Abstract 3713: Immunoengineering by design: generative AI and high throughput cell screening to develop therapeutic TCR-guided bispecifics against multiple targets

Abstract T cell receptors (TCRs) are a unique class of targeting molecule. Through highly specific interactions with peptide:HLA complexes present on the cell-surface, they offer a means of therapeutically targeting any protein, with this specificity highlighted by their ability to selectively recognize cancer-associated single amino acid mutations. However, identifying TCRs that have the desired specificity toward a target peptide has traditionally been challenging; often screening large numbers of peripheral T cells to find only a handful of potential molecules. Even when a target peptide-specific TCR has been identified, its selectivity becomes crucial in developing a successful therapeutic, with reactivity toward off-target peptides having potentially serious safety implications. To increase the speed and success of TCR discovery, we have developed a in silico TCR design process leveraging recent advances in machine learning for protein sequences combined with high-throughput, rapid cell-based screening, where both specificity and selectivity are considered at the point of design. Targets for TCR generation are assessed for potential off-target reactivity through a combination of immunopeptidomic data, sequence similarity, and machine learning predicted binding similarity. Large protein language models are fine-tuned on experimental data which captures binding specificities between known TCRs and their peptide ligands, learning meaningful latent representations of both sequences. These trained models are then used to generate TCRs in silico that are optimized for selective target binding, leveraging an energy-based diffusion process that has been specially adapted to handle multiple conditioning. In this way, we can guide the model towards TCR sequences that bind to multiple target peptide-HLA sequence(s), enabling reactivity to heterogeneous neoantigens, cross-HLA recognition and off-target (self-peptide) avoidance. Designed sequences are then synthesized and expressed in a cell-based high throughput screen, providing binding data on up to a million TCR sequences per cycle, enabling further model refinement based on desired binding properties. Our pipeline has been applied to design novel TCRs against public neoantigens derived from the common KRAS mutations at the position 12. We show how a generative model produces TCRs optimized for multiple-antigen specificity, binding to one or more of the target peptide(s) but not self-peptides including the unmutated wild type KRAS protein. The TCRs are optimized for high-affinity interactions to drive development of TCR-guided bispecific therapies. Combining state of the art generative AI with high throughput screening empowers a TCR design process that considers both selectivity and specificity, opening up the next generation of rapid development of TCR-based therapeutics. Citation Format: Alex S. Powlesland, Aaron Sim, Dubravka Pezic, Valerie Coppard, Victoria Koullourou, Joshua Meyers, Lewis Cornwall, James Day, Neil Dalchau, Lilly Wollman. Immunoengineering by design: generative AI and high throughput cell screening to develop therapeutic TCR-guided bispecifics against multiple targets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3713.