← Back to Home

More Stable T-Cell Receptor Pairing: 18 Modifications Improve Experimental Anticancer Activity

An Israeli team used computational protein design to improve the shared scaffold of T-cell receptors, reproducing functional improvements across six receptors and improving tumor control in mice. Key questions remain about whether the approach can be used safely in humans.

By SURL BioNews

Introducing a receptor that recognizes cancer cells into an immune cell does not guarantee that it will work properly. The new receptor may assemble poorly or even pair incorrectly with the cell’s existing components, weakening its ability to attack. An Israeli research team targeted this step by improving the structure of T-cell receptors (TCRs), enhancing antitumor activity in cell experiments and mouse models.

The study was published in Science Advances on September 23, 2026, and was a collaboration between Cyrille Cohen’s laboratory at Bar-Ilan University and Sarel Fleishman’s computational protein design team at the Weizmann Institute of Science. Reports from The Jewish Chronicle and the university both noted that the findings remain limited to preclinical research, with efficacy in humans yet to be demonstrated.

TCRs typically consist of two protein chains, α and β. Their variable regions recognize targets, while their constant regions help with assembly and connection to the signaling complex. When chains from an introduced receptor pair with the cell’s existing chains, they can reduce surface expression of the intended receptor and potentially create unintended recognition capabilities. Ensuring that the correct two chains preferentially pair is therefore an important hurdle in engineering T cells.

The team used the PROSS and FuncLib protein design algorithms to improve protein stability and the interface between the two chains, respectively. Experimental screening then yielded a “structurally enhanced TCR” (SET) containing 18 amino acid substitutions. The modifications were concentrated in the constant regions, with the aim of allowing receptors with different recognition functions to share the same structural adjustments.

The study reproduced functional improvements across six TCRs, including receptors that recognize cancer-associated antigens and antigens from Epstein–Barr virus and SARS-CoV-2. The modified cells showed stronger cytokine secretion responses, and anticancer experiments also showed increased killing activity. However, the experimental responses to viral antigens do not yet establish that the approach can treat viral infections.

In experiments involving mice implanted with human melanoma cells, with seven mice per group, the SET group’s mean tumor volume on day 83 was 35% smaller than that of the groups receiving the existing modified version D14 or wild-type receptors. On day 127, all mice in the SET group were alive, compared with fewer than half in the other comparison groups. These data show improved tumor control in a specific model and cannot be directly translated into a survival benefit for patients.

SET’s appeal lies in the possibility that the same set of modifications could serve multiple therapeutic targets, reducing the work needed to optimize each receptor individually. However, results from six receptors are still insufficient to establish broad applicability, and mispairing was reduced rather than eliminated. Whether the additional amino acid changes trigger immune responses in humans requires further assessment. Safety remains an essential hurdle in moving from more effective experimental cells to a usable treatment.

References

  1. The Jewish Chronicle
  2. Science Advances (via PubMed Central)
  3. Mirage News (Bar-Ilan University release)