Evolution of oncogene amplification across 86,000 cancer cell genomes

Read full text
Authors: Jake June-Koo Lee, Jake June-Koo Lee, Sohrab Salehi, Matthew A. Myers, Marc J Williams, Melissa A. Yao, Duaa Hassan Al-Rawi, Jin Lee, Jin Lee, Eric G. Sun, Kerstin Thol, Seongmin Choi
+22 more Eliyahu Havasov, Asher Preska Steinberg, Michelle Ka Yan Wu, Nicole Rusk, Caitlin Timmons, Cheryl Zi Jin Phua, Stephen Martis, Neeman Mohibullah, Parvathy Manoj, Esther Redín, Àlvaro Quintanal-Villalonga, P. Razavi, Samuel A. J. R. Aparicio, Natasha Rekhtman, Viviane S. Tabar, Mark M. Awad, Helena A. Yu, Kenny Kwok Hei Yu, Andrea Ventura, Andrew McPherson, C M Rudin, Sohrab P. Shah
Cited by: 2

Abstract

High-level copy-number (CN) amplification (HLAMP) is a major mechanism of oncogene activation in human cancer. Despite progress in therapeutically targeting amplified oncogenes, the processes underlying amplicon evolution remain incompletely understood, leaving critical knowledge gaps in their etiology and mechanisms of therapeutic response. To address this, we analyzed the evolutionary trajectories of HLAMPs using single-cell whole-genome sequencing data from 86,239 cancer cells across 93 patients and 9 experimental systems. We found that cell-to-cell CN variability provides a quantifiable readout of HLAMP mechanism, clearly distinguishing extrachromosomal circular DNA (ecDNA) from intrachromosomal amplification (ICamp) through characteristic CN distributions that reflect distinct modes of segregation and correspond to clonal architecture. Notably, ICamp events frequently showed multiple amplitude peaks specific to subclones, indicating punctuated shifts in oncogene dosage through numeric or structural modulatory mechanisms with transcriptional impact. In contrast, ecDNAs exhibited broad, continuous CN distribution with extreme high-copy outliers, consistent with asymmetric segregation. The CN and structural diversity of ecDNA regions enabled systematic deconvolution of ecDNA subspecies and estimation of their per-cell abundance, revealing the history of ecDNA-mediated oncogenesis at single-nucleotide resolution. We observed ecDNA diversification through internal rearrangements across cases and, notably, convergent evolution in glioblastoma cases marked by multiple, recurrent acquisition of EGFR-targeting ecDNAs. Finally, single-cell genome-based identification of ecDNAs showed substantial discrepancy with bulk genome graph-based predictions and reliably distinguished actively maintained ecDNAs from historical genomic footprints after chromosomal re-integration. These findings reveal marked tissue-type specificity of ecDNAs, suggesting that ecDNA-mediated oncogenesis may depend on a permissive tissue context.

Fields of Study & Subjects

Cancer Genomics and Diagnostics Single-cell and spatial transcriptomics Genomic variations and chromosomal abnormalities

Publication details

Journal / source
bioRxiv (Cold Spring Harbor Laboratory)
Publisher
openRxiv
Publication date
14 February 2026
Language
EN
Document type
Preprint