Fig 2: Phenotypically selected clones show specific and differential activation of NFAT signalling in the presence of hMSLN+ cells. CAR activation profiling using phenotypically selected scFv clones reformatted as CAR-mGFPs in Jurkat NFAT-Lucia luciferase reporter cells. (a) Stimulation in response to MSLN+ (H-226, AsPC-1) and MSLN- (HEK293-6E, Raji) target cells. (b) Stimulation in response to AsPC-1 wt and a polyclonal AsPC-1 KO cell pool enriched for MSLN CRISPR disruption events. Assay controls: FMC63 (CD19+, Raji); P4 (MSLN+ lines). RLU, relative light units; data are represented as mean ± SD from triplicate discrete assay points.

Fig 3: Phenotypically enriched anti-hMSLN CAR-mGFP clones trigger effector and killing functionality in transduced primary T cells. (a, b) Expression of inflammatory (IFN?) and a surface activation marker (CD25) in the presence of H-226 tumor cells. (c, d) Killing of H-226 Antares transduced target cells as determined by gated cell survival (c) and liberation of the Nanoluc luciferase reporter by cell lysis (d). Mock, out-of-frame CAR-mGFP ORF vector; HEK293-6E and Raji Antares, MSLN- control lines. All data are represented as mean ± SD from triplicate discrete assay points; RLU, relative light units; LB, lysis buffer (100% cell death control). Live Cell Counts in (c) indicate remaining DAPI- Antares+ target cells.

Fig 4: Enrichment of an anti-hMSLN CAR-active clone population by phenotypic sorting. (a) Low-frequency transduction of Jurkat NFAT-mCherry cells with a CAR-mGFP library incorporating ~ 105 scFvs pre-selected on MSLN (see text). (b) Phenotypic sorting and enrichment workflow for anti-MSLN CAR-mGFP. NS, negative sorting of the transduced library following co-incubation with MSLN- HEK293 cells to subtract for non-MSLN reactive clones (GFP+ mCherry- gate); CS01/CS02, positive cell sorting steps to enrich for the Jurkat CAR-mGFP clone population responding to MSLN+ target cells (GFP+ mCherry+ gate). (c) Representative iQue flow cytometer plate screening data for MSLN-dependent target cell binding by discrete scFvs recovered from the phenotypic library sorting and expressed to bacterial media. Left panel, Mean Fluorescence Intensities (MFI) for scFvs binding cognate MSLN+ H-226 cells with subtraction of the corresponding binding to HEK293 MSLN- control cells; Right panel, Mean Fluorescence Intensities (MFI) for scFvs binding cognate MSLN+ AsPC-1 cells with subtraction of the corresponding signals obtained against an AsPC-1 polyclonal line enriched for MSLN KO events by CRISPR. Cell binding of scFvs was detected using anti-Histag-Alexa647. Empty bars, anti-MSLN positive controls.

Fig 5: Binding affinities of MSLN CAR variants.(A) Schematic representation of the lentiviral vectors containing MSLN CAR variants. (B) Relative binding of MSLN CAR variants was estimated by staining CAR-expressing Jurkat T cells with 25 nM of Alexa Fluor 647–conjugated monomeric human and mouse MSLN, using NT cells as control. (C) Determination of the equilibrium binding affinities of CAR variants using a saturation binding assay with Alexa Fluor 647–conjugated human and mouse MSLN (n = 1).