NEW YORK – By following the dynamics of extrachromosomal DNA (ecDNA) — which often contains oncogenes — during cell division, new research has demonstrated the ways that random ecDNA inheritance can impact tumor development and treatment resistance.
"The findings presented in this article reveal that ecDNA uniquely shapes each of the foundational principles of Darwinian evolution, that is, random inheritance by descent, enhanced variation through random segregation, and selection, thereby accelerating tumor cell evolution and increasing adaptability," the authors wrote. "Such observations may explain why clinical activity from therapies targeting oncogenic amplification events are so limited in tumors such as GBM, where ecDNAs are so prevalent."
As they reported in Nature Genetics on Monday, researchers at Stanford University, the University of California San Diego, Queen Mary University of London, and other international centers relied on a combination of theoretical modeling, imaging, CRISPR-based ecDNA tagging, and a CRISPR-based ecDNA production approach known as CRISPR-C to follow ecDNA patterns in cancer cell lines and tumor samples.
"[W]e integrated computer simulations, mathematical modeling, evolutionary theory, unbiased image analysis, CRISPR-based ecDNA tagging with live-cell imaging, and CRISPR-C to generate ecDNA, as well as longitudinal analyses of patients' tumors, to better understand ecDNA inheritance and its functional consequences," the authors explained.
While past studies have unearthed apparent ties between ecDNA and everything from oncogene amplification and tumor development to treatment response and cancer patient outcomes, the team noted that prior research has provided fewer clues related to the consequences of non-chromosomal inheritance of ecDNAs and of oncogenes they may contain.
"It has been suggested that ecDNAs, because they lack centromeres, are unequally segregated to daughter cells during cell division," the authors noted. "However, the impact of non-chromosomal oncogene inheritance in cancer — random identity by descent — on intratumoral genetic heterogeneity, accelerated tumor evolution, enhanced ability to withstand environmental stresses, and rapid genome changes on therapeutic resistance, is not well understood."
With the help of fluorescence in situ hybridization probing, immunostaining, and imaging, the researchers quantified ecDNA distribution to daughter cells produced after mitotic cell division in cell lines representing prostate cancer, gastric cancer, colon cancer, neuroblastoma, or glioblastoma (GBM). Across those cancer types, they found that the results lined up with theoretical predictions modeling random ecDNA segregation.
The team went on to shore up the results with live-cell imaging on prostate cancer cells containing fluorescent protein-tagged versions of ecDNA, before exploring the evolutionary and tumor biology effects of random ecDNA segregation.
Based on their mathematical modeling, simulation analyses, and ecDNA copy number predictions in half a dozen cancer cell lines — combined with FISH experiments, clinical data, and tumor tissue profiles for six GBM cases and four cases of neuroblastoma — the investigators saw dramatic differences in ecDNA representation from one cancer cell to the next, consistent with the notion that the extrachromosomal sequences can contribute to heterogeneity within tumors.
Along with follow-up analyses of CRISPR-C and other data, the team found that the presence of ecDNA in specific cancer cells may help tumors dodge and adapt to treatment, subsequently impacting cancer growth and survival.
These and other results suggested that "ecDNA inheritance can predict, a priori, some of the aggressive features of ecDNA-containing cancers," the authors concluded, noting that "[t]hese properties are facilitated by the ability of ecDNA to rapidly adapt genomes in a way that is not possible through chromosomal oncogene amplification."