New Protein Design Methods Boost T Cells’ Fight Against Cancer and Viruses

A joint team from Bar-Ilan University and the Weizmann Institute of Science has used advanced computational protein design to strengthen a core piece of the machinery T cells rely on to detect and destroy diseased cells. The work, published in Science Advances, focused on the T-cell receptor, or TCR—the natural sensor on T cells that recognizes foreign targets and triggers an immune attack.

Using protein design calculations, the researchers built a modified TCR scaffold they call SET (Structurally Enhanced TCR). Engineered T cells carrying this redesigned receptor produced stronger immune responses and killed cancer cells more effectively in laboratory experiments than T cells built with the original receptor. “Computational protein design is giving us a new and much more effective way to engineer the immune system,” says Prof. Cyrille J. Cohen, who led the research. “In this study, we used computational design to make T cells substantially more powerful at attacking their targets, and we saw this effect across both cancer and viral targets.” 

The effect held up in mice carrying human tumors. By day 83, tumors treated with SET-containing T cells were roughly 35% smaller than those in untreated controls, and by day 127, every mouse that received the enhanced cells was still alive, compared with fewer than half of the untreated mice.

Rather than working for just one target, the improvement showed up across a range of them. The team applied the same design approach to T cells built to recognize several cancer-associated targets, including ones tied to melanoma and other hard-to-treat cancers, as well as targets from Epstein-Barr virus and SARS-CoV-2, and the enhanced receptor outperformed the original version in every case. “One of the most exciting findings was that the same strategy could improve T-cell function across very different targets,” says Maria Radman, the study’s first author. “We saw enhanced activity not only against cancer-associated targets, but also against viral targets. This suggests that the approach could be broadly useful for engineering T cells with stronger and more effective immune responses.”

The computational design work was led by Prof. Sarel Fleishman who noted that the same mutations worked across different receptor types: “What’s particularly striking here is that a single set of mutations, strategically designed in conserved regions of the TCR, favorably impacts very different TCR types. Instead of spending years identifying enhancing mutations for every single therapeutic candidate, we expect the designed mutations to improve any TCR, dramatically accelerating the process of translating promising leads into therapeutics.”