AACR IO 2026 Keynote Highlights: The Tale of Tregs Told by Nobel Laureate Fred Ramsdell

Nobel laureate Fred Ramsdell, PhD, helped reveal that FOXP3 is the master switch for regulatory T cells (Tregs)—the immune system’s peacekeepers—and that breakdowns in this pathway can unleash damaging autoimmunity. He famously missed the prize announcement itself, backpacking far from cell service when Stockholm called. At the closing keynote of the second annual AACR Immuno-Oncology Conference (AACR IO) in Los Angeles, Ramsdell walked the audience through the full arc of his discovery, from a janitor’s closet repurposed as a mouse facility in the mid-1990s to a clinical trial now showing the first signs that Tregs infused into patients with severe rheumatoid arthritis (RA) can dampen a disease that years of drugs had failed to control.

A Walk Down Memory Lane: Discovering Regulatory T Cells

It was, by design, a historical talk. “I’m going to put everybody in the ‘wayback machine’ a little bit,” Ramsdell told the crowd, acknowledging that he had sat through the conference’s opening keynotes by Elizabeth Jaffee, MD, FAACR, and Philip Greenberg, MD, FAACR, and arrived at the closing session in a state of what he cheerfully described as imposter syndrome. “As I sat through those two spectacular talks from two amazing scientists, I was really regretting my choice and wondering why I’m at this meeting.” Jaffee and Greenberg had wrestled with how to build better T cells for attacking cancer, and Ramsdell’s decades of work had centered on the opposite question: How does the immune system know when to stop? And how can we harness that braking system—damaged in autoimmunity, weaponized by tumors—to treat disease?

The connection between damaged autoimmunity and Tregs came from a single experiment. Tregs are a population of cells that had been described in 1995 by Shimon Sakaguchi, MD, PhD, who shares the prize with Ramsdell and Mary E. Brunkow, PhD. Tregs remained at the margins of mainstream immunology, partly because the field had been burned by a previous generation of “suppressor T cell” research in the 1980s. In August 1998, a colleague in Ramsdell’s group, Roli Khattri, PhD, head of immunology at Expedition Medicines, showed that FOXP3 was exclusively expressed in the CD25-high CD4 T cells that Sakaguchi had characterized. No other cell in the hematopoietic lineage expressed FOXP3. “This is when I knew,” Ramsdell said, “whatever’s happening in the peripheral immune system, this cell is the mediator of that peripheral tolerance—that was actually really exciting.”

To investigate what they could do with this information in a human system, Ramsdell and his team collaborated with clinical researchers studying a rare pediatric syndrome called IPEX—immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome. They found mutations in FOXP3 in affected families. Boys with IPEX had many of the features that had been observed in scurfy mice, a mouse model developed to study severe autoimmune disease. Just like the scurfy mice, they developed aggressive autoimmunity in multiple organs, required massive immunosuppression to survive into their teens, and carried mutations in the same gene at the same chromosomal location. Every coding mutation in FOXP3 the researchers could find resulted in IPEX. The biological impact was highly conserved between mice and humans. FOXP3-expressing Tregs, it was now clear, were the central checkpoint enabling peripheral immune tolerance in both species.

What Ramsdell recognized almost immediately, even in the late 1990s with his experiments with mice in a janitor’s-closet-turned mouse facility, was that this discovery pointed in two therapeutically opposite directions. “If you could make more of these cells, or make them more active, maybe we could treat autoimmune disease,” he said, “on the flip side—if we could inhibit the activity of these cells in the context of the tumor, maybe we could get a better immune response to the tumor.” 

The relationship between Tregs and cancer immunotherapy remains one of the field’s sharpest double-edged swords: The same cells that protect tissues from autoimmune destruction can shelter tumors from the immune system that could otherwise eliminate them.

In 2000, however, neither direction was therapeutically actionable. Gene therapy was not ready. Cell therapy for autoimmunity was not being done. Ramsdell’s group had a deep understanding of the biology and no practical path to the clinic. “That was incredibly frustrating to all of us,” he said.

The Scientific Community Connects the Dots

What changed over the following two decades was the rest of the field filling in the mechanistic picture. Researchers mapped the T-cell receptor (TCR) repertoire of Tregs and found it distinct from conventional T cells. They showed that Tregs home differently, use different survival signals, suppress immune responses through multiple parallel mechanisms—consuming IL-2 to starve effector cells, deploying TGF-beta, remodeling the metabolic program of the local microenvironment—and play important roles in tissue repair through factors like amphiregulin and neuregulin, which actively restore damaged tissue rather than simply quieting inflammation. “The way I think about it is I don’t have to be smart,” Ramsdell said. “This cell is smart, this cell knows what it’s doing.”

Over time, the cancer immunotherapy field established cell therapy as a viable clinical modality. “I always thank my oncology colleagues because they convinced the rest of the world that cell therapy is a reasonable, viable way to treat disease,” Ramsdell said. Once chimeric antigen receptor (CAR) T cells, tumor infiltrating lymphocyte therapy, and TCR-modified T cells had moved from concept to approved products, the infrastructure, the regulatory experience, and the institutional willingness to infuse engineered T cells into patients existed in a way it simply had not before. The challenge of manufacturing therapeutic regulatory T cells, once seemingly insurmountable, became a tractable engineering problem, which gave Ramsdell the opportunity he needed.

Treg-based Therapy Goes From Idea to Innovation

In 2019, Ramsdell cofounded Sonoma Biotherapeutics, with Jeffrey Bluestone, PhD, managing director at Vie Ventures; Alexander (Sasha) Rudensky, PhD, FAACR, Sakaguchi’s former collaborator and chair of the immunology program at the Memorial Sloan Kettering Cancer Center; and others. The company was centered on an approach to isolate regulatory T cells from a patient’s blood, expand them ex vivo in carefully controlled conditions, engineer them to carry a CAR, and infuse them back in the patient. Isolation is the first hurdle: Unlike in mice, where CD25-high CD4 T cells are a cleanly demarcated population, human Tregs must be CD127-low, CD25-high, and CD4-positive cells.

“If you sort incorrectly and get a less-than ideal purity of cells, these other cells will outgrow the Treg cells quite easily,” Ramsdell explained. Once sorted, the cells expand reliably, maintain their FOXP3 expression and Treg-associated gene signature after transduction with the CAR construct, and remain stable under pro-inflammatory conditions designed to mimic the cytokine environment of a RA joint—conditions that might otherwise push T cells toward inflammatory phenotypes.

For the CAR target, Ramsdell and his team chose citrullinated proteins, a class of self-antigens strongly associated with RA. In RA, an enzyme converts arginine residues in normal proteins like vimentin into citrulline, generating neo-antigens that the immune system can mistakenly attack. Citrullinated peptides and proteins are found in abundance in RA joints, and anticitrullinated protein antibodies are among the most specific biomarkers of the disease.

Ramsdell’s team screened hundreds of antibodies against these antigens in collaboration with researchers at the Karolinska Institutet, selected one that recognized citrullinated vimentin as well as other citrullinated proteins including histones, and built it into a CAR with signaling domains optimized for Treg function rather than the standard CD28 or 4-1BB configurations used in effector CAR T cells. The product they developed can be activated by synovial fluid from RA patients—a crucial in vitro demonstration that the relevant antigen is present and accessible in the joint environment.

The deuterium-labeling studies Ramsdell showed demonstrated persistence: CAR Tregs infused into patients could be detected for up to a year in circulating blood, exclusively within the Treg compartment, with no evidence of conversion into other lymphocyte lineages. This stability matters, explained Ramsdell, for a therapy that is conceptually trying to reestablish durable immune tolerance rather than provide temporary symptom control.

Early Clinical Signals With Tregs

The clinical data Ramsdell presented was early—nine patients dosed so far across three cohorts in a standard dose-escalation design, all women, all with severe and refractory RA. Several had disease that six or more distinct mechanistic classes of drugs had failed over years of treatment. The primary endpoint, as in any phase I trial, was safety, and Ramsdell reported no adverse events in any patient at any dose level based on current data.

Ramsdell also reported preliminary efficacy signals. Counts of swollen and tender joints—two standard clinical measures in RA—declined in most patients following infusion, with a dose-response pattern visible between the low and higher dose cohorts. Ultrasound imaging of joints, which detects inflammatory blood flow in real time, showed reduction in synovial inflammation by week four in responding patients. Data from biopsies confirmed that the infused CAR Tregs could be detected inside joint tissue—expressing the CAR, FOXP3, and CD4—weeks after infusion. Inflammation scores dropped from nine to one in one patient at the highest dose.

Ramsdell explained observations from patient experiences as well. One patient, who had not been able to run for five years, went for a run after feeling better from the treatment—a signal of something working. Another could open a jar of pickles for the first time in years—”not a validated endpoint,” Ramsdell acknowledged with evident warmth, “but great for that patient, especially if they liked pickles.” These are patients who had exhausted their options—all of them did not respond to anti-TNF therapy, most also did not respond to JAK inhibitors, rituximab (Rituxan), and abatacept (Orencia) across a dozen or more drug regimens spanning years. “Some of these patients have been suffering for a long time from this disease,” Ramsdell noted.

Whether the responses will prove durable, and at what dose, remain open questions. Ramsdell was direct about the limitations: The numbers are too small for statistical conclusions, the biopsy data is sparse, and a therapy requiring the isolation and expansion of a patient’s own cells cannot realistically be administered every two weeks. “We may actually be able to cure these patients,” he said. “It’s aspirational, but it’s certainly something I think we’re trying to achieve.”

Knowledge Begets Knowledge: Exploring Tregs to Treat Cancer

While Ramsdell and colleagues are taking their first steps into bringing CAR Treg technology into direct clinical applications, the brilliant scientific discoveries made in this field are now informing others. One such key area being cancer—the flip side of Treg research that Ramsdell mentioned.

For instance, this blog post about a study presented at the AACR Annual Meeting 2024 describes how researchers have used Treg’s infiltrative powers toward killing tumors cells in mice. A study in the AACR journal Cancer Discovery shows what happens when you lift the Treg brakes on the immune system: In a small clinical trial of a human papillomavirus‑linked throat cancer, a treatment regimen of two checkpoint drugs decreased the effector Treg levels and led to complete metabolic response in 94% of the patients in the study.

The tale of Tregs isn’t over yet—in oncology, it has only just begun.