Cytokines (proteins that help coordinate immune activity) have long shown promise as cancer treatments. Their ability to activate and reshape immune responses can drive powerful anti-tumor effects. Yet many cytokine therapies have been difficult to use in patients because the same immune signals that attack tumors can also trigger severe inflammation and other toxic side effects in healthy tissues.
Stephanie Dougan, PhD, Associate Professor of Immunology at the Dana-Farber Cancer Institute and Member of the Hale Family Center for Pancreatic Cancer Research, is working to overcome that barrier. Leveraging advances in artificial intelligence and computational protein design, she and her collaborators are creating synthetic cytokines, which are engineered versions of immune signaling proteins designed to amplify anti-tumor immunity with greater precision and fewer side effects. Their goal is to solve one of cancer immunotherapy’s longstanding challenges: harnessing cytokine potency without unleashing harmful systemic inflammation.
Preserving Immune Activity, Reducing Toxicity
Together with her husband, Michael Dougan, MD, PhD, a Dana-Farber faculty member, gastroenterologist at Beth Israel Deaconess Medical Center (BIDMC), and expert in immunotherapy toxicity, she is developing a platform for designing cytokines with properties not found in nature. Their aim is to help patients benefit from cytokine-driven immune stimulation while minimizing the toxicities that have historically limited these therapies.
“What attracted us to cytokines was their ability to remodel an entire immune response,” Michael Dougan explains. “You aren’t simply targeting one immune cell. You’re changing the broader immune environment.”
Redesigning Cytokines for Cancer Therapy
Yet the very features that make cytokines powerful can also make them dangerous. That’s why the Dougans have focused on redesigning these proteins to produce stronger, longer-lasting anti-tumor effects with improved safety.
“Once you learn how to engineer one cytokine successfully, you can begin thinking about how those same approaches might be applied more broadly.”
Stephanie Dougan, PhD
They partnered with David Baker, PhD, and colleagues at the University of Washington’s Institute for Protein Design, leveraging advances in AI-based protein design that make it possible to predict how proteins will fold and interact.
This enabled the researchers to create immune signaling proteins with greater control over how and when immune responses are activated, an achievement that would be difficult or even impossible to accomplish using naturally occurring proteins alone.
Building a Platform
In 2023, the team reported its first major advance in Nature Biotechnology. They engineered a mimic of interleukin‑2 (IL‑2), a cytokine that activates and expands cancer-fighting T cells. They designed the protein so it could be split into inactive components and reassembled only under specific conditions, offering a new level of control over cytokine activity.
Two years later, they published a second breakthrough in Science Immunology, this time focusing on interleukin‑21 (IL‑21), a cytokine that helps sustain long-term immune responses. Although IL‑21 has shown promise as an immunotherapy, its activity can be difficult to maintain long enough to produce durable clinical benefit.
The team developed 21h10, an engineered IL‑21 variant with greater stability and activity than the natural protein. They then tested 21h10 in pancreatic cancer, where tumors frequently develop resistance to KRAS inhibitors.
“For patients, that resistance can limit the long-term effectiveness of targeted therapies and leaves few treatment options once tumors begin growing again,” Stephanie Dougan says.
In a study recently accepted by Cell, the Dougans along with collaborators from the Baker lab and the Hale Family Center for Pancreatic Cancer Research combine 21h10 with KRAS-targeted therapy and found that the engineered cytokine converted temporary treatment responses into long-term remissions in preclinical models, including tumors that had already developed resistance. Immune cells isolated from human pancreatic tumors also responded to 21h10 in ways consistent with the anti-tumor activity observed in these studies.
Collectively, these findings demonstrate that cytokines can be redesigned to improve stability, activity, and precision, Stephanie Dougan stresses.
Taking a Broad Approach
She also points out that she and her colleagues view synthetic cytokines not as a single therapeutic strategy but as a platform for designing a new generation of immune-modulating therapies.
“Once you learn how to engineer one cytokine successfully, you can begin thinking about how those same approaches might be applied more broadly,” Stephanie Dougan says.
Therefore, rather than focusing on a single candidate, the team is applying these design principles across multiple cytokines and cancer types. They are developing next-generation cytokines that activate only within specific biological environments—such as tumors—helping concentrate immune activity where it is needed while limiting effects elsewhere in the body.
Their approach has shown activity in preclinical models of pancreatic cancer, KRAS‑mutant lung cancer, melanoma, and colorectal cancer. Michael Dougan notes that the team is now exploring whether these design principles can be extended to additional cytokines and immune pathways.
Moving Toward Patients
The researchers are now working to advance their most promising candidates, including 21h10, toward clinical testing.
“We’ve shown activity in mice, and we’ve shown that human immune cells respond,” Michael Dougan says. “The next step is determining whether we can achieve those same benefits safely in patients.”
Ultimately, the goal is to create safer, more effective immunotherapies for people with cancer. If successful, this work could establish a new class of programmable immune therapies—cytokines tailored to specific cancers and treatment settings, capable of delivering more durable anti-tumor responses with fewer side effects.
Read the full publication on Cell.
Team Members: Stephanie Dougan, PhD, Michael Dougan, MD, PhD, David Baker, PhD