Development of Novel Non-Immunoglobulin Centyrin™-Based CARs Burton E. Barnett, Jenessa B. Smith, Xinxin Wang, David Hermanson, Yening Tan, Eric M. Ostertag, Devon J. Shedlock ABSTRACT INTRODUCTION CONCLUSIONS Chimeric-antigen receptor (CAR)-T cell immunotherapy is a promising type of cancer therapy and substantial progress has been made in developing adoptive T cell approaches for B cell malignancies. B cell maturation antigen (BCMA) is an attractive target for patients with multiple myeloma (MM) due to its high level of expression on tumor cells and restricted expression on normal tissues. In addition, BCMA is essential for cell growth and survival, and is therefore likely not susceptible to antigen escape. Traditionally, the antigen-binding domain of a CAR is a single chain variable fragment (scFv) comprised of heavy chain (HC) and light chain (LC) variable fragments joined by a flexible linker that has been derived from a non- human monoclonal Ab (mAb). However, there are a number of disadvantages to scFv-based CARs including the limited availability of scFv, their potential to elicit antibody responses, and their association with tonic signaling due, in part, to inherent instability and flexibility of the structure and the potential for both HC/LC domain swapping and multimer formation through framework region interactions. Thus, replacement with alternative binding technologies may improve CAR-T efficacy in the clinic. Centyrins™ are alternative scaffold molecules that bind protein targets with high affinity and specificity, similar to scFv molecules. However, unlike scFv, Centyrins™ are smaller, derived from human consensus tenascin FN3 domains and are predicted to have decreased immunogenicity. Additionally, a monomeric Centyrin™ in CAR format (“CARTyrin” molecule) is less likely to engage in domain swapping or interact with other Centyrins™ at the cell surface, thereby limiting the potential for the tonic signaling that drives the functional exhaustion of CAR T cells. Finally, as Centyrins™ are 1/3 the size of scFv molecules, they provide flexibility for the creation of multi-specificity CARs. Centyrins™ can be isolated against virtually any antigen through ex vivo panning of an extensive Centyrin™ library, yielding many distinct binders with a range of affinities and target epitopes. We tested a panel of 12 anti-BCMA (11 monomeric and one multimeric “P9”) Centyrins™ by linking them to a standard second-generation CAR scaffold. High quality mRNA of each CARTyrin construct was produced in order to rapidly screen CARTyrin cell surface expression and functionality in human pan T cells against BCMA + targets. We also constructed scFv-based CARs against CD19 and BCMA for comparison. CD3/CD28-stimulated T cells were electroporated (EP) with mRNA encoding each of the 12 anti-BCMA CARTyrins and, the following day, analyzed for surface expression of CARTyrin and their ability to degranulate against BCMA + tumor cells. All 12 CARTyrins were detected on the cell surface and the 11 monomeric CARTyrins imparted BCMA-specific killing capacity to T cells. Notably, in these assays, CARTyrins were functionally comparable to scFv-based CARs against BCMA or to CD19-specific scFv-based CARs in a parallel assay with CD19 + tumor cells. The 11 functional anti-BCMA CARTyrins were further characterized for functional avidity by determining their activity against a panel of target cells with titrated levels of surface BCMA expression. To create this panel, various amounts of high quality BCMA mRNA were electroporated into BCMA - K562 tumor cells. After 4 hours of co- culture with the panel of BCMA expressing cells, CARTyrin + T cell activity was measured as a function of CD107a expression. We observed a range of activities by each CARTyrin and show that this assay can be utilized to determine the minimal effective dose of BCMA needed to induce killing by CARTyrin + cells. Furthermore, we establish that certain BCMA-specific CARTyrins are responsive to target cells with extremely low levels of surface BCMA expression. The data shown in this poster are complemented by functional studies in vitro and in vivo that may be found in the posters listed below. These results confirm that Centyrins™ are viable replacements for scFv in the construction of functional CARs and establish their potential utility in generating novel BCMA-specific CAR molecules, as well as other novel targetable tumor antigens. Poseida Therapeutics, Inc. 4242 Campus Point Court, Suite 700 San Diego, CA, 92121 (CARTyrins) Targeting Human BCMA Signal Peptide Transmembrane Hinge Costimulatory Signaling Domain Centyrin™ Figure 1: Surface CARTyrin Expression in RNA-electroporated human T cells Surface expression of CARTyrin using a biotinylated BCMA-Fc protein and fluorescent streptavidin 24 hours post electroporation in primary human T cells. T cells were previously stimulated with anti-CD3/CD28 mAb-coated beads, allowed to rest, and thawed O/N prior to electroporation with 10ug CARTyrin mRNA. a) CARTyrin expression (y-axis). CARTyrin+ frequency (b) and mean fluorescence intensity (c). BCMA-specific CAR molecules (“scFv 1” and “scFv 2”) were included as positive controls. • All CARTyrins tested were expressed on the surface of primary human T cells • All BCMA-specific CARTyrins (except F9 comprised of multimeric Centyrin™) exhibited specific effector function against BCMA+ tumor cells, demonstrating comparable functionality to BCMA scFv-based CARs • BCMA-specific CARTyrins exhibit high functional avidity and a titratable dose response against BCMA antigen, demonstrating a range of potency down to picogram amounts of BCMA • Testing RNA CARs is an effective screening platform for identification of lead CAR candidates • The potency exhibited by Centyrin™-based CARTyrins combined with their advantages in size, immunogenicity, flexibility, and lack of tonic signaling suggest they may provide an excellent alternative to scFv based CARs Figure 3: Electroporation with lower levels of CARTyrin mRNA permit greater functional distinction between CARTyrin candidates (a) Surface expression of in T cells 24 hours post electroporation with either 10 ug or 5ug of CARTyrin mRNA. (b) Expression of CD107a (y-axis) and CD8 (x-axis) in live (7AAD-), CD3+ T cells 4 hours post co-culture with tumor cells at an effector to target ratio of 1:4. Each bar represents one CARTyrin molecules, in order of P5, P12, P8, P2, P11. Assays were performed 24hrs post electroporation. Figure 4: Comparative analyses of BCMA-specific CARTyrins against BCMA+ tumor cells (a) Surface expression of CARTyrin in T cells electroporated with 5ug of CARTyrin mRNA. (b) Frequency of CD107a+ in live, CD3+, CD8+ T cells 4 hours post co-culture with tumor cells at an effector to target ratio of 1:4. Assays were performed 24hrs post electroporation. Error bars represent 3 independent experiments. Centyrin™ is a registered trademark of Janssen Pharmaceuticals, Inc. Poseida has licensed certain rights to the Centyrin™ technology platform from Janssen Pharmaceuticals, Inc. for use in autologous T cell therapeutics (a) (b) K5 62 H92 9 U266 %CD107a+ K562 H929 U266 %CD107a+ (a) (b) Figure 2: T cells electroporated with BCMA CARTyrins demonstrate specific activity against BCMA+ tumor cells (a) Expression of CD107a (x-axis) in live (7AAD-), CD3+ T cells 4 hours post co-culture with tumor cells at an effector to target ratio of 1:4. Frequencies of CD107a+ cells is shown on the plots. (b) Bar graphs represent CD107a+ frequency of CD8+ T cells. Assays were performed 24 hours post electroporation with BCMA-specific CARTyrins. (c) BCMA expression via flow cytometry in the indicated tumor cells. A CD19-specific scFv CAR (“CD19”) was included as a positive control for its activity against Nalm6 cells. P7 P1 2 P9 P11 P1 0 P2 P 4 P6 P5 P3 P1 P8 BC M A sc Fv 1 BCMA scFv 2 CD19 s cFv P7 P12 P9 P 1 1 P10 P2 P4 P6 P5 P3 P1 P8 BCMA s c Fv 1 BCMA scFv 2 CD 1 9 s c Fv P7 P1 2 P9 P11 P1 0 P2 P4 P6 P5 P3 P1 P8 BC M A sc Fv 1 BCMA scFv 2 C D19 s cFv (a) (b) (c) P7 P12 P9 P1 1 P 10 P 2 P4 P6 P5 P 3 P 1 P8 0 10 20 30 40 50 60 70 80 90 100 P7 P 12 P 9 P11 P10 P2 P4 P6 P 5 P3 P1 P8 0 10 20 30 40 50 60 70 80 90 100 P 7 P 12 P 9 P11 P10 P 2 P 4 P 6 P 5 P3 P 1 P 8 0 10 20 30 40 50 60 70 80 90 100 (a) (b) Figure 5: BCMA-specific CARTyrins function against picogram amounts of BCMA (a) Expression of CD107a (y-axis) in CD8+ T cells (pre-gated on live (7AAD-), CD3+) 4 hours post co-culture with K562 tumor cells electroporated with the indicated amount of BCMA mRNA (x-axis). Effector to target ratio of 1:4. (b) Expression of BCMA on BCMA-electroporated K562 cells. Mean fluorescence intensity (MFI) of BCMA expression is quantified on the y-axis. Assays were performed 24 hours post electroporation with BCMA-specific CARTyrins. L.O.D., limit of detection. 0.001 0.01 0.1 1 10 100 0 20 40 60 80 BCMA mRNA (ug) P7 P12 P9 P11 P10 P2 P4 P6 P5 P3 P1 P8 scFv1 scFv2 Mock (a) (b) I. PiggyBac™-Produced CAR-T Cells Exhibit Stem-Cell Memory Phenotype Session 703 Poster I II. A Novel Bcma-Specific, Centyrin™-Based CAR-T Product for the Treatment of Multiple Myeloma Session 653 Poster I CD107a P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 scFv 1 scFv 2 CD19 Mock CD107a P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 scFv 1 scFv 2 CD19 Mock CD107a P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 scFv 1 scFv 2 CD19 Mock CARTyrin P5 P4 P3 P6 P12 P9 P1 P11 P8 P10 P7 P2 scFv 1 scFv 2 Mock P 1 2 P 11 P7 P2 P4 P8 P3 P6 P10 P 1 P5 P9 s c F v1 s c Fv 2 no mRNA P1 2 P11 P7 P 2 P4 P8 P 3 P6 P 10 P 1 P5 P9 s c F v 1 s c Fv2 n o m RNA %CAR+ METHODS & RESULTS scFv Centyrin™ Size: ~250 aa ~90 aa Components: Heavy & Light Chain Variable Regions, Flexible Linker Fibronectin type III domain Derived: Mouse/RodentAbs Human Stability: Not as stable as original IgG format Stable