Tumor suppressors gone rogue: a new class of targets for TCR-based cell therapies

One of the central challenges in developing T cell-based cancer immunotherapies is finding targets that are consistently abundant on tumor cells but absent from healthy tissue, a distinction that is important for therapies seeking to eliminate cancer cells without causing serious toxicity to the patient. Most efforts have focused on mutated proteins unique to tumors, or proteins that are highly expressed by cancer cells. A new study published in the Journal for ImmunoTherapy of Cancer by Tom Schmitt and colleagues in the Chapuis and Greenberg labs at Fred Hutch takes a counterintuitive approach: targeting proteins that normally suppress cancer but end up highly expressed due to broken negative feedback loops.

The proteins the researchers focused on are p16^INK4A and p14^ARF, both encoded by a gene called CDKN2A. Under normal circumstances, these proteins act as cell cycle gatekeepers, sensing abnormal growth signals and triggering cell cycle arrest or programmed cell death to avoid uncontrolled growth. But in many tumors, the downstream targets these of proteins, such as the tumor suppressors Rb and p53, are mutated or deleted. When that happens, the normal feedback loop that keeps p16^INK4A and p14^ARF in check gets disrupted, and these proteins accumulate to abnormally high levels. In particular, p14^ARF is overexpressed in most tumors carrying mutations in p53, practically all HPV-positive cervical cancers and more than two-thirds of breast cancers. Importantly, in healthy adult tissues, p14^ARF is barely detectable.

“We usually think of tumor suppressor proteins as the ‘good guys’ that protect cells from becoming cancerous, so they are not the kinds of molecules people typically consider targeting therapeutically. But this work highlights an important paradox in cancer biology: when downstream pathways are disrupted, tumor suppressors like p14^ARF can become highly overexpressed and ineffective at stopping tumor growth. In that setting, the very proteins that failed to suppress the cancer may become useful antigenic targets for T cell therapy,” lead author Tom Schmitt explains.

The researchers sought to address whether p14^ARF could be exploited as a therapeutic target. The strategy they pursued is called TCR-T cell therapy: engineering a patient's T cells to carry a T cell receptor (TCR) that recognizes cells expressing p14^ARF, which directs the immune system to destroy these cells. The challenge is finding the fragment of the target protein that gets displayed on the cancer cell's surface by human leukocyte antigen (HLA, the human form of MHC I) in a way the engineered T cells can recognize.

The team started by identifying fragments computationally predicted to bind HLA-A2, a common allele that cancer antigens are presented on. After systematically screening fragments of both p16^INK4A and p14^ARF, the team identified one fragment of p14^ARF, a nine-amino-acid stretch (noted ARF_35-43), that consistently appeared on the surface of tumor cells from a range of cancer types but was absent from normal tissue samples. When this peptide was loaded onto dendritic cells for antigen presentation to donor T cells, T cells were efficiently stimulated and expanded. They then used fluorescently labeled reagents that carry the specific peptide-HLA-A2 combination to isolate high-affinity T cell receptors capable of recognizing this fragment, engineered T cells to carry these receptors, and put them to the test.

In cell-based experiments, the engineered T cells efficiently killed cancer cell lines from colon, pancreatic, lung, cervical, liver, and breast cancers. Importantly, these experiments showed that since T cells responded to the tumor cells without any peptide being artificially added, this means ARF_35-43 is naturally processed by the proteasome of target cells and presented by HLA-A2. In mouse models, two injections of the engineered T cells significantly reduced tumor burden in both colon and pancreatic cancer models. The team also analyzed tumor samples from 20 cervical cancer patients, finding the ARF_35-43 fragment displayed on the surface of cancer cells in 16, at levels consistent with other successfully targeted tumor antigens.

Perhaps the most reassuring finding came from safety testing for any potential off-target effects. Using a mouse model in which p14^ARF-expressing cells were tagged with a fluorescent marker, the researchers confirmed that T cells directed against ARF-expressing cells caused no detectable effects on normal healthy tissues, consistent with low p14ARF expression in normal tissues and reiterating the clinical promise of this approach.

Schmitt looks to the future implications of this work: “These findings also raise an interesting question beyond cancer: in what other settings is ARF expression a marker of cells that should be eliminated? p14^ARF is highly expressed in many advanced tumors, but it can also be induced in damaged or senescent cells. That suggests that therapies targeting p14^ARF may eventually have applications outside oncology, for example in diseases where p14^ARF-expressing senescent cells contribute to tissue dysfunction.  One example is diabetic kidney disease, where chronic cellular stress causes p14^ARF-mediated senescence, leading to tubular cell damage and progressive renal fibrosis.”

The study unveils exciting new possibilities for tumor suppressors in cancer immunotherapy and beyond.

Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Members Drs. Aude Chapuis and Philip Greenberg contributed to this research.

The spotlighted research was supported by funding from Cullinan Therapeutics (Fred Hutch Reference Number: SRA191002).

Schmitt TM, Furiya K, Black C, Vazquez A, Sharma J, Hailemariam M, Paushter DH, Trieu L, Lam J, Lee B, Rakhra K, Whalen KA, Mehta NK, Sauer K, Baeuerle PA, Michaelson JS, Greenberg PD, Chapuis AG. 2026. Dysregulated expression of the tumor suppressor p14ARF in cancer provides an effective target for TCR-T cell therapeutics. J Immunother Cancer. doi: 10.1136/jitc-2025-013520.