The study builds on earlier work by the same group, which had identified SARS-CoV-2 antigens showing both sequence and conformational homology to tumor-associated antigens, the molecular flags that tumors display on their surfaces and that the immune system can, in principle, attack. Cross-reactive T cells had already been detected in people who had either been infected with SARS-CoV-2 or vaccinated with the BNT162b2 mRNA vaccine. What remained unknown was the fine structure of that cross-reactivity: which T cell receptor genes were involved, what the hypervariable recognition loops looked like, and whether the same receptor architectures could be elicited by both the viral and the tumor-derived peptide. The new study set out to answer those questions directly.

To do so, the researchers selected a matched pair of epitopes presented by the same class of major histocompatibility complex molecules. The viral peptide, LLLDDFVEI, comes from a SARS-CoV-2 antigen, while its tumor counterpart, LLLDDLLVS, is derived from PRDX5, a tumor-associated antigen. The two nine-amino-acid sequences differ by only two residues, and both substitutions are conservative, meaning the overall shape and chemical character of the peptides are largely preserved. When processed and displayed on the cell surface in the peptide–MHC complex, the two epitopes present a sufficiently similar molecular surface that a single T cell receptor can plausibly engage both — the defining premise of molecular mimicry.

The experimental design was deliberately simple and controlled. Peripheral blood mononuclear cells were obtained from five healthy donors and cultured in the laboratory for ten days, during which the cells were stimulated every three days with 10 micrograms of one or the other of the two peptides. This ex vivo priming allowed rare cross-reactive T cells to expand to numbers that could be reliably detected and characterized. After the stimulation period, the researchers performed single-cell RNA sequencing paired with T cell receptor sequencing, a technique that captures both the transcriptomic identity of each cell and the full nucleotide sequence of its receptor’s alpha and beta chains, including the crucial CDR3 regions that make direct contact with the peptide–MHC complex.

The results, focused on CD8-positive effector T cells, revealed a striking pattern. In each donor, treatment with either epitope amplified or newly induced a small, reproducible set of TRAV and TRBV gene segments — the variable regions of the receptor’s alpha and beta chains — together with characteristic CDR3α and CDR3β amino acid motifs. Critically, the very same clonotypes appeared in both stimulation conditions: T cells expanded by the viral peptide carried the same receptor signatures as those expanded by the tumor peptide. These signatures were strictly individual, differing from donor to donor, and none of the identified motifs matched any entry in publicly available T cell receptor repositories, underscoring how personal the anti-viral and, by extension, the cross-reactive anti-tumor repertoire can be.

The cross-reactive clones were not confined to the effector compartment. The researchers found the identical clonotypes within the CD8-positive proliferating subset, indicating that both epitopes drove the same cells into a highly activated, dividing state. In immunological terms, this plasticity matters: it suggests that exposure to the viral epitope does not merely expand a narrow, terminally differentiated population but recruits versatile effector cells capable of further proliferation and functional diversification — properties desirable in a vaccine-elicited response intended to attack tumor cells presenting the mimic antigen in vivo.

To test whether the shared clonotypes reflected genuine structural cross-recognition rather than coincidental co-expansion, the team performed conformational prediction analyses of the peptide–MHC–TCR complexes. The modeled structures showed a perfect structural overlap between complexes built on the viral epitope and those built on the tumor epitope. In other words, when the same receptor docks onto either peptide presented by MHC, the geometry of the interaction is essentially indistinguishable. This structural convergence provides a mechanistic explanation for the functional cross-reaction observed in the cell cultures and elevates the finding from a correlational observation to a validated molecular event.

The authors emphasize the functional significance of the result in light of a well-known feature of T cell biology: degeneracy. A single T cell receptor can often recognize multiple different peptides, and a single epitope can be recognized by many different receptors. Given this inherent promiscuity, finding identical CDR3αβ motifs elicited by two homologous but non-identical epitopes is, as the researchers describe it, of the highest functional relevance. It demonstrates that the mimicry between the SARS-CoV-2 peptide and the PRDX5-derived tumor peptide is tight enough to select for the same receptor solutions across independent stimulation conditions and across genetically distinct donors’ repertoires.

The translational implication is the most consequential part of the work. Tumor-associated antigens are self antigens, and the immune system is generally tolerant of them, which has long hampered therapeutic cancer vaccines: self-derived peptides tend to be weak immunogens that expand low-affinity or exhausted T cells, if they expand anything at all. Microbial epitopes, by contrast, are foreign, and the immune system responds to them vigorously. The study’s results provide, for the first time, a clear experimental validation that microbial epitopes mimicking tumor-associated antigens can be used to develop off-the-shelf preventive or therapeutic vaccine formulations. A vaccine built on a viral mimic would exploit the immune system’s readiness to mount strong responses against non-self sequences while redirecting the resulting effector cells, through cross-reactivity, against tumor cells displaying the homologous self antigen.

Several caveats and open questions remain before such a strategy reaches the clinic. The study involved five healthy donors and a single epitope pair, so the breadth of the mimicry effect across other tumor antigens, HLA backgrounds, and patient populations will need to be established. Whether vaccine-elicited cross-reactive T cells can actually infiltrate tumors and mediate clinically meaningful anti-tumor activity in vivo is the decisive next test, and the risk that broadly cross-reactive responses could target healthy tissues expressing PRDX5 must be carefully evaluated. The work was funded by the Italian Ministry of Health through institutional research programs and a PNRR project on molecular mimicry to improve liver cancer immunotherapy, signaling that hepatocellular carcinoma is one intended application. Even with those qualifications, the study marks a milestone: it converts molecular mimicry from an epidemiological curiosity — the observation that some infections correlate with altered cancer risk — into a precisely characterized receptor-level mechanism that can be engineered on purpose. If follow-up studies confirm that viral-mimic vaccines can reliably expand functional, tumor-directed CD8 T cells, the pandemic coronavirus that reshaped global health may also leave behind an unexpected gift for oncology.

Subject of Research: Cross-reactivity between SARS-CoV-2-specific T cells and tumor-associated antigens via molecular mimicry

Article Title: SARS-CoV-2-specific T cells cross-react with tumor-associated antigens via molecular mimicry

Article References: Ragone, C., Mauriello, A., Cavalluzzo, B., Mangano, S., Cembrola, B., Ciotola, N., Tagliamonte, M., & Buonaguro, L. (2026). SARS-CoV-2-specific T cells cross-react with tumor-associated antigens via molecular mimicry. Journal of Translational Medicine, 24(1), Article 1179. https://doi.org/10.1186/s12967-026-08905-5

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08905-5

Keywords: SARS-CoV-2, molecular mimicry, tumor-associated antigens, T cell receptor, cross-reactivity, cancer vaccine, CD8 T cells, single-cell RNA sequencing, PRDX5, immunotherapy, epitopes, Journal of Translational Medicine