We often think of our DNA as a nice tightly wound package that exists within the nucleus of the cell. Yet, in various cancers, there are fragments of our genetic code that reside outside of the contiguous genome—called extrachromosomal DNA (ecDNA)—that play a critical role in carcinogenesis and chemotherapy resistance. For instance, glioblastoma (GBM), the most common and aggressive form of brain cancer, has an extremely poor response to standard-of-care treatment, with a two-year survival rate of only 15%. New research is beginning to provide a better understanding of the processes underlying DNA disparities within GBM tumors—a crucial finding because these differences contribute to therapy resistance.
Now, a multi-institutional research team led by investigators at the Henry Ford Health System's Hermelin Brain Tumor Center in Detroit and The Jackson Laboratory (JAX) has tracked genomic alterations detected in patient samples during tumor cell evolution in culture, in patient-derived xenograft (PDX) mouse models from the cultures, as well as before and after treatment in patients. Findings from the new study—published recently in Nature Genetics, in an article entitled “Discordant Inheritance of Chromosomal and Extrachromosomal DNA Elements Contributes to Dynamic Disease Evolution in Glioblastoma”—that tumor progression was often driven by cancer-promoting genes, known as oncogenes, on extrachromosomal pieces of DNA.
ecDNA elements were first observed directly under microscopes in cancer cells more than 50 years ago, but it remains unknown how they arise in the first place. Technological limitations have impeded studies of ecDNA in detail, despite a recent publication in Nature suggesting their presence in nearly half of cancers. In fact, their role in disease has not been extensively studied, but it's an important topic. Unlike chromosomal DNA, ecDNA is inherited inconsistently as a tumor grows. That is, when a cancer cell divides, the DNA on the chromosomes almost always gets accurately duplicated and remains the same in the daughter cells. But ecDNA inheritance appears to be far more random. Sometimes both daughter cells inherit ecDNA, but sometimes all or most of it will end up in one cell and not the other.
"The process quickly creates important differences between cells within the same tumor, and it helps accelerate the evolution of the cancer," noted co-senior study investigator Roel Verhaak, Ph.D., professor and associate director of computational biology at JAX. "It provides the cells with more ways to evade stress. Therefore, there's a better chance that at least some of the cells will survive severe stress, such as stresses caused by a chemotherapy or radiation."
In the current study, the scientists performed detailed analyses of the tumor cells from patients to culture to mouse, revealing that most cells retained the same genomic lesions. This is good news overall, as it indicates that PDX mice can provide a relatively accurate and effective experimental platform for GBM. However, the primary caveat was the finding that in a few cases, the numbers of oncogene copies differed between tumors and the cultures and PDX mouse samples derived from them. If an oncogene is increased or amplified, that can both cause and maintain cancer, so differences in gene amplification can be very important.
The underlying question that the researchers sought to uncover is why did the levels of oncogene amplification change? What the researchers found was that the differences were caused by oncogenes that weren't part of chromosome sequences as usual. Instead, they were on separate circular pieces of ecDNA. These pieces of DNA are not found in normal cells and cause major increases in the expression of oncogenes. More detailed investigation showed that many instances of oncogene amplification found in the glioma tumors involved ecDNA elements.
"The selective advantage conferred to tumor cells by the regulation of oncogene copy number in ecDNA has not been sufficiently addressed in interpreting results in the laboratory or in clinical trials,” explained co-senior study investigator Ana deCarvalho, Ph.D., assistant professor at the Hermelin Brain Tumor Center. “Using the GBM patient-derived models carrying ecDNA amplification of the most frequent oncogenes, we are developing and testing novel combination therapies specific for each unique tumor.”
One reason ecDNA has been relatively ignored is that it's hard to detect using standard sequencing methods, which don't accurately detect and separate it from chromosomal DNA. But it's now attracting more attention, and the work moving forward will likely help explain why cancers such as GBM are difficult to treat and evolve therapy resistance so rapidly.
"We think targeting ecDNA has huge potential for the development of new cancer treatments," Dr. Verhaak concluded. "We're now working to develop sequencing-based protocols to identify ecDNA more efficiently. The bigger goal is to learn how and why ecDNA elements form. If we can block those mechanisms, we'll have a way to prevent the evolution, and perhaps even the formation, of many cancers."