Engineering virus-specific T cells in vivo for CAR T cell therapy

Introduction:: Engineered T cells that express chimeric antigen receptors (CARs) have shown remarkable efficacy against hematological malignancies. However, broad implementation of CAR T cell therapies is limited by the lengthy (3–5 weeks) and costly ($350K–450K per treatment) ex vivo manufacturing pipeline. Here, we developed antigen-presenting nanoparticles (APNs) for in situ programming of virus-specific T cells for rapid and cost-efficient CAR T cell manufacturing. Virus-specific T cells present a promising opportunity to enhance CAR T cell therapy, as they have improved persistence and proliferation potential, and allow for viral vaccination to augment CAR therapy through their endogenous receptors. We focused on influenza A virus (IAV)-specific T cells to exploit the existing seasonal influenza vaccination to boost CAR activities. To deliver CAR to IAV-specific T cells, APNs will comprise lipid nanoparticles (LNPs) that encapsulate CAR-encoded mRNA and are decorated with peptide-major histocompatibility complex (pMHC) displaying influenza peptide epitopes (Fig. 1). We used APNs to deliver human B-cell maturation antigen (BCMA) CAR in the context of multiple myeloma with future goals to expand to other CAR specificities and indications, including CD19 positive cancers.


Materials and Methods:: We expressed the heavy chain and light chain of pMHC by E. coli and refolded into pMHC with selected peptide antigen. In parallel, we synthesized CAR mRNA-loaded LNPs using the NanoAssemblr® Ignite (Precision Nanosystem). We subsequently functionalized the surface of LNPs with pMHC through thiol-Maleimide chemistry to form APNs. To test APNs for transfecting antigen-specific T cells with CAR mRNA, we injected APNs into P14 TCR transgenic mice whose CD8+ T cells express a TCR that specifically recognizes the LCMV GP33/Db antigen¹. At 24 hr after injection, we isolate CD8 T cells from splenocytes and co-incubated with EL4 cells expressing mouse CD19 antigen. We evaluated the resulting cytotoxicity using LDH cytotoxicity assay. To test APNs for transfecting human IAV-specific T cells with αBCMA CAR mRNA, we functionalized APNs with HLA-A2.1 – an allele expressed by >30% of the US population² – to present an immunodominant IAV peptide (GILGFVFTL). We injected this APNs into NSG mice pre-infused with IAV-specific T cells. At 24 hr after injection, splenocytes were isolated and stained with PE-conjugated recombinant BCMA proteins for evaluating CAR transfection efficiency through flow cytometry.


Results, Conclusions, and Discussions:: Results & Discussions

To evaluate APNs for in situ CAR gene delivery, I administered GP33/D^b APNs encapsulating αCD19 CAR mRNA to P14 mice, which are TCR-transgenic for GP33 peptide epitope (Fig. 2A), or GP100/D^b APNs as the non-cognate control. After 24 hours, splenic CD8 T cells were stained for CAR expression and co-incubated with target CD19+ or CD19- EL4 cells. GP33/D^b APNs achieved ~30% transfection efficiency (Fig. 2B) and transfected T cells led to significantly higher cytotoxicity when co-incubated with CD19+ EL4 cells ex vivo compared to T cells transfected with non-cognate GP100/D^b APNs (Fig. 2C). These data demonstrate the ability of APNs in programming antigen-specific T cells with functional mouse αCD19 CAR in vivo.

To target human IAV-specific T cells, we first expanded IAV-specific T cells by peptide pulse to mimic the influenza vaccine that will be given to patients before APN injection in the intended use of the proposed approach. To mimic the IAV-specific T cell frequency after vaccination in human (~0.5-1%)³, we infused NSG mice with 12 million IAV-peptide pulsed PBMC (containing ~1.75 million IAV-specific CD8 T cells), resulting in IAV-specific CD8 T cells to be ~1% of total splenocytes at 48 hr after the cell transfer (Fig. 3A, B). Even at such low frequency, intravenous injection of HLA/IAV APNs to these NSG mice resulted in ~90% transfection of IAV-specific CD8 T cells with αBCMA CAR. By contrast, PBS, non-cognate control APNs, and APNs carrying αCD19 CAR mRNA did not result in detectable αBCMA CAR expression in IAV-specific CD8 T cells (Fig. 3C, D).



Conclusions

We demonstrate an approach to engineer antigen-specific T cells with functional CAR mRNA in vivo. Our future work includes validating the anti-cancer efficacy of in vivo transfected CAR T cells using mouse models recapitulating human multiple myeloma, and evaluating the potential of using influenza vaccination to boost CAR T cell therapy. Looking forward, we envision that our approach for antigen-specific T cell modulation would be an alternative for ex vivo engineered T cell therapies and serve as a modular tool to program T cell immunity to effect therapeutically relevant outcomes.


Acknowledgements (Optional): : This work was supported in part by NIH grants R01CA237210, R21CA260247, R01EB032822, DP1CA280832, R01AI171892, and U01CA265711; NSF grant ECCS-1542174; The International Myeloma Society (IMS) and Paula and Rodger Riney Foundation Translational Research Award; and the Donaldson Charitable Trust Research Synergy Fund (G.A.K.). F.Y.S. acknowledges support by the Postdoctoral Fellowship from the Department of Biomedical Engineering at Georgia Tech and the College of Engineering at Peking University, China.


References (Optional): : 1 Puglielli, M. T. et al. In vivo selection of a lymphocytic choriomeningitis virus variant that affects recognition of the GP33-43 epitope by H-2Db but not H-2Kb. Journal of virology 75, 5099-5107, doi:10.1128/JVI.75.11.5099-5107.2001 (2001).

2 Ellis, J. M. et al. Frequencies of HLA-A2 alleles in five U.S. population groups: Predominance of A∗02011 and identification of HLA-A∗0231. Human Immunology 61, 334-340, doi:https://doi.org/10.1016/S0198-8859(99)00155-X (2000).

3 He, X.-S. et al. Phenotypic Changes in Influenza-Specific CD8+ T Cells after Immunization of Children and Adults with Influenza Vaccines. The Journal of infectious diseases 197, 803-811, doi:10.1086/528804 (2008).