In vivo programmed myeloid cells expressing novel chimeric antigen receptors show potent anti-tumor activity in preclinical solid tumor models

Abstract

Introduction: The approval of chimeric antigen receptor (CAR) T cell therapies for the treatment of B cell malignancies has fueled the development of numerous ex vivo cell therapies. However, these cell therapies are complex and costly, and unlike in hematological malignancies, outcomes with most T cell therapies in solid tumors have been disappointing. Here, we present a novel approach to directly program myeloid cells in vivo by administering novel TROP2 CAR mRNA encapsulated in lipid nanoparticles (LNPs).

Methods: The CAR comprises a TROP2 specific single-chain variable fragment (scFv) fused to a truncated CD89 which requires association with the FcRγ signal adapter to trigger myeloid-specific cell activation. The mRNA encoding the TROP2 CAR was encapsulated in LNPs. Co-immunoprecipitation, flow cytometry and enzyme-linked immunosorbent assay (ELISA) were used to measure CAR expression and functional activity in vitro. Anti-tumor efficacy of the TROP2 CAR mRNA/LNP was evaluated after intravenous administration in various murine tumor models.

Results: In vitro, transient expression of the TROP2 CAR on monocytes triggers antigen-dependent cytotoxicity and cytokine release. In tumor bearing mice and cynomolgus monkeys, the TROP2 CAR mRNA/LNP are primarily expressed by myeloid cells. In a mouse xenograft model, intravenous administration of TROP2 CAR mRNA/LNP results in tumor growth inhibition and in a B16/F10-OVA immunocompetent melanoma mouse model, anti-tumor efficacy of a gp75-specific CAR correlates with increased number of activated T cells, activation of dendritic cells and a humoral response against B16/F10-OVA melanoma tumors.

Discussions: These findings demonstrate that myeloid cells can be directly engineered in vivo to kill tumor cells and orchestrate an adaptive immune response and guide clinical studies for the treatment of solid tumors.

Keywords: RNA therapeutics; cancer therapy; cell therapy; lipid nanoparticles; myeloid cells.

Figure 1

Development of CD89 (FcαR) based CAR targeting TROP2. (A) Design of the CD89 fusion CAR. Human IgA receptor FcαR (CD89) forms multi-chain complexes with endogenous FcRγ. CD89-based CAR is constructed by replacing the ectodomain of CD89 with a tumor-targeting scFv. (B) TROP2-CD89 CAR was delivered to primary human monocytes via mRNA electroporation and CAR expression was measured using flow cytometry. Data shown are representative of two different donors. Also see Supplementary Figure 1A . (C) Expression level and duration of TROP2 CAR are influenced by the presence of FcRγ. TROP2 CAR mRNA was transfected into Jurkat, Huh7 or 293T cells with or without co-transfection of FcRγ mRNA. Expression of TROP2 CAR was measured using flow cytometry over 3 days. (D) Association of TROP2 CAR with endogenous FcRγ detected by co-immunoprecipitation. Western Blot was performed against the FcRγ chain with αGAPDH as loading control. (E) TROP2 CAR induced NF-κB and IFN-I pathways activation as measured in the THP1-Dual™ reporter cells. Stimulation with the TLR4 agonist LPS or cGAS agonist 2’3’-cGAMP are shown as positive control. Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test for comparing unstimulated and TROP2-stimulated conditions, and an unpaired t-test (two-tailed) for comparing no agonist vs agonist stimulated conditions. (F) Co-culture of PBMCs electroporated with TROP2 CAR with SKOV3 tumor cells results in elevated tumor killing compared to mock electroporated PBMCs. Statistical analysis was performed using an unpaired t-test (two-tailed). (G) Co-culture of PBMCs electroporated with TROP2 CAR with SKOV3 tumor cells results in production of pro-inflammatory cytokines and chemokines. Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test. All data shown are mean ± SD. Statistical comparison with p value <0.05 are labeled.

Figure 2

Myeloid cell-targeted delivery of TROP2 CAR mRNA/LNP. (A) In vivo CAR expression in immune competent mice resulted from MT-302 infusion. Non-tumor bearing C57BL/6 mice were given a single i.v. injection of MT-302 at 0.5 mg/kg or 1 mg/kg. Blood, spleen, bone marrow and liver were collected from treated mice at 6 hours post-infusion. Single cell suspensions were prepared and stained for CAR expression among immune subsets (T cells, B cells, NK cells, Neutrophils, Monocytes, DCs) by FACS. Data shown were average and STD of each group (n=3 mice per group). Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test. (B) In vivo CAR expression in non-human primate resulted from MT-302 infusion. Naïve cynomolgus monkeys were infused with MT-302 (i.v.) (0.1, 0.5, or 1 mg/kg) for 1h. Blood was collected 12 hours after infusion and TROP2 CAR surface expression of different immune populations of PBMCs was analyzed by FACS. Data shown were average and STD of each group (10 monkeys, 5 females and 5 males). Statistical analysis was performed by Two-Way ANOVA followed by Dunnett’s multiple comparisons test. (C) CAR expression in human blood resulted from in vitro transfection. Whole blood from healthy human donors was incubated with 125 µg/mL of MT-302 for 3 hours in presence of ApoE3 (at 1.5 µg/mL). Following RBC lysis and additional 3 hours culture, surface TROP-2 CAR expression was assessed by FACS for T cells, B cell, NK cells, Granulocytes and Monocytes. Data shown were average and STD of each group (n=3 donors). Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test.

Figure 3

MT-302 shows robust anti-tumor activity in TROP2+ xenograft models. (A) Anti-tumor efficacy of MT-302 in vivo. NCG mice with established HCC-1954 tumor (TROP2+, s.c.) were treated with MT-302 (1 mg/kg/dose, total 5 doses) or PBS every 4 days (Q4D). Tumor growth was monitored by caliper measurement twice a week. Data shown were average and STD of each group (n= 5 mice per group). (B) CAR expression detected in monocytes in vivo following MT-302 infusion. NCG mice with established HCC-1954 tumor (tumor volume approximately 500 mm³) were dosed i.v. with MT-302 (1 mg/kg) or PBS. Six hours post infusion, blood, spleen, bone marrow and tumor were harvested. Single cell suspensions were prepared from collected tissues and stained for CAR expression. Data shown were average and STD of each group (n= 3 mice per group) for monocytes (CD45+Ly6C+ cells). Statistical analysis was performed by unpaired T-test (two-tailed) for tumor samples, and by Two-Way ANOVA followed by Sidak’s multiple comparisons test for blood, spleen, and bone marrow samples. (C) Anti-tumor efficacy of bi-weekly injection of MT-302 in vivo. NCG mice with established HCC-1954 tumor (TROP2+, s.c.) were treated with MT-302 (1 mg/kg/dose, total 3 doses) or PBS every 2 weeks (Q2W). Tumor growth was monitored by caliper measurement twice a week. Data shown were average and STD of each group (n= 5 mice per group). Statistical analysis was performed by Ordinary One-Way ANOVA.

Figure 4

gp75 CAR mRNA/LNP shows anti-tumor activity in the B16/F10 syngeneic melanoma model and remodels the TMEs. (A) Anti-tumor efficacy of surrogate gp75 CAR mRNA/LNP in syngeneic mouse melanoma model. C57BL/6 mice were inoculated s.c. with B16/F10-OVA tumor cells on day 0. Upon tumor establishment, PBS, Empty LNP or gp75 CAR mRNA/LNP were injected i.v. at 2 mg/kg/dose every 2 days. Data shown are mean ± SEM (n=6). (B, C) Flow cytometry analysis of CD8⁺ T cell phenotypes in tumors 24 h post 4th dose. (B) Treatment with gp75 CAR mRNA/LNP significantly increased the frequency of proliferating (CD8⁺ Ki67⁺) and cytolytic (Granzyme B⁺) CD8⁺ T cells, and simultaneously reduced the percentage of TIM3⁺ and TOX⁺ exhausted T cells. Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test. (C) Treatment with gp75 CAR mRNA/LNP reduces the percentage of memory CD44⁺ CD8⁺ T cells with exhaustion phenotype indicated by high PD1 expression. Statistical analysis was performed by Two-Way ANOVA followed by Sidak’s multiple comparisons test. (D) Treatment with gp75 LNP significantly increased activated dendritic cells (CD40⁺ CD86⁺). Statistical analysis was performed using an unpaired t-test (two-tailed).