THOMAS FARNETTI/WELLCOME IMAGES
In 1993, Charles Swanton’s clinical training was set to commence. He had just finished his preclinical work at St. Bartholomew’s and the Royal London Medical School when the 21-year-old opted to complete a yearlong cellular pathology program at University College London (UCL) Medical School and earn his bachelor’s degree.
There, learning about prior discoveries in cell biology from his professors, Swanton got a taste for the pursuit of scientific discovery. “I vividly remember the first lecture. The professor was setting up the overhead projector and slides, and I was...
I think it is harder to be a successful scientist without experiencing truly prolonged failures.
In the 1990s, researchers were beginning to understand the ways cells regulate transitions into the four phases of the cell cycle: G1, synthesis (S) phase, G2, and mitosis. One of the major figures moving this field forward was future Nobel Prize–winning geneticist Paul Nurse, who at that time was director of research at the Imperial Cancer Research Fund (ICRF, now Cancer Research UK) . Having joined Nic Jones’s group at ICRF as a UCL graduate student in 1994, Swanton heard Nurse give a lecture on this cell-cycle work. Swanton was stunned to learn how similar fundamental cellular processes were between humans and yeast, and how cell-cycle regulation is related to cancer development.
According to Swanton, his first 18 months in Jones’s lab were “an unmitigated disaster.” He had been trying to understand the functions of three related cyclin D family proteins, important for directing early cell-cycle transitions. But despite long days and late nights at the lab bench, he failed to produce any results. One Friday, Swanton recalls, his graduate advisor told him to hand in his midterm progress report on Monday morning. Soon after, with apprehension, he presented his failed experiments to his graduate committee, who recognized his efforts but also hinted that it might be time for Swanton to call it quits on his PhD degree.
“I remember thinking, ‘I am in so deep already, and I enjoy being in the lab and addressing problems relevant to human disease. What have I got to lose by plowing on?’” His committee asked him what he wanted to do and, quick on his feet, Swanton thought of a new project so that he could continue his PhD.
Cyclins form complexes with cyclin-dependent kinases (CDKs), unique to each cell-cycle phase, and activate specific genes to drive cells through their cycles. Swanton proposed to explore how one such protein, cyclin D, interacts with proteins called p21 and p27 that bind to the cyclin-CDK complexes and can inhibit progression.
With the help of Jones, structural biologist Neil McDonald, and cancer researcher Gordon Peters, Swanton spent the summer of 1994 mutating every surface amino acid of the human cyclin D protein individually to create a library of cyclin D–mutated proteins. He then measured the ability of his 40 cyclin Ds to interact with the inhibitory proteins in vitro. One mutant could still bind to its protein-binding partner, a CDK, but not to wildtype p21 or p27, suggesting that the mutation resulted in an always-active complex that drove constant cell division—a hallmark of malignant cells. “This was my first and only result in two years and made me realize again how fun science was,” says Swanton.
He then stumbled upon a paper that had identified a cyclin-like protein in a group of herpesviruses that could induce malignancy upon infecting mammalian cells. Swanton aligned the sequences of his mutated human protein and the viral one and found that they were nearly identical within the domain he had altered. Swanton decided to compare his mutated cyclin D to the viral protein in a test tube. He still remembers getting the result on a Saturday morning; the abilities of the two proteins to each bind to their CDK binding partners were the same, as were their inabilities to bind to either p21 or p27. “I called Nic on that Saturday to tell him the result. We both still talk about that phone call.”
The work, published in Nature, presented a novel way mammalian virus proteins evolved, adapting to resemble those found in mammalian cells. In this case, the viral protein managed to deregulate the cell cycle and induce oncogenesis.
For Swanton, both the lows and highs of his graduate career were valuable. “I think it is harder to be a successful scientist without experiencing truly prolonged failures. Only when you’ve been through the terrible stuff do you learn to unravel a problem and develop resilience.”
Discovery on the brain
Swanton was born in 1972 in Dorset, a southwest county in England on the coast of the English Channel, where his father was a cardiologist at a local hospital and his mother a historian. When he was two years old, his family moved to southwest London. “I was always into the outdoors—biking, cricket, and football—although I was not good at either team sport,” says Swanton. “One of the comments from a teacher in school went something along the lines of, ‘Charlie’s athletic contributions are rarely matched by his verbal ones.’ I tended to talk a lot and not do very much,” he says. “Some would say things haven’t changed!”
We showed that a single sample really dramatically underestimates the evolutionary complexity of a patient’s disease.
Swanton also says he was not a great student, uninspired by the way even subjects he was interested in were taught. He liked biology but was keen on discovery and experiments rather than the scripted lectures and textbook material his teachers presented. In the evenings, Swanton liked to build with Legos and other building sets. “I think that science is a bit like Lego building. You build yourself an ever bigger and bigger model and hope that it remains standing,” Swanton reflects.
After graduating from high school in 1990, Swanton took a year off and traveled. Afterward, he entered Bart’s and the London School of Medicine and Dentistry and began his premed studies. His attraction to oncology was solidified in his first year, when his father was diagnosed with a high-grade B-cell lymphoma. After rounds of chemotherapy and radiation, Swanton’s father was in remission, and 25 years later, still works in the UK’s National Health Service at the age of 74. “It’s really a remarkable story for 1991.”
Cancer treatment failure
After receiving his PhD in 1998, Swanton went back to clinical training. Having developed “an addiction to the lab,” he set out on the long path to becoming a physician-scientist.
Because of the many clinical training requirements in the U.K. Swanton didn’t return to the lab bench for another seven years, practicing general medicine, surgery, and neurology before specializing in oncology. “I missed the bench massively,” he says. “I enjoyed medicine, but it was treading water. I didn’t feel we were making progress.”
In 2004, he joined Julian Downward’s lab at the Francis Crick Institute in part to understand why cancer patients become resistant to standard treatments. Using an RNA interference screen, Swanton and his colleagues identified a set of genes involved in the regulation of mitotic arrest and in the metabolism of ceramides, lipid molecules abundant in cell membranes that influence whether tumors are sensitive to certain chemotherapy agents. The work also showed that tumors with high chromosomal instability, which can lead to tumor cell diversity through chromosomal rearrangements, are least sensitive to these chemotherapies.
See “Chromosomal Instability Drives Metastasis”
The project demonstrated to Swanton that lab work can inform why cancer drugs fail. “Cancer cells have this fast way of gaining and losing whole chromosomes, adapting in the face of cancer therapy,” says Swanton.
In 2008, Swanton set up his own lab at the Francis Crick Institute to study how chromosome instability can occur and how cancer cells can tolerate the genetic chaos that causes normal cells to self-destruct. Again, his first project didn’t go as well as he might have hoped.
The evolution of a cancer
Swanton’s new lab set out to identify specific genes that, when inhibited, result in the death of tumor cells that displayed aneuploidy, meaning they had more or less than the normal set of 46 chromosomes. But Swanton never found any such genes.
His first success came in 2012, when his lab provided an explanation of why cancer is such a difficult disease to eradicate. Swanton and his team took biopsies from four kidney cancer patients at various locations within the same tumors, and from metastases, at different times during their course of treatments. When the researchers sequenced the samples for genetic mutations and analyzed chromosome structure, they could trace the tumors’ evolutionary histories, much as evolutionary biologists trace the origins of organisms back to their common ancestors based on fossils deposited in different geologic eras.
Swanton dubbed the founding mutations in the original tumor that persist in most tumor cells “trunk” mutations, and subsequent alterations, present in only a proportion of tumor cells, “branches.” In all four patients, the investigators identified two lineages, one that seeded the metastasis from the original tumor and the other that allowed the original tumor to grow in place.
“It is very important in science not to claim you were the first in anything. The old adage that we stand on the shoulders of giants is so true. But I think the [tumor evolution] study was a bit of a wake-up call. People were relying very heavily on single-cell analysis to derive tumor information,” says Swanton. “We showed that a single sample really dramatically underestimates the evolutionary complexity of a patient’s disease.” The research could also explain why informative cancer biomarkers are difficult to identify—the tumors transform and change too much—and why rapidly mutating tumors find ways to grow despite aggressive treatment.
Swanton’s lab has since followed the work with comprehensive spatial and temporal genetic analyses of other tumor types, including colorectal cancer, showing the ways that faulty DNA replication promotes chromosomal instability in cancer. For Swanton, “these studies led to the idea that integrating genomics and cell biology could start to inform mechanisms of disease and ways to target those mechanisms with therapies.”
A turn to immunotherapy
As the cancer field saw successes with immunotherapies that could boost patients’ immune responses to cancer, Swanton turned to studying how heterogeneity within a tumor influences its interaction with the immune system. His lab demonstrated that neoantigens (mutated proteins unique to cancerous cells) present in most or all cells within a tumor are much more likely to be effectively recognized by the immune system. Additionally, the greater the number of these trunk neoantigens, the more likely the patient will respond to immune checkpoint inhibitor therapy, an antibody-based intervention that unleashes T cells to attack tumors.
Swanton is now focused on identifying the important trunk mutations present in most tumor cells. And, as a cofounder of Achilles Therapeutics, he’s working to commercialize these discoveries into adoptive, cell-based therapies.
This past October, led by postdoc Nicholas McGranahan and graduate student Rachel Rosenthal, Swanton’s lab found one way that lung cancers evolve to escape detection by the immune system. Forty percent of patient-derived tumors his team examined had tumor cells that stopped producing human leukocyte antigen (HLA)—a molecule necessary for the presentation of antigens on a cell’s surface, to be recognized by immune cells. The loss of HLA often occurred relatively late in the tumors’ evolution and resulted in an expansion of neoantigens within the tumors, predicted to bind to the lost HLA allele. If the immune system has no way to detect cancer-specific antigens, immune cells won’t be able to mount an attack.
“This loss-of-HLA mechanism suggests that, when designing vaccines and cell-based therapies, we need to target antigens that are presented to the immune system,” says Swanton. “The knowledge of which HLA molecules are lost will be critical to develop such effective therapies and choose the right antigens to target.”
Swanton’s goal is to map tumor evolution and adaptation over space and time. His plan is to sample thousands of tumors from hundreds of cancer patients across their disease course to track where immune checkpoint signaling molecules are distributed. The team will capture the tumor samples’ genomic and transcriptomic data as well as the clinical outcomes and drug responses of each patient, starting with 842 participants and the already-collected tumor biopsies from 3,000 tumor regions among them. Those data are part of the Tracking Cancer Evolution through therapy (TRACERx) program, which recruited Swanton’s patients with lung cancers. Swanton and his colleagues have expanded the program to include renal cancer and melanoma patients.
When not in the lab, Swanton spends time with his family, including his two daughters, ages 14 and 11, cycling, playing with their dog, or going to museums. His wife is an academic in gynecology, so dinner-table conversations typically turn to science and medicine.
Recently, Swanton gave a talk at his older daughter’s school on his recent work that uncovered HLA loss as a way tumors avoid being recognized by the immune system. He mentioned that tumor cells will perish if they lose all six copies of their HLA genes.
“At the end, a girl stood up and asked whether we should be targeting the HLA molecule in the tumor as a way to kill tumors, since tumors cannot lose all of their HLA genes. Here is a young student that applied interest and logic to a problem and came up with a solution—one that, I must confess, I hadn’t considered deeply enough until she asked me,” says Swanton. “That was a wonderful light-bulb moment that I witnessed. I think there is a scientist in all of us that is just waiting to be inspired.”