A platform for engineering immune cells with combinatorial antigen recognition circuits for precise cancer targeting

Background

Adoptive cell therapies, particularly those involving engineered T cells such as chimeric antigen receptor (CAR) T cells, have emerged as a powerful approach in cancer treatment. These therapies harness the patient’s own immune system to selectively target and destroy cancer cells by recognizing specific antigens on the tumor cell surface. Despite their promise, the field faces a critical challenge: most tumor-associated antigens are not uniquely expressed on cancer cells but are also present, albeit at lower levels, on normal tissues. This overlap significantly increases the risk of off-tumor toxicity, where healthy cells are inadvertently attacked, leading to severe and sometimes fatal side effects. As a result, there is a pressing need for strategies that can enhance the specificity of immune cell targeting to maximize tumor eradication while minimizing collateral damage to normal tissues.

Current approaches in CAR T cell therapy predominantly rely on the recognition of a single antigen to trigger immune cell activation and tumor cell killing. However, this single-antigen targeting paradigm is inherently limited by the lack of truly tumor-specific markers, resulting in a narrow therapeutic window and frequent off-target effects. Attempts to improve specificity by identifying better single antigens have yielded only incremental gains, as most candidate antigens still exhibit some expression in normal tissues. Moreover, the complexity and heterogeneity of tumor antigen expression further complicate the identification of universally safe and effective targets. These limitations underscore the inadequacy of current single-antigen strategies and highlight the need for more sophisticated solutions that can discriminate between tumor and normal cells with greater precision.

Technology

This technology is a platform for developing engineered cancer immunoherapies by integrating combinatorial antigen recognition circuits. It utilizes a tailored computational pipeline, applied on a large collection of transcriptomics data, to prioritize optimal combinations of cell surface antigens that distinguish cancer cells from healthy cells while maximizing specificity and sensitivity. Ranked combinations are specified for engineered T cells approaches that utilize synthetic recognition circuits, such as AND, AND-NOT, and OR logic gates, which require the presence or absence of specific antigen combinations for activation. This technology details these circuits implemented using a variety of synthetic biology tools, including synNotch receptors, chimeric antigen receptors (CARs), inhibitory CARs, T cell receptors, and polyspecific-immune-inducing polypeptides (PIIPs) for diverse cancer types.

What differentiates this technology is its data-driven approach to antigen selection and its customizable modular engineering of immune cell recognition logic. Unlike traditional CAR T therapies that target a single antigen, this solution leverages Boolean logic to require 2-3 combinatoric antigenic signatures for immune cell activation, improving specificity and safety. The computational pipeline evaluates millions of antigen pairs and triplets, optimizing for tumor-versus-normal discrimination using robust statistical and machine learning methods. Experimental validation confirms that these combinatorial circuits outperform single-antigen approaches, and the platform’s compatibility with existing gene delivery and cell engineering techniques ensures broad applicability. This integration of computational design and synthetic biology enables the rapid development of highly specific, next-generation cellular immunotherapies for virtually any cancer type.

Benefits

• Improved specificity of cancer cell targeting by requiring recognition of multiple antigens, reducing off-target toxicity to healthy tissues.
• Expanded therapeutic window and enhanced safety profile compared to single-antigen CAR T therapies.
• Modular and programmable immune cell activation using Boolean logic gates (AND, AND-NOT, OR) for precise tumor discrimination.
• Computational pipeline enables identification of optimal antigen combinations tailored to diverse cancer types.
• Validated experimental proof-of-concept demonstrating selective killing of cancer cells while sparing normal cells.
• Compatibility with existing gene delivery and immune cell engineering technologies for clinical translation.
• Potential to address challenging cancers by using multi-antigen circuits that achieve near-ideal tumor discrimination.

Applications

• Personalized CAR T cell therapies
• Off-the-shelf engineered immune cells
• Targeted solid tumor immunotherapy
• Minimizing off-target cancer toxicity

Additional info

This system utilizes Boolean logic circuits to engineer immune cells for precise tumor targeting. By integrating synthetic receptors like synNotch and CARs, cells require specific multi-antigen combinations (AND, AND-NOT gates) for activation. A computational pipeline identifies optimal antigen pairs or triplets from transcriptomic datasets, maximizing tumor discrimination. This combinatorial approach minimizes off-target toxicity and expands the therapeutic window for cellular immunotherapies across various malignancies.