Computational and Molecular Engineering Platforms for Optimizing Cell Therapies

A computational method optimizes CAR T-cell design by engineering hinge domains for better tumor targeting.

Background

Cancer immunotherapy, particularly Chimeric Antigen Receptor (CAR) T-cell therapy, represents a significant advancement in treating various malignancies. This approach modifies a patient’s T-cells to express synthetic receptors that can recognize and attack cancer cells. While CAR T-cells have achieved remarkable success in certain blood cancers, there remains a critical need to enhance their effectiveness and expand their applicability to a broader range of tumors, especially those characterized by low target antigen expression or resistance to current treatments. Improving the precision and potency of CAR T-cells is paramount for better patient outcomes.

Current CAR T-cell design methodologies often rely on empirical testing and a limited understanding of the precise biophysical interactions crucial for optimal T-cell activation. A significant challenge lies in adequately controlling the spatial relationship between the CAR on the T-cell and its target antigen on the tumor cell surface. The “hinge domain” of the CAR, which dictates the extension and flexibility of the antigen-binding site, is often chosen without systematic optimization. This oversight can lead to suboptimal intermembrane distances, hindering effective immunological synapse formation, reducing T-cell sensitivity, and impairing anti-tumor responses, particularly against tumors with sparse antigen expression.

Technology

This computational workflow revolutionizes the design of Chimeric Antigen Receptor (CAR) T-cell therapies by precisely engineering the optimal spatial distance between CAR T-cells and their target cells. It employs a two-step computational process: first, AlphaFold2 models CAR hinge domains to predict their intrinsic flexibility and spatial reach. Second, these predicted distributions are mathematically combined with the known distributions of target receptors on tumor cells, enabling the identification of the ideal interaction geometry for effective CAR-target engagement and robust T-cell activation.

What differentiates this approach is its foundational reliance on the discovery that a specific, optimal intermembrane distance is crucial for successful CAR-target binding and subsequent T-cell activation. This shifts CAR-T design from empirical, trial-and-error methods to a predictive, structure-based engineering paradigm. By allowing for the in-silico selection of specific hinge domains, this technology provides precise spatial control, significantly maximizing CAR T-cell sensitivity, even against tumor cells expressing low levels of target antigens, a major challenge in current therapies.

Benefits

  • Enhances CAR T-cell therapy effectiveness by optimizing intermembrane distances for improved target engagement
  • Enables in-silico prediction and selection of hinge domains to maximize CAR T-cell sensitivity, even against low antigen expression
  • Improves tumor cytokine response and cell killing efficacy 
  • Uses advanced computational modeling (AlphaFold2 and molecular simulations) for unique structural and functional predictions
  • Facilitates rational CAR design, reducing reliance on empirical trial-and-error methods
  • Supports expansion to multiple target receptors and epitope locations for broader therapeutic applications
  • Demonstrates commercial readiness with validated experimental and in-vivo results

Applications

  • Personalized CAR-T therapy design
  • Preclinical CAR-T candidate screening
  • Optimizing CARs for low-antigen tumors
  • Automated CAR hinge domain selection

Additional info

Publication

Enhancing CAR T cell therapy through in silico modeling of hinge sequence length based on epitope location

Biophysical Journal Volume 123, Issue 3, Supplement 1552aFebruary 08, 2024Open Archive

Justin M. Mirazee ∙ Akashnathan Aranganathan ∙ Sooraj Achar ∙ Dongya Jia1∙ Xiang Chen ∙ Bodhi P. Vani ∙ Christopher D. Chien ∙ Marie Pouzolles ∙ Kniya DeDe ∙ Philippe Youkharibache ∙ Kylie Walters ∙ Pratyush Tiwary ∙ Gregoire Altan-Bonnet ∙ Naomi Taylor