Shining a light on extrachromosomal DNA
In 2021, Stanford ChEM-H welcomed physician scientist Paul Mischel, professor of pathology, as an Institute Scholar. Mischel has pioneered the field of extrachromosomal DNA (ecDNA), circles of genes that float outside chromosomes, the well-organized home for genes in healthy cells. These long-overlooked structures, present in an estimated up to one-third of all cancer patients, allow tumors to grow and develop resistance to drugs. He talks to us about his many lives as a physician scientist, looking where the light isn’t being shone, and how genetic SpaghettiOs give cancers an advantage.
You were originally trained as a physician, and now you lead a research lab studying fundamental and translational cancer biology. Can you tell me how you got here?
I’ve lived many lives in my career. I lost my dad to stomach cancer when I was 14. It was a very difficult time, and I wanted to do something about it. I went to medical school and became a pathologist so that I could look the enemy in the eyes. But I realized that I might spend my whole life looking the enemy in the eyes and never find a new treatment.
So, I trained as a scientist and started a research lab. That was in the early days of precision medicine, and we were using science to develop better drugs for patients and using patient results to reveal more about the fundamental biology. The problem was that the patients weren’t getting better. I realized that I was going to have to look at the problem through a different lens.
This thought coincided with something quite personal, which was that my dad had died at 51, and I was turning 50. I decided to look at cancer from a completely different angle, to look where the light was not being shone. This brought me to shift how my lab was looking at the problem, paying a lot of attention to the fundamental biology of chromosomes, to uncover why patients weren’t getting better. We discovered that in many patients, cancer-causing genes, or oncogenes, are found not only on chromosomes, where nearly all genes are found in healthy cells, but also on circles of DNA outside those chromosomes. This led us to understand the role of this extrachromosomal DNA (ecDNA) in tumor growth and drug resistance.
What inspired you to go looking for ecDNA?
We first looked at glioblastoma, an aggressive brain cancer. These tumors defied our assumptions. According to classical genetics, as a mother cell divides it gives its genes equally to daughter cells. It takes time for resistance-driving mutations in DNA to develop, and more time for those drug-resistant cells to become the dominant cell type in the tumor.
In these brain tumors, the pattern was right, but the timing was completely wrong. It was happening way too fast. It made no sense.
So, we did what hadn’t been done in a long time, and we looked inside the cells. And there, hiding in plain sight in the nucleus of the cell, were oncogenes right there on extrachromosomal pieces of DNA. We knew that this was the key.
ecDNA has been known but overlooked for 50 years, and for a long time, people thought it was rare and of unclear significance, but that was mostly because they lacked the tools we have today to identify both the molecular makeup and cellular location of genes. Luckily for us, Vineet Bafna, a computer scientist who had worked with Craig Venter on the sequencing of the human genome, became interested in this problem and together we integrated the molecular and the computational with the visual to learn what the genes were and where they were. Later, we were joined by Stanford professor Howard Chang, who is a pioneer in understanding how genes can be turned “on” or “off” in cancers based on how accessible they are to the machinery that binds to and reads DNA. Howard too, arrived at the importance of ecDNA by looking at the pattern of genome organization, and we began to work closely together.
Together, we decoded what ecDNA really looks like; like doughnuts or SpaghettiOs, ecDNA circles were staring us straight in the face. Unequivocally, they were circles.
How does having ecDNA allow cancers to become more drug-resistant?
Circles are nature’s currency. These genetic circles don’t play by the same rules as chromosomes. When a mother cell divides, the ecDNA might be divided unequally between daughter cells. This means that each cell division is like a coin flip, which allows some cancer cells to contain many copies of a single oncogene while also maintaining variation. And that variation—that coin flip—is the fuel of Darwin’s natural selection. It’s like an evolutionary machine on steroids.
The circularity also changes the organization. Instead of buried in tightly-wound chromosomes, the oncogenes are more accessible, driving oncogene expression to unprecedented levels.
The advantage of ecDNA extends beyond cancers. This is nature's way of adapting quickly to changing conditions. There's even data suggesting that weeds that become resistant to Roundup do so by amplifying the resistance gene on circular DNA. It's everywhere. Healthy human cells can’t do it; cancer cells can.
How have other researchers responded to these findings?
When we first published this in Science in 2014, the response was a giant snooze. Many people thought it was an anomaly, something specific to the brain tumors we had studied and not applicable to other tumors. That assumption is partly due to our trust in genome maps. The power of next generation sequencing is that you can take a piece of a tumor, grind it up to extract the DNA, identify the relevant oncogene and infer where in the full genome that gene is found. The problem is that these maps use normal, healthy cells as a reference point, and you have to assume that things in a cancer cell are where they would be in a healthy cell. That’s a huge inference.
I kept thinking about Ptolemy’s map of the solar system. Astronomers used very precise measurements to track planets across the night sky. The measurements were correct, but the map was totally wrong because they put Earth in the center.
It took 1400 years to get from Ptolemy’s map to Copernicus’, but it’s taken less time to convince people that our genome maps might be leading us astray. Since we and others have shown the impact of ecDNA on cancer progression, and that it’s present in at least 15% but maybe as high as 33% of all cancers, people have started to appreciate that this is a big deal.
You know that ecDNA exists, that it’s present in many cancer cells, and that it gives those tumors an advantage when clinicians try to drug them. So, what can we do about it?
That’s what we are working on. As with everything in nature, each advantage brings with it a vulnerability. The good news is that we’re beginning to see that this problem is tractable. We are trying to answer questions about the fundamental processes, questions like: How does ecDNA form? How is it maintained? What do we need to do to turn on our own defenses? But also, more importantly, how can we diagnose patients? How can we monitor them? And how can we treat them?
These questions will take years, but we don’t have years to spare, so we’ve built an interdisciplinary team of global experts to attack this problem from all angles. Together, we are focusing not just on knowledge, but on hope.
Why did you decide to join ChEM-H?
How could I not? The ChEM-H structure is built for the future of science, for a world in which we interact closely as teams, embed experts in one another’s labs, and are open to finding solutions to problems in unlikely places.
When Chaitan Khosla first described to me what they were building here at ChEM-H, I practically fell off my chair. Who would not want to be part of this? If you believe in the power of science to transform medicine, there is no place in the world you would rather be. You can feel the intellectual ferment and excitement that's brewing right here. I am also grateful to Tom Montine for bringing me into the Department of Pathology, whose vision and mission align so beautifully with ChEM-H. The timing was especially wonderful as my wife, Deborah Kado, was also being recruited to the School of Medicine to be Chief of Geriatric Research.
The first phase in my life was purely clinical phase. The second phase in my life was translational phase. The third phase in my life was a pure science phase. I'm now ready and excited to put the pieces together, and I can't imagine a better place to be doing it than at ChEM-H.