β-Adrenergic Signaling in Mice Housed at Standard Temperatures Suppresses an Effector Phenotype in CD8+ T Cells and Undermines Checkpoint Inhibitor Therapy

Abstract

The immune context of tumors has significant prognostic value and is predictive of responsiveness to several forms of therapy, including immunotherapy. We report here that CD8⁺ T-cell frequency and functional orientation within the tumor microenvironment is regulated by β₂-adrenergic receptor (β-AR) signaling in host immune cells. We used three strategies-physiologic (manipulation of ambient thermal environment), pharmacologic (β-blockers), and genetic (β₂-AR knockout mice) to reduce adrenergic stress signaling in two widely studied preclinical mouse tumor models. Reducing β-AR signaling facilitated conversion of tumors to an immunologically active tumor microenvironment with increased intratumoral frequency of CD8⁺ T cells with an effector phenotype and decreased expression of programmed death receptor-1 (PD-1), in addition to an elevated effector CD8⁺ T-cell to CD4⁺ regulatory T-cell ratio (IFNγ⁺CD8⁺:Treg). Moreover, this conversion significantly increased the efficacy of anti-PD-1 checkpoint blockade. These data highlight the potential of adrenergic stress and norepinephrine-driven β-AR signaling to regulate the immune status of the tumor microenvironment and support the strategic use of clinically available β-blockers in patients to improve responses to immunotherapy. Cancer Res; 77(20); 5639-51. ©2017 AACR.

Figure 1. Reducing β-adrenergic signaling by physiologic, pharmacologic, and genetic strategies controls tumors at 22°C comparable to 30°C

(A) Norepinephrin e levels in the serum of 4T1 tumor-bearing mice housed at 22°C and 30°C. (B) Experimental design for β blocker studies with propranolol at 22°C and 30°C. Mice were acclimated to 22°C or 30°C and were then treated with or without daily propranolol 4 days before tumor challenge until the experimental endpoint. (C) B16-OVA and (D) 4T1 tumor growth in wildtype mice housed at 22°C or 30°C treated with or without propranolol. (E) 4T1 tumor growth in wildtype BALB/c and Adrb2^−/− mice housed at 22°C or 30°C or (F) housed at 22°C treated with or without propranolol. Data are presented as mean ± SEM. Comparison of norepinephrine levels by Student’s t-test. N = 4–5 per group. Tumor growth statistics analyzed using two-way ANOVA with Tukey analysis. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. N = 4–8 per group.

Figure 2. Reduced tumor growth in propranolol treated mice depends on the adaptive immune system and is CD8⁺ T cell dependent

(A) B16-OVA and (B) 4T1 tumor growth in SCID mice housed at 22°C or 30°C treated and treated with PBS or. (C) B16-OVA and (D) 4T1 tumor growth in wildtype C57BL/6 mice and wildtype BALB/c mice respectively depleted of CD8⁺ T cells and treated with or without propranolol. (E) 4T1 tumor growth in wildtype BALB/c and Adrb2^−/− mice depleted of CD8⁺ T cells. Data are presented as mean ± SEM. Tumor growth statistics analyzed using two-way ANOVA with Tukey analysis. * P < 0.05, *** P < 0.001, **** P < 0.0001. N = 5–8 per group.

Figure 3. Frequencies of intra-tumoral effector T cells are increased in β-blocker treated mice

Single cell suspensions were made from B16-OVA tumors from mice housed at 22°C treated with PBS or propranolol and analyzed by flow cytometry. (A) Representative flowplots of T-bet expression in CD8⁺ TILs. (B) Frequency and (C) number of T-bet⁺ CD8⁺ T cells per mg of tumor. (D) Representative flowplots of IFN-γ expression in CD8⁺ TILs. (E) Frequency and (F) number of IFN-γ⁺ CD8⁺ T cells per mg of tumor. (G) Representative flowplots of GzmB expression in CD8⁺ TILs. (H) Frequency and (I) number of GzmB⁺ CD8⁺ T cells per mg of tumor. (J) Representative histogram of GzmB expression in CD8⁺ TILs. (K) Quantification of the GzmB MFI in the GzmB⁺ CD8⁺ TIL populations. All data are presented as mean ± SEM. Statistics analyzed by Student’s t-test: * P < 0.05, ** P < 0.01, *** P < 005. N = 5–7 per group.

Figure 4. Treating B16-OVA tumor-bearing mice at 22°C with β-blockade significantly increases the ratio of CD8 T cells to suppressive Tregs

Single cell suspensions were made from B16-OVA tumors from mice housed at 22°C treated with or without propranolol and analyzed by flow cytometry. (A) Representative flowplots of FoxP3 expression in CD4⁺ TILs. (B) Percentage CD4⁺ TILs positive for FoxP3 staining. The ratio between (C) total CD8⁺ T cells and (D) effector IFN-γ⁺ CD8⁺ T cells to Tregs in B16-OVA tumors. All data are presented as mean ± SEM. Statistics analyzed by Student’s t-test: * P < 0.05, *** P < 005. N = 5–7 per group.

Figure 5. Expression of PD-1 by effector CD8⁺ TILs is reduced by propranolol treatment

Single cell suspensions were made from B16-OVA tumors from mice housed at 22°C treated with PBS or propranolol and analyzed by flow cytometry. Representative flowplots of PD-1 expression on CD8⁺ TILs isolated from (A) 4T1 and (C) B16-OVA tumors. Quantification of CD8⁺ TILs expressing PD-1 isolated from (B) 4T1 and (D) B16-OVA tumors. Representative histograms of PD-1 expression on CD8⁺ TILs isolated from (E) 4T1 and (G) B16-OVA tumors. Quantification of PD-1 MFI on CD8⁺ TILs isolated from (F) 4T1 and (H) B16-OVA tumors. The percentage of effector CD8⁺ TILs isolated from 4T1 tumors that co-express both PD-1 and (I) T-bet or (J) IFN-γ. All data are presented as mean ± SEM. Statistics analyzed by Student’s t-test: *** P < 005. B16-OVA studies N = 5 per group. 4T1 studies N = 8–10 per group.

Figure 6. β-blockade increases the efficacy of anti-PD-1 immunotherapy in the 4T1 murine tumor model

(A) Experimental design for combination propranolol and anti-PD-1 studies at 22°C. Mice at 22°C were challenged with 4T1 tumors. Mice were randomized to receive anti-PD-1 or isotype antibody (200μg) with or without daily propranolol treatments (200μg) beginning the day after tumors became detectable. Mice received 5 doses of anti-PD-1 or isotype antibody on days 0, 3, 6, 9, and 12. (B) 4T1 tumor growth. Data are presented as mean ± SEM. Tumor growth statistics analyzed using two-way ANOVA with Tukey analysis. **** P < 0.0001. N = 6–7 per group.

Figure 7. Anti-PD-1 and β-blocker therapy synergize to increase the frequency of IFN-γ expressing intra-tumoral effector CD8⁺ T cells in 4T1 tumors

Single cell suspensions were made from 4T1 tumors harvested from mice housed at 22°C treated with anti-PD-1 or isotype antibody with or without propranolol. IFN-γ expression was determined in intra-tumoral CD8⁺ T cells using flow cytometry. (A) Representative flowplots of IFN-γ expression in CD8⁺ TILs. (B) Quantification of IFN-γ expressing CD8⁺ TILs in all treatment groups. (C) The number of IFN-γ⁺ CD8⁺ T cells per mg of tumor. All data are presented as mean ± SEM. Log-transformed data analyzed using one-way ANOVA with Tukey adjusted post-hoc tests. In figures 7B and 7C, there was a significant overall association with treatment group (P = 0.002 and P = 0.001, respectively). * P < 0.05, ** P < 0.01. N = 6–7 per group.