive local fitness landscapes for aneuploid karyotypes [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr PR018.