A research team searches for every gene that helps tumors evade immunotherapy
Efforts to probe the entire genome of cancer cells in animals have led to potential new drugs — now in early-stage clinical trials — that could make cancer immunotherapy more effective for more patients.
In 2011, Robert Manguso was working in a cell biology lab when his mother was diagnosed with Merkel cell carcinoma, a rare and aggressive skin cancer.
Manguso, who’d recently graduated from college and was conducting research at the University of Copenhagen as a Fulbright scholar, moved back to the Boston area to be with his mother as she underwent treatment.
He also read everything he could about her disease, including emerging evidence that suggested that the immune system could recognize and kill Merkel cell carcinoma. His mother had a presentation of the disease that suggested her immune system was already on the job. But immunotherapy was not yet widely used and had not been applied clinically to Merkel cell carcinoma, so she received traditional chemotherapy and radiation therapy, suffering life-threatening complications along the way.
Though the treatment eliminated her cancer, Manguso was struck by the promise of immunotherapy, which offered the potential to harness the immune system to fight cancer without the harsh side effects his mother had experienced. As he searched for graduate school programs that same year, Manguso decided to move away from his interest in fundamental cell biology and focus instead on cancer immunotherapy.
“When I began my PhD I completely switched my focus, and I haven’t looked back,” said Manguso. “That’s the only thing I was interested in doing.”
He soon joined the lab of W. Nicholas Haining at the Dana-Farber Cancer Institute as a graduate student. By then, a few cancer immunotherapies had entered clinical trials and only a handful had been approved by the FDA. It was already becoming clear that some cancers responded better than others to immunotherapy. Manguso had an ambitious goal: using CRISPR-based tools on an almost unprecedented scale to find genes that might explain why.
As a first step, Manguso set his sights on treatments called PD-1 checkpoint inhibitors, which at the time had shown tremendous promise in clinical trials and would receive their first FDA approval in 2014. The drugs work by blocking a protein, PD-1 — which helps keep immune responses in check — and stimulating the immune system to attack tumors. They are used to treat cancers including melanoma and non-small cell lung cancer, but work only for a small fraction of patients.
Over the next five years, Manguso collaborated with members of Broad’s Genetic Perturbation Platform to use CRISPR-Cas9 genome editing to systematically delete nearly 2,400 tumor genes in mice, in parallel, and look for those that made melanoma tumors more or less sensitive to treatment with PD-1 inhibitors.
The team found that deleting the gene Ptpn2 made tumors sensitive to PD-1 therapy, and they published their findings in Nature in 2017. The researchers soon partnered with Calico Life Sciences, which in collaboration with AbbVie, discovered drug candidates that block the PTPN2 protein. By early 2021, Calico and AbbVie had launched two early-stage clinical trials testing two PTPN2 inhibitors, either as a single therapy or in combination with existing PD-1 immunotherapies.
As Manguso was wrapping up his PhD work in 2017, he, Haining (now at Arsenal Biosciences), and Kathleen Yates, an immunologist in Haining’s lab with a track record of executing large, complex projects, began to think big about the approach they were taking to discover new immunotherapy targets. They talked about how they might expand Manguso’s approach to cover the entire genome. Could they test the effect of every gene, in every cancer, to find every mechanism of resistance to immunotherapy?
“We wanted to see whether there were as yet undiscovered drug targets that would lead to amazing cancer immunotherapies,” Manguso said.
Together, Haining, Manguso, and Yates went on to found the Tumor Immunotherapy Discovery Engine (TIDE) at the Broad Institute of MIT and Harvard in 2017 with support from Calico as a collaborating partner. Today, Manguso and Yates co-lead a team of 22 systematically working towards Manguso’s goal. They scaled up their CRISPR screening approach to look across the whole genome in lab animals. Using that technology, they have already probed the role of every gene in immunotherapy resistance across eight cancer models. They’ve developed a new method to improve CRISPR screens in animal models, and have identified several new mechanisms of immune evasion shared by multiple cancers.
Calico, together with AbbVie, has worked with TIDE to advance some of its drug discovery and development programs and now has projects — including the PTPN2 inhibitors — in various stages of preclinical and clinical development. Along the way, TIDE is showing how careful, systematic science at scale can quickly home in on promising drug targets.
“Every part of TIDE has been the fruit of collaboration and multidisciplinary team-based science,” Yates said. “Everybody just really wanted to help each other succeed. We’re deeply grateful for that.”
Studying the impact of a few thousand genes on a single cancer is challenging enough in an animal model. But scaling up those experiments in mice to probe the entire transcriptome — the expression of some 20,000 protein-coding genes — was another task entirely.
“We weren’t 100 percent sure we could do it,” Manguso remembered. “Most people would’ve considered even one of the experiments we were proposing to do basically impossible.”
Yates, however, was more confident. She’d been fascinated by immunology since middle school. After college, she spent two years working for a biotechnology company before joining Haining’s lab, where she rose through the ranks from technician to senior scientist. She’d helped Haining build his lab and manage large, ambitious projects, and had a good idea what it would take to scale up Manguso’s experiment.
When Manguso and Yates approached different scientists at Broad, they were struck by how excited everyone was by their idea. Together, they talked through all aspects of the project, such as how long it might take to optimize the technology, what the experiments might look like at scale, and how the researchers could measure progress.
“There was an optimism and enthusiasm for being able to do things that seemed out of reach for academic research environments that was infectious,” said Manguso.
With the help of Calico, the TIDE team first decided to work on expanding their protocol to probe the entire genome. At the time, most people studying how cancers evade the immune system focused on a targeted list of genes in one or two mouse models, or they used genome-wide screens in cell culture. But genome-wide screens in animals were much more challenging: The team would need to fine-tune a swath of parameters, from the number of cancer cells they could graft in animals, to the length of time the tumors would grow in animals, to the number of cells they’d need to collect from the animals to study every gene in a statistically significant way.
Yates remembers many conversations with scientists on their team and in the Genetic Perturbation Platform, designing and troubleshooting pilot experiments.
“This wasn’t something we could automate,” she said. “It just took a lot of people crammed in a small room, working on their feet all day, learning about ways we could make the system more efficient.”
“Pretty much anyone can do one screen, but it takes both long-term scientific vision and a true dedication to operational excellence to pull off what the TIDE team accomplished,” said John Doench, director of research and development in the Genetic Perturbation Platform.
Using CRISPR screening in mice presented additional challenges. The animals’ immune systems often recognize and reject large proteins like Cas9 — the enzyme “scissors” of CRISPR that helps deactivate genes — making it difficult for researchers to efficiently turn genes off in the mice.
The team studied the immune response to CRISPR and designed a new technology that they published in 2021 — a way of removing immune-triggering molecules from CRISPR-edited cells. The method prevents the mouse’s immune system from rejecting the CRISPR machinery while also leaving behind RNA barcodes that can be sequenced to determine which gene had been deactivated in each cell.
“I heard from people about this paper as often as any of our target discovery work, because it solved a problem many were encountering but that nobody was really talking about,” Yates said.
Across every gene
With the addition of their new technology, the TIDE team finally had an approach they could use to screen whole genomes of mice.
First, the researchers introduce into cancer cells a library of “guide” RNAs that target specific genes of interest. These cells also express the Cas9 protein, which helps disable the genes targeted by the guide RNAs. The scientists then grow the cells, use the technology they developed to remove immune-triggering molecules, and implant the cancer cells in mice that have been engineered to mount one of three different levels of an immune response. Some mice have a disabled immune system, others have a functioning one, and a third group, also with a functioning immune system, receives the immunotherapy of interest. Next, the researchers extract the tumors from the animals and use the barcode system they designed to see which genes have caused cells to proliferate or die off.
TIDE has used their method to study mechanisms of immune evasion in eight different cancer models from five different tissues. Across the models, the team found both genes that help tumors evade the immune system and those that block the immune system from attacking cancer cells. The scientists were surprised to discover that an immune-signaling molecule called interferon-γ (IFNγ), which typically helps the immune system fight tumors, in some cases also helps tumors avoid immune attack. The findings suggest a complex, context-dependent relationship between IFNγ signaling and the anti-tumor immune response.
Overall, the work revealed both cancer-specific resistance pathways and mechanisms of immune evasion that are shared by multiple cancers and could be targets of treatments that work for a variety of cancers. The researchers say that in the future, doctors could look for these features in a pre-treatment biopsy from a patient to help predict a tumor’s resistance to a given treatment.
But Manguso knows that to find every possible mechanism by which tumor cells evade the immune system, he and his team will have to expand their strategy even more. In addition to their CRISPR screens that turn genes off, TIDE is also using screens that activate genes and increase their expression in cells, and has already found new connections between genes and resistance to immunotherapy. They also plan to look for synthetic lethalities — pairs of genes that result in cancer cell death only when both are mutated and are promising drug targets. Finally, because human immune pathways don’t all translate to mice, the team hopes to screen human cancer and immune cells.
Though Manguso and Yates have a lot of work ahead of them, they maintain the enthusiasm and energy that helped them launch and build TIDE.
“If you think about how long humans live and the systems we've evolved to navigate all the pathogens we encounter, the immune system has understandably had to become more complex as well,” Yates said. “I just think it’s incredible. There’s always more to learn.”
The power of the immune system
Today, more than a decade after her diagnosis with Merkel cell carcinoma, Manguso’s mother is thriving, and her story is still inspiring Manguso to look for better ways to harness the immune system to fight cancer. Merkel cell carcinoma is now known to be among the most sensitive types of cancer to immunotherapy, and Manguso thinks his mother’s immune system, even without receiving any immunotherapy at the time, was critical to her continuing remission.
“The statistics suggested that with chemo-radiotherapy alone her chances of long-term remission were low, but we know her immune system was already involved before she began receiving treatment,” Manguso said.
He remains struck by the contrast between the harshness of the chemotherapy and radiation his mother received and the promise of curative therapy when the immune system is involved.
“That has really left a lasting impression and continues to motivate me,” Manguso said. “I am convinced that durable remission is only possible by engaging the immune system. To me, my mom is evidence of that.”