6.1 CAR Structural Architecture, Generations & Costimulatory Domain Biology

CAR T-Cell Biology, Tumor Targets & In Vivo Cellular Kinetics

Quick Clinical Summary: Chimeric Antigen Receptor (CAR) T-cell therapy represents a revolutionary class of adoptive cellular immunotherapies wherein autologous (or allogeneic) T lymphocytes are genetically re-engineered to express synthetic receptors. Unlike physiologic T-cell receptors (TCRs) that require antigen processing and presentation on Major Histocompatibility Complex (MHC/HLA) molecules, CARs bind surface antigens directly in an HLA-independent manner. Understanding the structural domains of the CAR construct—specifically the metabolic and kinetic distinctions between CD28 and 4-1BB (CD137) costimulatory endodomains—is essential for anticipating clinical toxicity onset, cellular expansion peaks, and long-term immunologic persistence.

---

1. CAR Structural Architecture & Molecular Engineering

A Chimeric Antigen Receptor is a modular, recombinant fusion protein engineered through four distinct structural domains spanning the extracellular, transmembrane, and intracellular compartments of the T lymphocyte:

                             CAR MOLECULAR ARCHITECTURE
  [ Extracellular ]  ├── Single-Chain Variable Fragment (scFv)  <-- Antigen Binding (HLA-independent)
                     └── Hinge / Spacer Region (CD8a / IgG)     <-- Flexibility & Reach
  [ Transmembrane ]  ─── Transmembrane Domain (CD8a / CD28)     <-- Membrane Anchoring & Stability
  [ Intracellular ]  ├── Costimulatory Domain (4-1BB or CD28)   <-- Proliferation, Survival & Metabolism
                     └── CD3-zeta Activation Domain             <-- Primary ITAM Signaling Cascade

1. Extracellular Antigen-Recognition Domain (scFv)

• Composition: Derived from the variable light (VL) and variable heavy (VH) chains of a monoclonal antibody, connected via a flexible hydrophilic peptide linker (such as (Gly4Ser)3).
• Mechanism: Directly binds native, intact surface proteins (such as CD19 or BCMA) on tumor cells with high affinity without requiring antigen cleavage or MHC presentation. This bypasses major tumor immune escape mechanisms such as MHC class I downregulation or HLA loss.
• Epitope Accessibility: The affinity and length of the linker determine the precise spatial alignment between the CAR T-cell and the target tumor cell membrane.

2. Extracellular Hinge / Spacer Region

• Composition: Typically derived from immunoglobulin constant regions (IgG1, IgG4) or surface molecules such as CD8-alpha or CD28.
• Function: Extends the scFv away from the T-cell membrane, providing steric flexibility and optimal rotational reach to engage membrane-proximal or membrane-distal tumor epitopes.

3. Transmembrane Domain

• Composition: Hydrophobic alpha-helical segment derived from CD28, CD8-alpha, or CD3-zeta.
• Function: Anchors the synthetic receptor firmly within the T-cell lipid bilayer and facilitates the transmission of conformational binding signals from the extracellular scFv to the intracellular signaling machinery.

4. Intracellular Signaling Domains (Signal 1 + Signal 2)

• Primary Activation Domain (Signal 1): The cytoplasmic end of the CD3-zeta chain contains three Immunoreceptor Tyrosine-based Activation Motifs (ITAMs). Upon antigen binding, these ITAMs undergo phosphorylation by Src-family kinases (Lck), recruiting ZAP-70 to initiate downstream cytotoxic signaling, calcium mobilization, and transcriptional activation.
• Costimulatory Domain (Signal 2): Critical endodomain that provides survival and proliferative signals, preventing anergy and rapid activation-induced cell death.

Evolution of CAR Generations

• First-Generation CARs: Contained only the CD3-zeta intracellular domain. In clinical trials, these demonstrated poor in vivo proliferation, minimal cytokine secretion, and rapid T-cell anergy.
• Second-Generation CARs (Current Commercial Standard): Incorporate one costimulatory domain (either CD28 or 4-1BB / CD137) linked in tandem with CD3-zeta, providing robust in vivo expansion, durable cytotoxic killing, and high clinical efficacy.
• Third-Generation CARs: Incorporate two costimulatory domains in series (such as CD28 + 4-1BB + CD3-zeta) to combine rapid expansion with prolonged persistence.
• Fourth-Generation CARs (TRUCKs / Armored CARs): Engineered to constitutively or inducibly secrete immune-modulating cytokines (such as IL-12, IL-15, IL-18) to remodel the immunosuppressive tumor microenvironment.

---

2. Costimulatory Domain Biology: CD28 vs. 4-1BB

Second-generation products pair CD3-zeta with one costimulatory domain. CD28 signaling commonly favors brisk glycolytic activation and earlier expansion, while 4-1BB signaling commonly supports mitochondrial fitness and longer persistence. These are useful tendencies, not a bedside clock: construct details, manufacturing, disease, tumor burden, lymphodepletion, cell dose and host biology influence CRS/ICANS timing and persistence. Use the actual product label and baseline risk rather than assuming all CD28 or all 4-1BB products behave identically.

3. FDA-Licensed CD19 and BCMA CAR T Products

As of the source review, U.S.-licensed autologous CAR T products include:

• CD19: tisagenlecleucel (Kymriah; 4-1BB), axicabtagene ciloleucel (Yescarta; CD28), brexucabtagene autoleucel (Tecartus; CD28), lisocabtagene maraleucel (Breyanzi; 4-1BB), and obecabtagene autoleucel (Aucatzyl; 4-1BB), across label-defined B-cell leukemias and lymphomas.
• BCMA: idecabtagene vicleucel (Abecma; 4-1BB) and ciltacabtagene autoleucel (Carvykti; 4-1BB) for label-defined multiple-myeloma settings.

This is a current-awareness table, not an indication memorization substitute. FDA adds indications and revises dose, monitoring, warning, proximity and driving language. Confirm the current prescribing information, especially age, disease subtype, prior therapy, cell dose, lymphodepletion, release, administration and toxicity instructions.

4. Antigen Escape and Why One Marker Cannot Define Relapse

The same target specificity that makes a CAR potent creates its most important failure mode. Relapse after CD19- or BCMA-directed therapy divides into two biologically distinct categories, and they are managed differently.

Antigen-positive relapse occurs when the target is still expressed but the CAR T cells have contracted, lost function, or were suppressed. Re-treatment with a CAR-directed strategy may remain reasonable.

Antigen-negative relapse occurs when the malignant clone survives by no longer presenting the target. Recognized mechanisms include selection of pre-existing antigen-negative subclones, alternative splicing or mutation that removes the epitope the scFv binds while the protein is still nominally present, and lineage switch, in which a B-lineage leukemia relapses as myeloid disease and is therefore invisible to a CD19-directed product. Trogocytosis, the physical transfer of target antigen from the tumor cell onto the T cell, can also lower surface density below the threshold required for killing.

The clinical consequence for surveillance is direct: relapse cannot be excluded by a single antigen measurement. A flow panel restricted to CD19 will miss an antigen-negative or lineage-switched relapse entirely, which is why disease assessment uses marrow morphology, a broad immunophenotypic panel, cytogenetics, and measurable residual disease testing rather than the target antigen alone.