CAR ICD Refactoring: AI-Guided Design of krgr–ζ1–40 for T-Cells

bioRxiv (Generate Biomedicines) USA

Overview

AI agents were used to refactor chimeric antigen receptor (CAR) intracellular domains (ICDs), resulting in the compact krgr–ζ1–40 design. This 45-amino-acid ICD increases surface expression and IFN-γ release compared to the clinical 4-1BB–CD3ζ (BBζ) ICD. The findings demonstrate that iterative agent-guided design can sustain CAR T-cell function and enhance B-cell depletion.

In Depth

Key Findings

The primary outcome of this research is the development of krgr–ζ1–40, a refactored intracellular domain (ICD) that significantly enhances CAR expression and function. Compared to the clinical 4-1BB–CD3ζ (BBζ) ICD, krgr–ζ1–40 increased CAR surface expression and antigen-driven IFN-γ release. Notably, this increase occurred without detectable antigen-independent IFN-γ release.

Technical Details

The design process utilized AI agents to propose and refine CAR ICDs over three rounds of iterative measurements. The resulting krgr–ζ1–40 is a 45-amino-acid ICD, making it less than a third of the length of the BBζ ICD. Technical performance improvements include:

• Extended CAR expression and target-cell killing following transient mRNA delivery.
• Support for the expression of four CARs from a single shared mRNA dose.
• Enhanced B-cell depletion when delivered via lipid-nanoparticle and lentiviral methods.

Background & Context

Chimeric antigen receptor (CAR) intracellular domains are critical components that couple the recognition of an antigen to the activation of T-cells. Traditional designs, such as the clinical BBζ, serve as the benchmark for these functions. This study sought to optimize these domains by leveraging AI-guided iterative design and experimental feedback to create more compact and efficient structures.

Significance & Outlook

The ability to produce compact ICDs like krgr–ζ1–40 suggests a pathway toward more efficient CAR T-cell therapies. By reducing the length of the ICD while simultaneously increasing surface expression and sustaining function, this approach demonstrates the utility of AI agents in protein engineering. The results indicate that iterative refinement based on experimental data can yield designs that improve both the delivery and the therapeutic efficacy of CAR T-cells.

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.

• Your email and selected fields are used only to deliver the newsletter.
• We never share your information with third parties.
• You can unsubscribe anytime via the link in each email.

See our Privacy Policy for details.

Agree & Continue ▶

Takes about a minute · Unsubscribe anytime