Unveiling hidden threats: clustering loops in CAR-T cells impact cancer therapy

By Vijay Kumar MalesuReviewed by Lily Ramsey, LLMAug 14 2023

In a recent study published in Nature Communications, a group of researchers determined how complementarity-determining region (CDR) loops in the single-chain variable fragment (scFv) influence chimeric antigen receptor (CAR)-T cell behavior, leading to antigen-independent activation and potential dysfunction.

Study: Complementarity-determining region clustering may cause CAR-T cell dysfunction. Image Credit: Gorodenkoff/Shutterstock.com

Background

CAR-T cell therapies have transformed blood cancer treatments. The CAR design, comprising components like the scFv for antigen recognition and other domains, is vital for CAR-T cells' effective function and sustainability in patients.

Some CARs can cause chronic activation without ligands or other elements linked to specific scFv features, which leads to CAR clustering and uncontrolled signaling.

Interestingly, the differentiation 19 (CD19) cluster-targeting scFvs, especially from the FMC63 clone used in approved CAR-T products, do not induce this signaling, possibly explaining their efficacy against B-cell malignancies.

Hence, designing more effective and safer CAR-T therapies that could benefit patients with blood cancers and possibly other malignancies is essential.

About the study

The present study explored the extracellular domain of human interleukin-13 receptor alpha 2 (IL13Rα2).

Using the multiBac expression system, a biotinylated receptor was created by merging the domain with BirA biotin ligase. After obtaining genes from GeneArt Thermo Fisher Scientific, they were structured, sequenced, and expressed in Sf9 cells. The resultant protein was purified and verified using various techniques.

Using the SciLifeLab synthetic library, the researchers targeted the biotinylated receptor, identifying eleven key scFv candidates. Binding kinetics and other properties were thoroughly examined. The research also detailed cell culture conditions, cell engineering methods, and applications on human T cells.

The researchers further transduced and cultured T cells, later examining CAR's distribution on these cells using confocal microscopy. To analyze CAR-T cell behavior, the assessed CAR-T cell binding to recombinant IL13Rα2, stained T cell surface markers, and investigated CAR expression both at the surface and intracellularly.

Additionally, they explored CAR-T cell proliferation and cytokine secretion and analyzed gene expression in different CAR-T constructs, highlighting their differences.

In vivo experiments involved orthotopic implantation of tumor cells in mice, followed by CAR-T treatment and tumor growth monitoring via in vivo imaging system (IVIS). Statistical analysis was performed using GraphPad Prism with standard significance markers.

Study results

In the present study, the researchers selected five scFv molecules, labeled from [A] to [E], using phage display panning, specifically binding to human IL13Rα2 but not the abundantly found IL13Rα1.

These molecules showed comparable thermal stability, with scFv[B] being a slight outlier. Testing results revealed that scFv[A], [B], and [C] are attached within IL13Rα2's ligand-binding domain. However, scFv[D] and [E] did not.

These scFvs were used to develop five specific CAR-T cell constructs. After expression, all CARs showed the binding capacity to IL13Rα2. In vitro evaluations found that all five CAR-Ts effectively killed the U-87MG human glioblastoma cell line known for high IL13Rα2 levels.

CAR[D]-T and CAR[E]-T, with CAR[C]-T to some extent, also displayed strong cytotoxicity against other glioblastoma cell lines with lower IL13Rα2 levels, while CAR[A]-T and CAR[B]-T did not.

Most notably, the CAR-Ts did not display cytotoxicity against IL13Rα2-negative Mel526 melanoma cells. Additionally, CAR[C]-T, CAR[D]-T, and CAR[E]-T proliferated significantly upon exposure to U-87MG tumor cells. In contrast, CAR[A]-T and CAR[B]-T showed diminished proliferative abilities. Overall, CAR[C]-T, CAR[D]-T, and CAR[E]-T showcased superior effectiveness compared to CAR[A]-T and CAR[B]-T.

Upon examining the different functionalities of CAR-Ts, it was noticed that CAR[A]-T and CAR[B]-T showed reduced CAR surface expression compared to CAR[E]-T, hinting at the instability in these CAR molecules. Further, these two CARs tended to cluster on the T-cell surface, which led to a heightened activated state.

This activation, despite the absence of specific stimuli, indicated antigen-independent tonic signaling in CAR[A]-T and CAR[B]-T. Such activation led to faster exhaustion of these CAR-T cells, while CAR[E]-T responded more effectively to antigen-specific stimuli.

Gene expression studies compared the functional CAR[E]-T with the less effective CAR[B]-T. Principal component analysis depicted that CAR[E]-T gene expression closely resembled Mock-T, while CAR[B]-T had a unique gene expression profile.

A total of 136 genes were found to be expressed differently between CAR[B]-T and both CAR[E]-T and Mock-T. CAR[B]-T displayed upregulated genes related to T cell activation, inhibition, exhaustion, and apoptosis, confirming its pre-activated and exhausted state, even without antigen-specific stimuli.

The scFvs utilized to create different CARs shared identical frameworks but varied in a few amino acids present in the CDRs.

Alanine substitution experiments on these CDRs revealed that specific amino acids in CDR loops could significantly influence CAR molecule stability. This stability further impacted CAR clustering and subsequent antigen-independent activation and exhaustion.

Conclusions

In the present research, the intricacies of a CAR construct's components, notably its extracellular scFv, linker, and additional domains, were evaluated for their impact on CAR-T cell performance.

It was determined that the CDR loops within the scFv are responsible for antigen recognition. Surprisingly, certain frameworks in the scFv could unintentionally activate CAR signals, potentially impairing T cell function.

The study results also uncovered that interactions within the CDR loops could further impact CAR stability. Despite all derived CAR-Ts proving effective against high IL13Rα2-expressing glioblastoma cells, their efficiency waned on cells with lesser expression.

Notably, CAR[A]-T and CAR[B]-T faced challenges due to CAR aggregation. The researchers further emphasized that minor variances in the CDR loops can profoundly alter CAR-T cell effectiveness and proposed a new rapid screening approach for CAR behaviors.