OC 69.2 - Costimulatory domains of second- and third-generation CAR-Tregs affect the type of FVIII-specific Treg responses and might be able to stabilize the Treg phenotype

Background: Development of inhibitors against FVIII is one of the major complications in Hemophilia A patients treated by protein replacement therapy. Treatments for tolerance induction towards FVIII besides ITI are still missing.

Aims: This project aims to study the potential of second- and third-generation FVIII-specific CAR-Tregs to reverse inhibitor-formation against FVIII.

Methods: A single-chain variable fragment (scFv JE10) was used for the generation of a second-generation FVIII-specific CAR (CAR-2). ScFv affinity for FVIII was measured by affinity ELISA. Addition of a second costimulatory domain (OX40, CTLA-4, 4-1BB and ICOS) resulted in third-generation CARs (CAR-3) (Fig.1). CAR construct-containing lentiviral vectors were used to transduce Tregs. CAR-Treg performance was investigated stimulating rested CAR-Tregs with FVIII. Tregs were analyzed by flow cytometry. Cytokine analysis (including IL-10, INFγ and TNFα) was assayed by bead array.

Results: Second- and third-generation CAR-Treg performance was studied following FVIII stimulation. Despite variable transduction rates and CAR surface expression levels, all CAR-Treg variants responded to FVIII. Treg-specific surface activation marker 4-1BB was upregulated following FVIII stimulation. Expression of important Treg homeostasis and stabilty markers (Foxp3 and Helios transcription factors) upon FVIII stimulation of CAR Tregs was mainly comparable to untransduced Tregs. Thus, analyzed CARs may not negatively impact Treg phenotype stability. In particular, CAR-3-JE10-ICOS-Tregs showed the highest Foxp3 expression in response to FVIII. Hence, this CAR variant may improve the Treg phenotype. Cytokine analysis revealed that CAR-2-JE10, CAR-3-JE10-OX40 and CAR-3-JE10-ICOS secreted the highest amount of the immunosuppressive cytokine IL-10.

Conclusion(s): Analyzed CAR-Treg variants showed potent cellular responses to FVIII. In particular, our data suggest that CAR-3-JE10-ICOS might be the most promising CAR for functional FVIII-specific CAR-Treg generation. Overall, variable responses among CAR variants in the context of Tregs might be influenced by CAR surface expression levels, and/or signaling pathways induced by different costimulatory domain combinations.