Session V — Lifetime Achievement: Michel Sadelain

From a CAR Molecule to an Evolving Therapeutic Platform

Michel Sadelain traces CAR-T evolution from the 1991 zeta-chain fusions through CD19 to three new designs — 1XX, HIT, and pre-TCR — that address calibration, sensitivity, and expansion.

---

Origins

• 1991: three groups cloned CD3-zeta and showed it could trigger T-cell activation — but in leukemia lines
• 1993: Zelig Eshhar replaced extracellular stub with scFv — antibody specificity, T-cell trigger
• Sadelain’s key contribution: retroviral methods for primary T cells (not leukemia lines), revealing that zeta alone → activation-induced cell death or anergy
• Solution: second-generation CARs fusing zeta with CD28 (or 4-1BB) for co-stimulation
• 2003: coined the term “CAR” and “living drugs”

Choosing CD19

• CD20 was the famous target (rituximab already a drug); Sadelain chose CD19 instead
• CD19: earlier in B-cell ontogeny, higher and more homogeneous surface density, broader malignancy coverage
• Bonus thesis: depleting B cells would dodge anti-CAR immunity

Three New CAR Designs

1XX — Calibrated CD3-zeta signaling

• Inactivated two of three ITAMs → single live ITAM outperformed wild-type on durability without sacrificing potency
• With Takeda: 28 patients, durable responses at doses as low as 25 million cells (vs typical 100-300M)
• Knocked into TRAC locus by CRISPR: responses at 10 million cells

HIT — HLA-Independent T-cell receptors

• Swaps antibody V regions into TCR alpha/beta chains, knocked into TRAC locus
• 10–50× lower antigen-density threshold than conventional CARs
• Published in Science, ~early 2026

Pre-TCR Signaling

• Borrows thymocyte expansion logic: 100–200-fold proliferation without exhaustion
• Extreme potency in glioblastoma at very low cell doses
• Published in Cell, ~early 2026

Design Principles

1. Use real T cells, not leukemia lines
2. Build co-stimulation into the receptor
3. Calibrate, don’t maximize — the 1XX lesson
4. Sensitivity is a design parameter
5. Pick targets on lineage biology, not existing fame
6. Steal from development (pre-TCR, thymocyte biology)