Summary
A bioRxiv preprint uses a polymer-physics model to explain how ecDNA copy number may shape regulatory contacts in cancer-cell condensates. In the model, contacts in oncogene-associated regions grow stronger and become less variable as copy number increases.
A bioRxiv preprint posted September 26, 2026, describes a polymer-physics model in which increasing the copy number of circular extrachromosomal DNA (ecDNA) strengthens and stabilises some regulatory contacts linked to cancer genes. The model, built around MYC-harbouring COLO320-DM ecDNA, suggests how condensates can vary in overall shape while retaining recurring patterns of contact between DNA molecules.
The authors propose that this combination could create a more consistent regulatory environment for oncogenes located in particular regions of ecDNA condensates. The work is a theoretical model, rather than a clinical study or a direct test of a treatment.
How the model forms ecDNA condensates
EcDNA consists of DNA outside the cell’s chromosomes; the study models it as rings that can carry oncogenes. The authors represent BRD4-like complexes as binders that attach to matching sites along the rings and can bridge those sites. Above a critical binder concentration, the model’s ecDNA molecules separate into a dense condensate.
The simulated condensates take on diverse three-dimensional conformations, which the authors attribute to the number of energetically accessible ways the DNA can fold. Yet their contact patterns are not wholly random. The model produces recurring interaction scaffolds, including in-trans associated domains, or I-TADs: regions enriched for regulatory contacts between distinct ecDNA molecules.
This distinction matters because the overall shape of a condensate and the contacts within it are not the same property. In the model, substantial variation in shape can coexist with reproducible patterns of interaction.
Copy number and oncogene contacts
The authors report that the condensate’s three-dimensional architecture follows scaling relations as ecDNA copy number, n, changes. They say these relations remain robust when model parameters are varied. Within I-TADs, regulatory contacts increase linearly with n. Compared with size-matched control regions outside I-TADs, those contacts are one order of magnitude stronger, while their relative fluctuations decline as copy number rises.
The model therefore links more ecDNA copies not only to stronger contacts in I-TAD regions, but also to less variable contacts there. The authors propose that this could give oncogenes embedded in I-TADs, including PVT1-MYC fusions, an enhanced, lower-noise regulatory environment. They contrast these with the canonical MYC copy, which they place outside I-TADs and describe as less amplified, consistent with experimental observations cited in the preprint.
The proposed mechanism offers a way to connect ecDNA organisation with selective oncogene amplification. Its conclusions are specific to a minimal polymer-physics model of COLO320-DM ecDNA and BRD4-like binders; the reported results are model findings, not measurements of clinical outcomes.