A computationally designed chimeric antigen receptor provides a small-molecule safety switch for T-cell therapy

Abstract

Approaches to increase the activity of chimeric antigen receptor (CAR)-T cells against solid tumors may also increase the risk of toxicity and other side effects. To improve the safety of CAR-T-cell therapy, we computationally designed a chemically disruptable heterodimer (CDH) based on the binding of two human proteins. The CDH self-assembles, can be disrupted by a small-molecule drug and has a high-affinity protein interface with minimal amino acid deviation from wild-type human proteins. We incorporated the CDH into a synthetic heterodimeric CAR, called STOP-CAR, that has an antigen-recognition chain and a CD3ζ- and CD28-containing endodomain signaling chain. We tested STOP-CAR-T cells specific for two antigens in vitro and in vivo and found similar antitumor activity compared to second-generation (2G) CAR-T cells. Timed administration of the small-molecule drug dynamically inactivated the activity of STOP-CAR-T cells. Our work highlights the potential for structure-based design to add controllable elements to synthetic cellular therapies.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Structure-based computational design of a high-affinity CDH to control CAR-T-cell activity.
Fig. 2: Computationally designed heterodimeric STOP-CARs are stably expressed on the surface of Jurkat and primary human T cells.
Fig. 3: STOP-CARs are functional in primary human T cells, both in vitro and in vivo and activity can be tuned in a dynamic drug-dependent manner.

Data availability

The data supporting the findings of this study are available within the article and its Supplementary Information. Coordinates of the determined structure have been deposited in the PDB with accession code 6IWB. Other data are available from the corresponding authors upon reasonable request.

Code availability

All code used to perform the design simulation can be found at https://github.com/LPDI-EPFL/STOP-CAR.

References

  1. 1.

    Kalos, M. et al. T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci. Transl. Med. 3, 95ra73 (2011).

  2. 2.

    Brentjens, R. J. et al. CD19-targeted T cells rapidly induce molecular remissions in adults with chemotherapy-refractory acute lymphoblastic leukemia. Sci. Transl. Med. 5, 177ra138 (2013).

  3. 3.

    Maude, S. L. et al. Chimeric antigen receptor T cells for sustained remissions in leukemia. N. Engl. J. Med. 371, 1507–1517 (2014).

  4. 4.

    Turtle, C. J. et al. CD19 CAR-T cells of defined CD4+:CD8+ composition in adult B cell ALL patients. J. Clin. Invest. 126, 2123–2138 (2016).

  5. 5.

    Lim, W. A. & June, C. H. The principles of engineering immune cells to treat cancer. Cell 168, 724–740 (2017).

  6. 6.

    Lanitis, E., Dangaj, D., Irving, M. & Coukos, G. Mechanisms regulating T cell infiltration and activity in solid tumors. Ann. Oncol. 28, xii18–xii32 (2017).

  7. 7.

    Adachi, K. et al. IL-7 and CCL19 expression in CAR-T cells improves immune cell infiltration and CAR-T cell survival in the tumor. Nat. Biotechnol. 36, 346–351 (2018).

  8. 8.

    Tang, L. et al. Enhancing T cell therapy through TCR-signaling-responsive nanoparticle drug delivery. Nat. Biotechnol. 36, 707–716 (2018).

  9. 9.

    Morgan, R. A. et al. Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Mol. Ther. 18, 843–851 (2010).

  10. 10.

    Gargett, T. & Brown, M. P. The inducible caspase-9 suicide gene system as a “safety switch” to limit on-target, off-tumor toxicities of chimeric antigen receptor T cells. Front. Pharmacol. 5, 235 (2014).

  11. 11.

    Jones, B. S., Lamb, L. S., Goldman, F. & Di Stasi, A. Improving the safety of cell therapy products by suicide gene transfer. Front. Pharmacol. 5, 254 (2014).

  12. 12.

    Fedorov, V. D., Themeli, M. & Sadelain, M. PD-1- and CTLA-4-based inhibitory chimeric antigen receptors (iCARs) divert off-target immunotherapy responses. Sci. Transl. Med. 5, 215ra172 (2013).

  13. 13.

    Lanitis, E. et al. Chimeric antigen receptor T Cells with dissociated signaling domains exhibit focused antitumor activity with reduced potential for toxicity in vivo. Cancer Immunol. Res. 1, 43–53 (2013).

  14. 14.

    Cho, J. H., Collins, J. J. & Wong, W. W. Universal chimeric antigen receptors for multiplexed and logical control of T cell responses. Cell 173, 1426–1438 (2018).

  15. 15.

    Rodgers, D. T. et al. Switch-mediated activation and retargeting of CAR-T cells for B cell malignancies. Proc. Natl Acad. Sci. USA 113, e459–e468 (2016).

  16. 16.

    Cartellieri, M. et al. Switching CAR-T cells on and off: a novel modular platform for retargeting of T cells to AML blasts. Blood Cancer J. 6, e458 (2016).

  17. 17.

    Wu, C. Y., Roybal, K. T., Puchner, E. M., Onuffer, J. & Lim, W. A. Remote control of therapeutic T cells through a small molecule-gated chimeric receptor. Science 350, aab4077 (2015).

  18. 18.

    Riddell, S. R. et al. T cell-mediated rejection of gene-modified HIV-specific cytotoxic T lymphocytes in HIV-infected patients. Nat. Med. 2, 216–223 (1996).

  19. 19.

    Berger, C., Flowers, M. E., Warren, E. H. & Riddell, S. R. Analysis of transgene-specific immune responses that limit the in vivo persistence of adoptively transferred HSV-TK-modified donor T cells after allogeneic hematopoietic cell transplantation. Blood 107, 2294–2302 (2006).

  20. 20.

    Jensen, M. C. et al. Antitransgene rejection responses contribute to attenuated persistence of adoptively transferred CD20/CD19-specific chimeric antigen receptor redirected T cells in humans. Biol. Blood Marrow Transplant. 16, 1245–1256 (2010).

  21. 21.

    Boncompain, G. et al. Synchronization of secretory protein traffic in populations of cells. Nat. Methods. 9, 493–498 (2012).

  22. 22.

    Rollins, C. T. et al. A ligand-reversible dimerization system for controlling protein–protein interactions. Proc. Natl Acad. Sci. USA 97, 7096–7101 (2000).

  23. 23.

    Bouillet, P. & Strasser, A. BH3-only proteins: evolutionarily conserved proapoptotic Bcl-2 family members essential for initiating programmed cell death. J. Cell Sci. 115, 1567–1574 (2002).

  24. 24.

    Leverson, J. D. et al. Exploiting selective BCL-2 family inhibitors to dissect cell survival dependencies and define improved strategies for cancer therapy. Sci. Transl. Med. 7, 279ra240 (2015).

  25. 25.

    Hinds, M. G. et al. BIM, Bad and Bmf: intrinsically unstructured BH3-only proteins that undergo a localized conformational change upon binding to prosurvival Bcl-2 targets. Cell Death Differ. 14, 128–136 (2007).

  26. 26.

    Procko, E. et al. A computationally designed inhibitor of an Epstein–Barr viral Bcl-2 protein induces apoptosis in infected cells. Cell 157, 1644–1656 (2014).

  27. 27.

    Gemperli, A. C., Rutledge, S. E., Maranda, A. & Schepartz, A. Paralog-selective ligands for Bcl-2 proteins. J. Am. Chem. Soc. 127, 1596–1597 (2005).

  28. 28.

    Chin, J. W. & Schepartz, A. Design and evolution of a miniature Bcl-2 binding protein. Angew. Chem. Int. Ed. Engl. 40, 3806–3809 (2001).

  29. 29.

    Silva, D. A., Correia, B. E. & Procko, E. Motif-driven design of protein–protein interfaces. Methods Mol. Biol. 1414, 285–304 (2016).

  30. 30.

    Ku, B., Liang, C., Jung, J. U. & Oh, B. H. Evidence that inhibition of BAX activation by Bcl-2 involves its tight and preferential interaction with the BH3 domain of BAX. Cell Res. 21, 627–641 (2011).

  31. 31.

    Lanier, L. L. DAP10- and DAP12-associated receptors in innate immunity. Immunol. Rev. 227, 150–160 (2009).

  32. 32.

    Ghosh, A., Wang, X., Klein, E. & Heston, W. D. Novel role of prostate-specific membrane antigen in suppressing prostate cancer invasiveness. Cancer Res. 65, 727–731 (2005).

  33. 33.

    Kochenderfer, J. N. et al. Construction and preclinical evaluation of an anti-CD19 chimeric antigen receptor. J. Immunother. 32, 689–702 (2009).

  34. 34.

    Santoro, S. P. et al. T cells bearing a chimeric antigen receptor against prostate-specific membrane antigen mediate vascular disruption and result in tumor regression. Cancer Immunol. Res. 3, 68–84 (2015).

  35. 35.

    Tao, Z. F. et al. Discovery of a potent and selective Bcl-XL inhibitor with in vivo activity. ACS Med. Chem. Lett. 5, 1088–1093 (2014).

  36. 36.

    Feucht, J. et al. Calibration of CAR activation potential directs alternative T cell fates and therapeutic potency. Nat. Med. 25, 82–88 (2019).

  37. 37.

    Griswold, K. E. & Bailey-Kellogg, C. Design and engineering of deimmunized biotherapeutics. Curr. Opin. Struct. Biol. 39, 79–88 (2016).

  38. 38.

    Zhao, H. et al. Depletion of T cell epitopes in lysostaphin mitigates anti-drug antibody response and enhances antibacterial efficacy in vivo. Chem. Biol. 22, 629–639 (2015).

  39. 39.

    Gainza-Cirauqui, P. & Correia, B. E. Computational protein design — the next generation tool to expand synthetic biology applications. Curr. Opin. Biotech. 52, 145–152 (2018).

  40. 40.

    Lee, E. F. et al. High-resolution structural characterization of a helical αβ-peptide foldamer bound to the anti-apoptotic protein Bcl-XL. Angew. Chem. Int. Ed. Engl. 48, 4318–4322 (2009).

  41. 41.

    Fleishman, S. J. et al. RosettaScripts: a scripting language interface to the Rosetta macromolecular modeling suite. PLoS ONE 6, e20161 (2011).

  42. 42.

    Henikoff, S. & Henikoff, J. G. Amino acid substitution matrices from protein blocks. Proc. Natl Acad. Sci. USA 89, 10915–10919 (1992).

  43. 43.

    Kuhlman, B. & Baker, D. Native protein sequences are close to optimal for their structures. Proc. Natl Acad. Sci. USA 97, 10383–10388 (2000).

  44. 44.

    Wang, C., Schueler-Furman, O. & Baker, D. Improved side-chain modeling for protein–protein docking. Protein Sci. 14, 1328–1339 (2005).

  45. 45.

    Gainza, P., Roberts, K. E. & Donald, B. R. Protein design using continuous rotamers. PLoS Comput. Biol. 8, e1002335 (2012).

  46. 46.

    Miller, S., Janin, J., Lesk, A. M. & Chothia, C. Interior and surface of monomeric proteins. J. Mol. Biol. 196, 641–656 (1987).

  47. 47.

    Word, J. M. et al. Visualizing and quantifying molecular goodness-of-fit: small-probe contact dots with explicit hydrogen atoms. J. Mol. Biol. 285, 1711–1733 (1999).

  48. 48.

    Petros, A. M. et al. Solution structure of the antiapoptotic protein Bcl-2. Proc. Natl Acad. Sci. USA 98, 3012–3017 (2001).

  49. 49.

    McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674 (2007).

  50. 50.

    Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D 66, 213–221 (2010).

  51. 51.

    Dong, L. M. et al. Human apolipoprotein E. Role of arginine 61 in mediating the lipoprotein preferences of the E3 and E4 isoforms. J. Biol. Chem. 269, 22358–22365 (1994).

  52. 52.

    Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D 60, 2126–2132 (2004).

  53. 53.

    Brunger, A. T. et al. Crystallography and NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D 54, 905–921 (1998).

  54. 54.

    Gong, M. C., Chang, S. S., Sadelain, M., Bander, N. H. & Heston, W. D. Prostate-specific membrane antigen (PSMA)-specific monoclonal antibodies in the treatment of prostate and other cancers. Cancer Metastasis Rev. 18, 483–490 (1999).

Download references

Acknowledgements

This work was generously supported by the Biltema and ISREC Foundations, an Advanced European Research Council Grant to G.C. (no. 1400206AdG-322875), a Starting European Research Council Grant to B.E.C. (no. 716058), the National Center of Competence for Molecular Systems Engineering, the Ludwig Institute for Cancer Research, EPFL-Fellows grants funded by an H2020 Marie Sklodowska-Curie action to P.G. and J.B., as well as a Whitaker International fellowship to E.G.G. B.-H.O. was supported by the National Research Foundation of Korea (NRF-2018R1A2B3004764). We thank members of the Flow Cytometry Platform and the Animal Care Facility of the University of Lausanne for their excellent support. Computational calculations were performed using the facilities of the Scientific IT and Application Support Center of EPFL. We also acknowledge the EPFL Protein Production and Structure Core Facility for providing access to biophysical instrumentation. Finally, we thank S. Maerkl, M. Lutolf and E. Procko for critical reading of the manuscript and E. Oricchio for valuable discussions.

Author information

G.G.-A., E.G.-G., E. C. and S.S. designed and performed experiments and interpreted results. P.G. performed computational design and interpreted results. S.V., A.J.C.O. and P.R. assisted with the experimental work. J.B. assisted the computational design. S.K., M.K. and B.-H.O. performed X-ray crystallography and structure determination. G.G.-A., E.G.-G. and P.G. wrote the manuscript. M.I., G.C. and B.E.C. designed and supervised the study, interpreted results and wrote the manuscript.

Correspondence to Melita Irving or George Coukos or Bruno E. Correia.

Ethics declarations

Competing interests

EPFL, UNIL and the Ludwig Institute for Cancer Research have filed for patent protection on the technology described herein. G.G.-A., P.G., M.I., G.C. and B.E.C. are named as co-inventors on this patent (United States Patent and Trademark Office Provisional Application: 62/657,534).

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Integrated supplementary information

Supplementary Figure 1 Protein design protocol and sequence alignment of designed scaffolds.

(a) A 12-residue amino acid fragment from the BIM-BH3 interaction was matched against a database of > 11000 proteins using the MotifGraft protocol. Grafted scaffolds were then designed, with their amino acid identities restricted to common subsitutions according to the BLOSUM62 matrix. Designed scaffolds were filtered by three criteria: proteins with a human origin (or with a close human homologue), globularity, and packing of the BH3 motif within the scaffold. (b) Table of designs and scores for the scoring/filtering criteria. Scaffold PDB id: Protein Data Bank id for the protein that was used as a scaffold to design each binder. Scaffold protein name: Brief name of the protein that was used as a scaffold. Organism of scaffold: Species origin of the scaffold. Rosetta ddG: Computed delta-delta G interaction energy between LD[1-3] and Bcl-XL. Globularity: Globularity score for each design. vdW Dots to scaffold: Number of vdW contacts between the grafted motif and the scaffold. SASA of seed: Empirical score that denotes the buried surface area of the grafted motif in the scaffold. # manual reversions to WT: Number of designed positions that were reverted to the scaffold identity. Total # mutations on scaffold: Final number of residues in the scaffold that were mutated (amino acid substitutions) to a different amino acid identity during the design process. (c) Sequence alignment of the three designed scaffolds. A helical 12-residue fragment with the sequence IAXXLXXIGXXF (hotspot residues in orange) was grafted onto three different scaffolds for LD1, LD2, and LD3 respectively: Syntaxin 6 (Syn6, PDB ID: 1LVF); human focal adhesion targeting domain of Pyk2 (pyk2, PDB ID: 3GM2); and Apolipoprotein E (ApoE, PDB ID: 1LE4). Hotspot residues are shown in orange while designed residues are shown in bold. The sequence of BIM-BH3 is shown as a reference in the third line.

Supplementary Figure 2 Biochemical characterization of computationally designed binders.

(a) SPR sensorgrams of the three designs injected over immobilized Bcl-XL (measurement performed once for each complex). Black dashed curves show the sensorgrams and the red curves show the associated kinetic fits (2-state model was used to fit LD1, and 1:1 model was used to fit LD3). For LD1, the concentrations of analyte tested ranged from 1 µM to 31.25 nM varied in 2-fold dilutions. No binding was detected for LD2 upon the injection of concentrations up to 2 µM. LD3 binds to Bcl-XL with a KD of 3.9 pM, injections of analyte range from 250 nM to 7.8125 nM varied in 2-fold dilutions. (b) LD3 analyzed using Circular Dichroism spectroscopy showed a spectrum typical of a helical protein (experiment performed once). The melting temperature of LD3 was 59 °C. (c) SEC-MALS analysis (performed once) showed that Bcl-XL and LD3 are monomers in solution (left and center panels). (d) Apparent IC50s for the drug inhibiting activity were measured by Surface Plasmon Resonance (n=3 biological replicates). Different drug dilutions were pre-incubated with LD3, and the mixture was injected over immobilized Bcl-XL. Apparent IC50s were calculated by using the response units at 120 seconds. The structures of the drugs used for this study are shown.

Supplementary Figure 3 LD3:Bcl-2 crystal structure comparison with the model.

(a) Comparison of crystal density of LD3 (green mesh) with the LD3 model (blue tubes). The molecular surface of Bcl-2 from the crystal structure is shown in white. (b) Comparison of the grafted 12-amino acid motif between crystal density (green mesh) and model (blue). Bcl-2 from the crystal structure is shown in white tubes.

Supplementary Figure 4 STOP-CAR prototypes comprising either c-Myc alone or c-Myc plus the CH2-CH3 linker region in the ectodomain of the S-chain yield low transduction efficiencies in primary human T cells.

(a) Top, schematic of R- and S-chain structure for the first STOP-CAR prototype-1 (Proto-1) tested (no experimental replicate); bottom, cell-surface expression on Jurkat reporter cells following transduction with a single lentiviral vector encoding both chains. Anti-F(ab’) antibody recognizes the scFv on R-chain, anti-c-Myc-mAb recognizes the c-Myc tag on S-chain (b) Cell-surface localization of Proto-1 chains on the surface of Jurkat cells as determined by Amnis® flow imaging following staining with anti-human-F(ab’)-Ab-FITC and anti-c-Myc-mAb-APC (for R- and S-chains, respectively). Four representative single cells are shown, each in a different row (no experimental replicate). (c) Activation of 6xNFAT-mCherry-engineered Jurkat cells transduced by Proto-1 STOP-CAR (upon CAR engagement and cellular activation the transcription of mCherry takes place and the cells become red). PSMA+ MS1 were used as target cells and phorbol 12-myristate 13-acetate (PMA)/Ionomycin(IONO) stimulation was used as a reference for fully activated Jurkat cells. Left, Representative flow cytometry plots of the mCherry-expressing activated Jurkat cells. T cells incubated with PSMA+ MS1 cells were transduced with a single lentiviral vector encoding both chains. Right, Comparison of activation of UTD, cells co-expressing R-chain and S-chain transduced by the same vector (R-T2A-S), cells expressing only R- or S-chains transduced individually, and cells co-expressing R-chain and S-chain through co-transduction with two vectors (R+S-chain). Values are the mean of n=2 technical replicates. (d) IL-2 production as measured by ELISA for Jurkat cells expressing the different constructs. Values are the mean of n=2 technical replicates. (c), (d) The percentage of mCherry expressing Jurkat cells and relative IL-2 production were normalized relative to PMA/IONO stimulation, which was set at 100% and 1, respectively. (e) Left, Proto-1 stability in Jurkat cells over 30 days. Right, Amnis® imaging at day 30. Four representative single cells are shown, each cell in a different row (no experimental replicate). (f) Transduction efficiency of R- and S-chains of Proto-1 on primary T cells averaged 80% and 4%, respectively, as determined by flow cytometric analysis. Values are the mean ± s.e.m. of n=3 human donors. (g) Top, Vector scheme of prototype-2 (Proto-2). Bottom, Proto-2 cell-surface expression on Jurkat reporter cells (no experimental replicate). (h) Proto-2 transduction efficiency of R- and S-chains on primary T cells averaged 4% and 6%, respectively, as determined by flow cytometric analysis. Values are the mean of n=2 human donors.

Supplementary Figure 5 STOP-CAR prototype comprising the DAP10 ectodomain on the S-chain achieves efficient and stable expression on the surface of Jurkat and primary human T-cells over time.

(a) Cell transduction protocol. CD4+ and CD8+ T-cells bead-enriched by negative selection were stimulated overnight with anti-CD3/anti-CD28 beads in the presence of human (h)IL-2 and then lentivirally transduced. On day 5, the beads were removed and hIL-7 and IL-15 were added to the culture. Assays were performed on day 10. (b) STOP-CAR cell-surface expression by Jurkat reporter cells on days 15 and 30, as determined by flow cytometric analysis of R- and S-chain, detected with anti-F(ab’)-Ab-APC and anti-c-Myc-mAb-APC staining, respectively (two experimental replicates). (c) Representative STOP-CAR expression by primary human T-cells on day 15. (d) Left, center, 2G-CAR expression and R/S-chain (STOP-CAR) co-expression on primary human T cells on day 5. Right, Expression of the 2G and STOP CAR constructs in primary CD4+ and CD8+ T cells. (UTD: untransduced cells). Values are reported as mean ± s.e.m. of n=6 (2G) and n=7 (STOP) human donors. (e) Flow cytometric analysis on day 10 of differentiation markers in primary human STOP-CAR transduced CD3+ T cells. Left, gating scheme using CCR7 and CD45RA expression to delineate percentages of Central Memory (TCM), T Naïve (TN), Terminally Differentiated Effector Memory (TEMRA) and, T Effector Memory (TEM) cells. Center, representative data from one donor. Right, STOP-CAR-Ts and 2G-CAR-Ts exhibit a similar memory phenotype as UTD-Ts in the bulk T-cell population following expansion. Values are the mean ± s.e.m. of n=4 human donors.

Supplementary Figure 6 STOP-CAR-Ts and 2G-CAR-Ts targeting PSMA are only activated in the presence of PSMA+ tumor cells, and exhibit similar secretion of effector molecules.

(a) Flow cytometric analysis of PC3 cells and (b) PC3-PIP cells stained with anti-PSMA-Ab-PE (two experimental replicates). (c) IFN-γ production by STOP-CAR-Ts or 2G-CAR-Ts in the presence of PC3-PIP or PC3 cells. Values are reported as mean ± s.e.m. of n=5 human donors. (d) Cytokine bead array (CBA) and flow cytometric analysis to measure cytokine secretion by 2G-CAR and STOP-CAR-Ts upon exposure to PC3-PIP co-culture. Values are reported as mean ± s.e.m. of n=4 human donors. Two-sided, unpaired parametric t test was used to compare 2G vs STOP. Statistical significance: Granzyme A P =0.2154; Granzyme B P =0.3068; IL-2 P =0.6427.

Supplementary Figure 7 Concentrations above 10 μM of Drug-1 and -2 are toxic in vitro to PC3-PIP tumor cells and impair primary human T-cell function.

(a) Left, IncuCyte measurements of tumor cell death (as measured by total red area/μm2) for PC3-PIP cells at 24h co-incubation with decreasing concentrations of Drug-1 and -2. (NI: non-interpretable). Values are represented as mean of two technical replicates. Right, Representative images of tumor cell death under the different conditions at 24h (Scale = 300 μm). (b) IncuCyte measurements of CD4+ and CD8+ T cell death (as measured by total red area/μm2) over 24h co-incubation with decreasing concentrations of Drug-1 and -2. Values are the mean ± s.e.m. of n=4 human donors. (c) Fold-expansion and cell diameter of CD4+ and CD8+ Ts. At 100 μM and 1.5 mM Drug-2, physical properties of T-cells are significantly impaired as compared to untreated, or 10 μM Drug-2 treated cells. Values are the mean ± s.e.m. of n=3 human donors. A Two-way ANOVA with Post-hoc Tukey test was used for statistical analysis. Statistical significance for fold expansion: CD4+ **P =0.0016, CD8+ ***P =0.0007; for diameter: ****P ≤ 0.0001.

Supplementary Figure 8 STOP-CAR-T cytotoxicity is not attenuated in the presence of 10 μM Drug-1, while lower concentrations of Drug-2 partially impair STOP-CAR-T activity.

(a) IncuCyte analysis of STOP and 2G-CAR-Ts in presence of 10 μM Drug-1 or (b) 5 μM Drug-2. (a), (b) Values are the mean ± s.e.m. of n=4 (2G) and n=5 (STOP) human donors. A Two-way ANOVA with Post-hoc Tukey test was used for statistical analysis. Statistical significance in panel b comparing STOP vs STOP + Drug: 22h *P=0.0393; 24h **P=0.0075; 26h **P=0.0027; 28h ***P=0.0003; 30h ****P≤0.0001.

Supplementary Figure 9 STOP-CAR-Ts recognize and respond to PSMA+ 22Rv1 tumor cells.

(a) Flow cytometric analysis of anti-PSMA-Ab-PE stained 22Rv1 cells (experimental replicates=2). (b) mCherry expression in UTD, 2G-CAR and STOP-CAR-engineered Jurkat reporter cells following 48h co-culture with 22Rv1 cells. Effector:Target ratio 2:1. Values are the mean of technical replicates=2. (c) Relative IFN-γ and IL-2 production by STOP-CAR-Ts and 2G-CAR-Ts upon co-culture with 22Rv1 cells during 24h. Values are the mean of n=2 human donors. Cytokine production was normalized as a ratio relative to the maximum quantity produced by each donor. (d) IncuCyte evaluation of 22Rv1 cell-death by STOP-CAR-Ts and 2G-CAR-Ts in the absence and presence of 10 µM Drug-2 at 22h and 42h, Effector:Target ratio 2:1. Values are the mean ± s.e.m. of n=3 human donors. A Two-way ANOVA with Post-hoc Tukey test was used for statistical analysis. Statistical significance in panel d comparing STOP vs STOP + Drug and 2G vs 2G + Drug. Statistical significance: **P = 0.0042, P = 0.9617. Representative images of STOP-CAR-T and 2G-CAR-T killing at 0 and 42h in the absence of Drug-2 (Scale = 300 µm).

Supplementary Figure 10 CD19-directed STOP-CAR-Ts confer similar in vitro activity levels as 2G-CAR-Ts against target cells, and their activity can be tuned down by Drug-2.

(a) Schematic representation of the CD19-directed STOP CAR (19-STOP) constructs. (b) Flow cytometric evaluation of R- and S-chain co-transduction on primary CD4+ and CD8+ T cells. Left, dot plots of CD4+ and CD8+ T-cells stained by anti-human F(ab’)-Ab-APC and anti-c-Myc-Ab-FITC. Center, right, transduction efficiency of a 2G anti-CD19-CAR (19-2G) and 19-STOP-CAR. UTD T cells are shown as control. Values are the mean ± s.e.m. of n=6 (CD4+ 19-2G) and n=7 (CD4+ 19-STOP); n=6 (CD8+ 19-2G and 19-STOP) human donors. (c) UTD, 19-2G and 19-STOP T cells show similar expansion. Values are the mean ± s.e.m. of n=6 human donors. (d) Left, Expression level of the 19-STOP after 5, 15 and 30 days. Right, In vitro cytotoxicity assay of UTD, 19-STOP or 19-2G CAR T cells (all tested 30 days after CAR transduction) against BV173 cells. Values are the mean ± s.e.m. of n=3 human donors. Statistical significance was assessed using One-Way ANOVA; 19-2G vs 19-STOP, ****P ≤ 0.0001. (e) UTD, 19-2G and 19-STOP CAR-Ts present similar T cell memory/effector phenotypes (TCM: Central Memory, TN: T Naïve, TEMRA: Terminally Differentiated Effector Memory, TEM: T Effector Memory). Values are the mean ± s.e.m. of n=3 human donors. (f) CD19 expression on BV173 tumor cells assessed by flow cytometric analysis (experimental replicates=2). (g) Killing assay of BV173 tumor cells by UTD, 19-2G and 19-STOP-CAR-Ts (IncuCyte measurement, calculated from total red area/μm2) Values are the mean ± s.e.m. of n=3 human donors. Statistical significance was assessed using Two-Way ANOVA and Post-hoc Tukey test; 19-2G vs 19-STOP, P = 0.3513 at 40h. (h) Cytometric bead array and flow cytometric analysis to measure cytokine secretion levels by 19-2G- and 19-STOP-CAR-T cells upon co-culture with BV173 cells for 24h. Values are the mean ± s.e.m. of n=3 human donors. Two-sided, unpaired parametric t test was used to compare 19-2G vs 19-STOP . Statistical significance: Granzyme A P =0.8604; Granzyme B P =0.6716; IFNγ P =0.4158; IL-2 *P =0.0346. (i) Short term cytotoxicity assessment. FACS analysis of residual CD19+ target cells (BV173) after 5h of co-culture with UTD, 19-2G and 19-STOP-CAR-Ts preconditioned for 12 hours or not with 10 μM Drug-2. (j) Left, CD19 expression on Bjab tumor cells assessed by flow cytometric analysis. Right, Short term cytotoxicity assessment. FACS analysis of residual CD19+ target cells (Bjab) after 5h of co-culture with UTD, 19-2G and 19-STOP-CAR-Ts preconditioned for 12 hours or not with 10 μM Drug-2. (i-j) Values are the mean ± s.e.m. of n=3 human donors. A One-way ANOVA was used for statistical analysis. Statistical significance: (i) comparing 19-STOP vs 19-STOP + Drug **P = 0.0028 and 19-2G vs 19-2G + Drug P = 0.9991; (j) comparing 19-STOP vs 19-STOP + Drug **P = 0.0098 and 19-2G vs 19-2G + Drug P = 0.2001.

Supplementary Figure 11 In vivo testing of Drug-2 toxicity, and in vivo activity of STOP-CAR T cells in the presence or absence of Drug-2.

(a) No toxicity was observed in 8-12 week-old male NSG mice injected daily for 5 days with Drug-2 at 1.5 mg/kg or 2.5 mg/kg as assessed by body weight as well as behavioral and physical observations. (b) There was no impairment in subcutaneous PC3-PIP tumor growth in male NSG mice receiving 1 week of daily Drug-2 injections of up to 5 mg/kg, nor did the mice show signs of distress. (a), (b) Values are the mean ± s.e.m. of n=5 mice per group. Statistical significance was determined by Two-way ANOVA: for panel (b) vehicule vs 2.5 mg/kg P = 0.7773 and vehicule vs 5 mg/kg P = 0.1672. (c) Top, cartoon of the experimental design for in vivo activity experiment. Bottom, Winn assay in which NSG mice were co-injected with tumor cells and UTD or CAR-Ts. Values are the mean ± s.e.m. of n=6 mice per group. Statistical significance was determined by Two-way ANOVA: STOP vs UTD considering all Ts groups, at day 9 *P = 0.0450, day 11 *P = 0.0175, day 14 **P = 0.0019 and day 17 *P = 0.0139; comparing STOP vs UTD only, ***P = 0.0002 at day 11 and ****P ≤ 0.0001 at day 14-17. (d) Top, cartoon of the experimental design for in vivo activity experiment. NSG mice were subcutaneously inoculated with PC3-PIP tumor cells, and on day 5 received CAR-Ts or UTD-Ts, followed by daily subcutaneous peritumoral injections of 10 µM Drug-2, or vehicle. Bottom, In vivo tumor control by 2G and STOP-CAR-Ts in the absence or presence of Drug-2. These data show a second replicate of the experiment depicted in Fig. 3c. Values are the mean ± s.e.m. of n=7 mice per group. Statistical significance was determined by Two-way ANOVA: STOP vs STOP + Drug, at day 17 *P = 0.0429 when comparing all Ts groups; STOP vs STOP + Drug only, ****P ≤ 0.0001 at day 17. (e) Left, cartoon of the experimental design for in vivo dynamic response. NSG mice were subcutaneously inoculated with PC3-PIP tumor cells and on day 5 received CAR-Ts or UTD-Ts. One mouse group received 10 µM Drug-2 from day 5 until day 11 only, while another group received Drug-2 only from day 11 on. Right, Tumor control over time. Values are the mean ± s.e.m. of n=4 mice per group. These data show a second replicate of the experiment depicted in Fig. 3f. Statistical significance was determined by One-Way ANOVA: at day 11 STOP vs STOP + Drug day 5-11 *P =0.0232, STOP vs STOP + Drug day 11-15 P =0.8275; at day 15 STOP vs STOP + Drug day 5-11 P =0.0789, STOP vs STOP + Drug day 11-15 *P =0.0284.

Supplementary Figure 12 STOP-CAR-Ts regain effector molecule secretion post-withdrawal of disruptor drug.

CBA assay to measure cytokine secretion by STOP-CAR-Ts pre-exposed to Drug-2 (experimental set-up illustrated in Fig. 3d). Values are the mean ± s.e.m. of n=3 human donors. Statistical significance was determined by Two-sided, unpaired parametric t test comparing UTD vs Pre-Exp STOP: Granzyme A *P =0.0166; Granzyme B ***P =0.0005; IFN-γ ***P =0.0001; IL-2 *P =0.0357.

Supplementary information

Supplementary Information

Supplementary Figs. 1–12 and Supplementary Note 1.

Reporting Summary

Supplementary Table

Crystallography collection and refinement.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Giordano-Attianese, G., Gainza, P., Gray-Gaillard, E. et al. A computationally designed chimeric antigen receptor provides a small-molecule safety switch for T-cell therapy. Nat Biotechnol (2020). https://doi.org/10.1038/s41587-019-0403-9

Download citation