β-Adrenergic Signaling Impairs Antitumor CD8⁺ T-cell Responses to B-cell Lymphoma Immunotherapy

Immune-based therapies are revolutionizing the clinical practice of cancer treatment and a growing armamentarium of immune-modulating agents are being developed (1). However, treatment success varies. Improvements in immunotherapy use depend on better understanding of interactions between tumor cells, immune cells, and the microenvironment. Understanding factors that antagonize anticancer immunity should advance our ability to deliver effective immunotherapy, alone or in rational combination with other interventions.

The most widely used immunotherapies are monoclonal antibodies directed at immune checkpoint molecules such as PD-1 and CTLA-4, as well as stimulatory molecules such as 4-1BB, which have collectively treated many patients with various types of cancers (2–4). These therapies were designed to augment antitumor CD8⁺ T-cell responses. Agonistic anti–4-1BB does this by eliciting new cytotoxic CD8⁺ T cells with specificity toward tumor cells (3), whereas anti–PD-1 and anti–CTLA-4 blocking antibodies release existing tumor-specific CD8⁺ T cells from inhibition caused by the tumor microenvironment (2, 4). Other immunotherapies are also being designed to target innate antitumor effector cells like natural killer (NK) and natural killer T (NKT) cells. An example is the use of the glycolipid adjuvant, alpha-galactosylceramide (αGalCer), to stimulate NKT cells and downstream activation of NK cells (5). However, despite recent advances in cancer immunotherapy, effectiveness remains limited in many patients, for reasons that are not well understood.

The sympathetic nervous system (SNS) is a physiological regulator of immune function (6–9). The SNS is activated by psychological stress, inflammation, and physical activity, which leads to the release of catecholamines (e.g., noradrenaline and adrenaline), which activate adrenergic receptors on target cells. The effects of adrenaline and noradrenaline on immune cells are predominantly mediated by β-adrenergic receptors (βAR). Although most immune cells express βAR, reports vary on receptor density, affinity, and relative abundance (10–12). These receptors are critical for translating the effects of SNS stimulation into immune response outcomes (13, 14).

βAR signaling has been shown to alter the magnitude and quality of immune responses. For example, βAR signaling suppresses NK cell function (15–18), and it can also have an enhancing effect of βAR signaling on NK cell function (19, 20). These divergent results may be due to differences between acute and continuous signaling through βAR pathways. βAR activation also leads to reduced function of antigen presenting cells like dendritic cells (DC), including reduced ability to cross-present antigens to CD8⁺ T cells (21), and promotion of an anti-inflammatory phenotype (22, 23). βAR signaling affects T-cell function by modulating cytokine production and reducing T-cell proliferation in response to various stimuli (24–27), and inducting regulatory T cells (28). This raises the possibility that activation of βAR signaling may impair the activity of antitumor T cells in response to immunotherapy.

In addition to regulating immune cell function, βAR signaling affects progression of both solid and hematological malignancies. Studies in mouse models demonstrated that βAR signaling in solid tumors enhances tumor cell invasion and remodels the tumor microenvironment to enhance metastasis (29–32). Increased βAR signaling has also been shown to accelerate the progression of leukemia (17, 33), although the effect on lymphoma is yet to be investigated.

The objective of the current study was to investigate the role of βAR signaling on progression of B-cell lymphoma, and on innate and adaptive immune responses to immunotherapy. βAR signaling was modulated using a βAR agonist and the effects on immune activation and immunotherapeutic responses were monitored. Using immunotherapies that harness either innate immunity (NK and NKT cells) or adaptive immunity (CD8⁺ T cells), we demonstrated a significant adverse effect of elevated βAR signaling on CD8⁺ T-cell function and negative impact on CD8⁺ T cell–targeted immunotherapy.

C57BL/6J and B6.SJL.ptprca mice were purchased from the Animal Resource Centre (Perth, Australia). OT-1xSJL mice and mice deficient for both β1AR and β2AR (Adrb1^tm1BkkAdrb2^tm1Bkk/J), referred to as β₁β₂AR-KO, were sourced from The Jackson Laboratory and bred on site. All mice were housed under standard conditions with free access to food and water, on a 12 hour light/dark cycle. Eμ-myc mouse B-cell lymphoma cells were originally provided by Ricky Johnstone at the Peter MacCallum Cancer Centre (Melbourne, Australia). All experiments used the “4242” tumor clone derived from an Eμ-myc transgenic mouse. Checkpoint immunotherapy experiments used Eμ-myc 4242 cells that had been retrovirally transduced with the model antigen ovalbumin (Eμ-myc-OVA), permitting investigation of anti-OVA T-cell responses. Eμ-myc cells were inoculated into mice intravenously at a dose of 10⁵ or 10⁴ cells per recipient as denoted in individual experiments. All experiments were performed in accordance with the animal ethics guidelines provided by the National Health and Medical Research Council of Australia and approved by the University of Queensland Health Sciences Animal Ethics Committee (approvals 090-14 and 113-17).

Isoprenaline hydrochloride (Sigma-Aldrich) was administered by daily subcutaneous injection in PBS at 10 mg/kg. Therapeutic anti–4-1BB (clone 3H3; 100 μg) and anti–PD-1 (clone RPM1-14; 200 μg), or Rat IgG2a (clone 2A3; 200 μg) isotype control antibody (all from Bio X Cell) were given to mice by intraperitoneal injections in PBS.

Eμ-myc tumor cell vaccine immunotherapy (34) was prepared by culturing Eμ-myc cells in complete RPMI containing 500 ng/mL alpha-galactosylceramide (KRN7000, Avanti Polar Lipids) overnight. Cells were then irradiated with 50 Gy of gamma radiation in order to arrest tumor cell proliferation, and then washed thoroughly to remove free-form αGalCer before intravenous injection into recipients at 5 × 10⁵ cells per mouse. Ovalbumin immunizations to study CD8⁺ T-cell responses were carried out by subcutaneous injection of a mixture of 50 μg OVA protein (PeproTech) and 20 μg QuilA adjuvant (InvivoGen) diluted in PBS vehicle.

Blood samples were collected into heparinized Eppendorf tubes from anaesthetized mice by submandibular puncture. Red blood cells were removed using a RBC lysis buffer (ACK buffer; 150 mmol/L NH4Cl, 10 mmol/L KHCO3, 0.1 mmol/L EDTA). The remaining white blood cells were stained with various antibodies in staining buffer containing 2% FCS and 2 mmol/L EDTA. The following fluorescently conjugated anti-mouse monoclonal antibody clones were used for flow cytometry analysis; β2AR (polyclonal IgG), CD19 (1D3 or 6D5), NK1.1 (PK136), CD3e (145-2C11), CD4 (RM4-5), CD8a (53-6.7), CD11b (M1/70), CD69 (H1.2F3), Gr-1 (RB6-8C5), Ly6G (1A8), Ly6C (HK1.4), CD8b (YTS156.7.7), CD45.1 (A20), TCR-β (H57-597), CD44 (IM7), KLRG1 (2F1/KLRG1), CD27 (LG.3A10), CD127 (A7R34), CD62L (MEL-14), IFNγ (XMG1.2), CD11c (N418), MHC-II (M5/114.15.2), IL12p40 (C15.6), and CD86 (GL-1). For intracellular cytokine staining, cells were stimulated with 1 μmol/L peptide, 25 ng/mL PMA + 1 μg/mL ionomycin, or 100 ng/mL LPS, as indicated, in the presence of Brefeldin A for 4 hours, then fixed and permeabilized using the BD Cytofix/Perm kit (Becton Dickinson) as per the manufacturer's instructions. After antibody staining, cells were acquired on a Gallios flow cytometer (Beckman Coulter) and populations were analyzed using FlowJo v10 software (FlowJo LLC). Cell populations were enumerated using Fluorosphere Flow Count Beads (Beckman Coulter) using the following formula; number of cells in sample = (number of cells analyzed/number of beads analyzed) × number of beads in sample.

CD45.1 congenic B6.SJL recipient mice were irradiated with 11 Gy of gamma radiation and injected 24 hours later with 2 × 10⁵ bone marrow cells derived from either β₁β₂AR-KO mice or wild-type (WT) C57BL/6 mice. T cells were depleted from the bone marrow by immunomagnetic selection in order to avoid GVHD due to the mixed genetic background of β₁β₂AR-KO mice. For consistency, T cells were also depleted from C57BL/6 bone marrow before transfer. Mice were maintained on neomycin water (1 mg/mL) for 1 week before and 2 weeks after irradiation and BM transplant to prevent infections. Blood samples were analyzed by flow cytometry to confirm full immune reconstitution every 2 weeks for 8 weeks, at which time mice were used in experiments.

To assess the cytolytic capacity of anti-OVA CD8⁺ T cells in vivo, CD45.1⁺ donor splenocytes were pulsed with K_b-restricted peptide OVA_257-264 (SIINFEKL; Peprotech) at concentrations ranging from 10⁻⁶ to 10⁻¹¹ mol/L for 2 hours in complete RPMI media. Peptide-pulsed targets were coinjected with unpulsed cells as controls at a 1:1 ratio for all populations. All donor cells were distinguished from host cells by CD45.1 expression and labeled with either Celltrace Violet (20 μmol/L, 1 μmol/L or 0.05 μmol/L) or CFSE (20 μmol/L, 1 μmol/L or 0.05 μmol/L; Life Technologies) to distinguish individual target populations by flow cytometry. The survival of target cells relative to control cells was assessed in the blood or spleens of mice 18 hours after injection in order to determine killing efficiency, using the following formula: % specific killing = 100 × (1 − (target cells/control cells)).

Bone marrow was collected from femurs of C57BL/6 or β₁β₂AR-KO mice, depleted of red blood cells using ACK lysis buffer, and cultured at a density of 10⁶ cells per mL in complete RPMI supplemented with 10 ng/mL recombinant mouse GM-CSF (PeproTech). After 7 days, nonadherent cells were removed and the adherent cells (>80% CD11c⁺ MHC-II⁺ dendritic cells) were collected. BMDCs were matured using 100 ng/mL LPS from E. coli (InvivoGen), and exposed to endotoxin-free OVA (InvivoGen) where described and cocultured at 10⁵ cells/mL with purified CD8⁺ T cells for 3 days. CD8⁺ T cells were isolated using an EasySep CD8⁺ T-cell–negative selection kit as per the manufacturer's instructions (Stemcell Technologies).

Spleens from WT C57BL/6 or OT-I mice were harvested and processed into single-cell suspensions by mechanical dissociation followed by RBC lysis. CD8⁺ T cells were isolated as described above. Purified CD8⁺ T cells (>95% TCRβ⁺ CD8⁺) were washed and labeled with 20 μmol/L Celltrace Violet before culture with either BMDCs or plate-bound CD3 mAb (145-2C111, 1 μg/mL) with soluble CD28 mAb (37.51, 1–1,000 ng/mL) or 4-1BB (3H3, 1–1,000 ng/mL), as indicated. Proliferation (CTV dilution) and activation of CD8⁺ T cells (intracellular IFNγ) was assessed 3 days later by flow cytometry. Proliferation index represents the average number of cells arising from one parental cell during the culture and was calculated using the following formula: proliferation index = total cells/sum of (cells in each generation/2⁁generation).

All statistical analyses were performed using Graphpad Prism v6 and P < 0.05 was selected as the cutoff for statistical significance in all comparisons. Univariate analysis was carried out using either Student t test for data that were normally distributed (passed a D'Agnostino-Pearson omnibus normality test) or Mann–Whitney test for those which were not normally distributed. Grouped data were analyzed by two-way ANOVA using a Tukey post hoc test to correct for multiple comparisons, or by repeated measures two-way ANOVA with a Sidak post hoc test to correct for multiple comparisons in time-course analyses, as appropriate. Dose–response data were fitted by unconstrained three-parameter logistic curves (log-dose vs. response) using a least-squares method and differences between groups were determined by comparison of the best-fit values of the individual regressions. Survival analyses were carried out by the log-rank test.

To investigate the effects of βAR signaling on the growth of B-cell lymphoma, mice were administered the nonselective βAR agonist isoprenaline daily for 21 days, commencing 7 days prior to Eμ-myc tumor inoculation (Fig. 1A). βAR signaling accelerated growth of lymphoma, as demonstrated by elevated GFP⁺ tumor cells in circulation compared with vehicle-treated mice (Fig. 1B) and reduced survival of lymphoma-bearing mice (Fig. 1C). To investigate the cellular targets of βAR signaling, we used flow cytometry to quantify surface expression of β2AR, and confirmed high expression on lymphoma cells, T-cell subsets and B cells (Fig. 1D), consistent with previous observations that immune cells preferentially express β2AR (10–12). The presence of Eμ-myc lymphoma increased the number of circulating lymphoid and myeloid cell populations, except CD4⁺ T cells (Fig. 1E), a phenomenon reported previously (35). Chronic βAR signaling with isoprenaline treatment decreased peripheral B cells and increased granulocytes in healthy mice, supporting previous findings (13). In contrast, there was little effect of isoprenaline on circulating NK cell or T-cell numbers. Corresponding effects on lymphocyte numbers were observed in the spleens of chronic isoprenaline treated mice. In lymphoma-bearing mice, increased βAR signaling did not alter the already elevated number of B cells, T cells, monocytes, or NK cells; however, it further amplified lymphoma-induced granulocytosis (Fig. 1E). To determine if the lymphoma-promoting effect of βAR signaling was immune mediated, bone marrow chimera mice were created using donor hematopoietic cells derived from either WT or β₁β₂AR-KO mice to establish a setting in which the hematopoietic compartment can directly respond (WT) or not respond (KO) to βAR stimulation (Fig. 1F). βAR activation accelerated the growth of lymphoma in mice that received WT bone marrow (Fig. 1G), but failed to accelerate lymphoma progression in mice receiving β₁β₂AR-KO bone marrow (Fig. 1H). This demonstrates that βAR signaling in immune cells or other BM-derived cells is required for accelerated growth of lymphoma and raises the possibility that chronic βAR signaling suppresses antitumor immunity to drive lymphoma.

To investigate the impact of βAR signaling on innate lymphocyte function, mice treated with isoprenaline or vehicle-treated controls were immunized with a tumor cell–based vaccine comprised of irradiated Eμ-myc cells pulsed with the glycolipid adjuvant alpha-galactosylceramide (αGalCer; ref. 34). Forty-eight hours after vaccination, mice were inoculated with Eμ-myc lymphoma, and the progression of disease was monitored by expansion of GFP⁺ cells in the blood and by survival (Fig. 2A). Administration of αGalCer vaccine led to a significant suppression in disease development in both vehicle-treated and isoprenaline-treated mice, with no tumor dissemination in blood detectable for 2 weeks (Fig. 2B). However, once tumor cells were detectable, βAR signaling increased tumor burden compared with vehicle-treated controls. Although βAR activation resulted in decreased survival in both untreated and vaccinated mice, the survival benefit resulting from vaccination occurred regardless of βAR signaling (Fig. 2B). As expected, administration of αGalCer vaccine caused an increase in CD69 expression and IFNγ production by both NKT cells (Fig. 2C) and NK cells (Fig. 2D). βAR activation did not significantly alter CD69 expression by NK cells, but did impair vaccine-induced IFNγ production by NK cells (Fig. 2D). In contrast, βAR activation enhanced vaccine-induced CD69 expression by NKT cells (Fig. 2C) and vaccine-induced IFNγ production (Fig. 2C). To address the net outcome of the opposing effects of βAR signaling on NK versus NKT cell IFNγ production, we measured total systemic amounts of IFNγ in serum of vaccinated mice. The activation of βAR did not significantly change the total amount of IFNγ induced by vaccination (Fig. 2E). This finding supports the observation that βAR signaling does not adversely affect vaccine-mediated tumor growth suppression in the initial 2-week period (Fig. 2B), despite different effects on innate effector cell populations. However, subsequent accelerated relapse in isoprenaline-treated mice indicated a disruption to adaptive immune cell control.

The effect of βAR signaling on the CD8⁺ T-cell response to the vaccine was next investigated. OVA-specific (OT-I) CD8⁺ T cells labeled with Celltrace Violet proliferation dye were injected into mice that received daily injections of isoprenaline or vehicle. Twenty-four hours after OT-I transfer, mice were vaccinated with αGalCer pulsed, irradiated Eμ-myc lymphoma cells expressing the OVA antigen (Eμ-myc-OVA). Vaccination caused significant proliferation of OT-I cells by day 5. βAR signaling suppressed the OT-I cell proliferative response, in terms of both initiation and number of divisions (Fig. 2F). These results demonstrate that elevated βAR signaling significantly reduces the proliferative response of antigen-specific CD8⁺ T cells to α-GalCer-based vaccination and may explain the accelerated outgrowth of vaccinated mice exposed to isoprenaline.

Eμ-myc-OVA tumors were utilized to investigate the effects of βAR signaling on the efficacy of immunotherapy that targets 4-1BB costimulatory or PD-1 coinhibitory molecules on CD8⁺ T cells. Mice received adoptive transfer of CD45.1⁺ OT-I CD8⁺ T cells 1 day prior to tumor inoculation to permit assessment of antigen-specific CD8⁺ T cell-mediated treatment responses to immunotherapy. On days 4, 8, 12, and 16 after tumor inoculation, mice were administered agonistic mAb to 4-1BB, a blocking mAb to PD-1, or an isotype control antibody and were treated with isoprenaline or vehicle (Fig. 3A). In contrast to parental Eμ-myc tumor growth (Fig. 1B), isoprenaline did not alter growth kinetics of Eμ-myc-OVA tumors (Fig. 3B). 4-1BB mAb treatment caused significant inhibition of lymphoma growth and improved survival, with 11 out of 15 mice displaying long-term, tumor-free survival. However, in mice treated with isoprenaline, only 4 out of 15 mice responded to anti–4-1BB treatment (Fig. 3C). PD-1 mAb therapy was less effective than 4-1BB mAb at controlling lymphoma, with only 5 out of 13 treated mice achieving long-term survival. βAR signaling also abrogated the effect of anti–PD-1 therapy, reducing the number of responding mice to 2 out of 15 (Fig. 3D). The functional state of tumor-specific CD8⁺ T cells was assessed by IFNγ production and cytolytic activity from OT-I cells 10 days after tumor inoculation. Only 4-1BB mAb treatment caused an increase in IFNγ production (Fig. 3E) and antigen-specific cytotoxicity against OVA peptide–pulsed targets (Fig. 3F). Treatment with isoprenaline significantly reduced IFNγ production by OT-I cells in 4-1BB mAb–treated mice (Fig. 3E). In all treatment settings, βAR signaling caused a significant decrease in CD8⁺ T cell killing efficiency (Fig. 3F). Together, these results demonstrate that elevated βAR signaling leads to a decrease in therapeutic response to two immune-checkpoint immunotherapies, associated with decreased antigen-specific CD8⁺ T-cell function.

βAR signaling negatively impacted CD8⁺ T-cell responses to three immunotherapies (αGalCer vaccine, 4-1BB mAb, and PD-1 mAb). To further elucidate the effects of chronic βAR signaling on antigen-specific CD8⁺ T-cell activity in the absence of additional modulating effects of the tumor environment, a simplified model of in vivo antigen challenge in the absence of tumor was utilized. Mice were treated with isoprenaline or vehicle, received adoptive transfer of CD45.1+ OT-I T cells, and 1 day later were immunized with OVA protein plus QuilA adjuvant (Fig. 4A). Immunization caused rapid expansion of OT-I cells, whereas βAR signaling significantly impaired OT-1 expansion (Fig. 4B). The rate of OT-I expansion and subsequent contraction were similar between isoprenaline- and vehicle-treated groups, suggesting that βAR signaling modulates the magnitude of OT-I expansion, but not the kinetics. Furthermore, OT-I cells isolated from isoprenaline treated mice had reduced capacity for IFNγ cytokine production at all time points assessed (Fig. 4C). Thirty days after immunization, the cytolytic capacity of OVA-specific CD8⁺ T cells was tested against transferred SIINFEKL-pulsed target cells, containing varying peptide concentrations ranging from 10 pmol/L to 1 μmol/L. At all peptide concentrations, βAR signaling significantly reduced CD8⁺ T cell-cytolytic function, particularly at lower peptide concentrations (Fig. 4D). Together, these data demonstrate the suppressive effects of chronic βAR signaling on the generation of an antigen-specific CD8⁺ T-cell response in vivo.

Suppression of antigen presenting cell activity is one possible explanation for reduced activation of antigen-specific CD8⁺ T cells in response to immunization, under conditions of elevated βAR signaling. To determine if DC responses are suppressed by βAR signaling, BMDCs were matured with LPS in the presence or absence of isoprenaline. As expected, stimulation of DCs with LPS caused upregulation of the stimulatory molecule CD86 and the cytokine IL12, which was significantly reduced by βAR signaling (Fig. 5A and B). To determine whether reduced DC activation by isoprenaline led to subsequent reduction in CD8⁺ T-cell priming, WT or β₁β₂AR-KO DCs were matured with LPS and ovalbumin protein (a source of antigen for OT-I cells) in the presence or absence of isoprenaline, and then cocultured with OT-I T cells. OT-I cells responded to DC coculture with significant proliferation and IFNγ production. βAR signaling caused significant decreases in both proliferation (Fig. 5C) and IFNγ production (Fig. 5D) by OT-I cells. The suppressive effects of βAR signaling on OT-I cell proliferation and IFNγ production were also observed in cocultures with DCs that lacked βAR, although the magnitude of the inhibitory effect of isoprenaline was lower in this setting (Fig. 5C and D). These findings suggest that direct βAR signaling in CD8⁺ T cells is sufficient to suppress their response to DC-mediated activation, regardless of whether the DCs are responsive to βAR signaling.

Eμ-myc tumor-bearing mice responded poorly to agonistic 4-1BB mAb therapy if concurrently treated with isoprenaline (Fig. 3C), a phenomenon associated with impaired CD8⁺ T-cell responses (Fig. 3E and F). To investigate the direct effects of βAR signaling on the CD8⁺ T-cell response to 4-1BB stimulation, purified CD8⁺ T cells were stimulated in vitro with agonistic anti-CD3 and 4-1BB mAbs in the presence or absence of isoprenaline. The addition of 4-1BB mAb to CD8⁺ T cells caused significant proliferation (Fig. 6A) and IFNγ production (Fig. 6B), in a dose-dependent manner. βAR signaling reduced CD8⁺ T-cell proliferation in response to 4-1BB mAb (Fig. 6A), particularly at higher antibody concentrations. In contrast, βAR signaling inhibited IFNγ production at lower 4-1BB mAb concentrations (Fig. 6B). A similar inhibitory effect of isoprenaline on CD8⁺ T-cell stimulation was observed when CD28 mAb was substituted for 4-1BB mAb. Proliferation was reduced at all CD28 mAb concentrations tested (Fig. 6A). The inhibitory effect on IFNγ production was more pronounced at higher CD28 mAb concentrations (Fig. 6B). Costimulation with either anti–4-1BB or anti-CD28 supported the viability of CD8⁺ T cells in culture; however, treatment with isoprenaline led to a decrease in overall CD8⁺ T-cell viability regardless of stimulation (Fig. 6C).

In order to validate the effects of βAR signaling on CD8⁺ T cell stimulation in vivo, healthy mice receiving isoprenaline or vehicle treatment were given OT-I T cells and 24 hours later were immunized with OVA protein and 4-1BB mAb. Additional 4-1BB mAb was administered 4, 8, and 12 days after initial immunization in order to mimic the schedule of immunotherapy used previously in tumor-bearing hosts (Fig. 3A). Immunization with 4-1BB mAb resulted in OT-I cell expansion (Fig. 6D) and IFNγ production (Fig. 6E), and βAR signaling significantly impaired both of these responses. Together, these results demonstrate the suppressive effects of βAR signaling on CD8⁺ T cell costimulation, and provide mechanistic insight into the reduced efficacy of agonistic 4-1BB mAb therapy observed with βAR activation in Eμ-myc lymphoma.

Immune suppression is a major impediment to successful outcomes with immunotherapy for cancer. Here, we identify βAR signaling as a physiological regulator of immunosuppression in B-cell lymphoma and define its role in lymphoma progression and response to immunotherapy. βAR stimulation is activated by neural signaling, which has largely been overlooked as a component of the tumor microenvironment, despite known bidirectional interactions between the immune and neural systems. We found that chronic βAR signaling in mice, afforded by administration of a nonspecific β-agonist isoprenaline, enhanced growth of transplanted Eμ-myc B lymphoma in an immune-dependent manner. βAR signaling inhibited CD8⁺ T-cell function and suppressed antitumor CD8⁺ T-cell activity in response to clinically utilized immunotherapeutic treatments. These findings highlight the importance of neural signaling in the tumor microenvironment, for both progression of hematological cancers and their response to immunotherapy.

The current findings reveal the inhibitory effect of chronic βAR signaling on antitumor adaptive immune effector cells. Antigen-specific CD8⁺ T cells proliferated less and were less capable of producing proinflammatory cytokines and killing target cells, in response to cognate antigen stimulation. There are various means by which CD8⁺ T-cell function could be directly or indirectly affected by βAR signaling; these are not mutually exclusive and could reasonably occur simultaneously in vivo. Two major avenues were explored including inhibition of activation by direct βAR signaling in T cells, and indirect suppression via effects of βAR signaling on antigen presenting cells. Consistent with findings that βAR signaling impairs cross-presentation by DCs (21), we observed reduced maturation of DCs in response to TLR4 stimulation in vitro, which subsequently reduced activation of antigen-specific CD8⁺ T cells. However, in cocultures with βAR receptor knockout DCs that do not respond to βAR agonism, the activation of CD8⁺ T cells was similarly suppressed, indicating that direct βAR signaling in CD8⁺ T cells was sufficient to inhibit CD8⁺ T-cell activation. We found that βAR signaling reduced the proliferative and cytokine response to CD28 and 4-1BB costimulatory signals, demonstrating that one mechanism by which βAR signaling directly suppresses CD8⁺ T-cell function is via inhibition of costimulation. CD8⁺ T-cell stimulation through CD28 or alternative receptors such as 4-1BB is essential for T-cell activation and prevention of induced tolerance (reviewed in ref. 36). This extends previous findings that neural signaling impairs CD28 costimulation (24, 25), by also showing effects with 4-1BB and antigen-specific TCR signaling in vivo.

Given the broad roles for DCs and their priming of CD8⁺ T cells in anticancer immunity and immunotherapy, suppression of these cells by βAR signaling has similarly broad clinical implications. The consequence of suppressed CD8⁺ T-cell function during elevated βAR signaling was evident when treating Eμ-myc tumor-bearing mice with T cell-targeting immunotherapies. The efficacy of two common immune modulatory antibodies, anti–4-1BB and anti–PD-1, was dramatically reduced with βAR signaling. We have previously shown that therapeutic efficacy of immune checkpoint therapy is dependent on CD8⁺ T-cell activity (37). Elevated βAR signaling was associated with reduced IFNγ production and cytolytic killing by CD8⁺ T cells in response to anti–4-1BB and anti–PD-1 antibody treatments, demonstrating possible mechanisms for reduced therapeutic efficacy. Anti–PD-1 antibodies are approved for use in a number of cancer settings and 4-1BB targeting is currently being tested in clinical trials on several cancer types, including melanoma, non–small cell lung cancer, leukemia, and lymphoma (38–41). This raises the possibility that neural signaling might affect treatment response in diverse patient populations.

In contrast to T cell–modulating monoclonal antibody therapy, the efficacy of a NKT cell-targeting vaccine was largely maintained under chronic βAR signaling, despite opposing effects of isoprenaline on NKT cells and NK cell function. This suggested that the net outcome of innate immune protection elicited by this vaccine (34) was not adversely affected. However, βAR signaling suppressed CD8⁺ T-cell activation from this vaccine, which may explain the subsequent increase in tumor growth rate following the initial period of innate effector cell–mediated control. Thus, βAR stimulation has different effects on various lymphoid cell populations. One potential explanation is variances in surface βAR expression. Flow cytometry analysis of β₂AR reveal that CD8⁺ T cells express high levels of the receptor, whereas NK cell express little by comparison. It is likely that a combination of direct βAR signaling and indirect effects of systemic βAR activation results in functional alterations of individual cell populations in vivo. However, we were not able to determine the contribution of direct versus indirect signaling events in this study.

These results suggest that in situations where systemic βAR signaling is elevated—such as chronic inflammation, psychological stress, or treatment with β-agonists—the function of CD8⁺ T cells, and therefore the response to many immunotherapies, may be much reduced. These findings suggest a rationale for combining immunotherapies with βAR antagonists to block the adverse effect of βAR signaling on immune function and immunotherapy. Given the widespread clinical use of inexpensive and well-tolerated β-blocker drugs and their recent evaluation in cancer patients (42), we propose that the combination of these drugs with immunotherapies provides a promising translational avenue to increase the efficacy of immunotherapy and outcomes for cancer patients. Current efforts in our lab are focused on demonstrating this effect in a range of preclinical models of cancer and cancer immunotherapy.

E.K. Sloan is a consultant/advisory board member for Cygnal Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M.D. Nissen, E.K. Sloan, S.R. Mattarollo

Development of methodology: M.D. Nissen, E.K. Sloan, S.R. Mattarollo

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.D. Nissen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.D. Nissen, E.K. Sloan, S.R. Mattarollo

Writing, review, and/or revision of the manuscript: M.D. Nissen, E.K. Sloan, S.R. Mattarollo

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M.D. Nissen

Study supervision: E.K. Sloan, S.R. Mattarollo

This work was supported by the NHMRC (APP1044355), the David and Lorelle Skewes Foundation, the Peter Mac Foundation, and the National Cancer Institute (CA160890). M.D. Nissen was supported by an Australian Government Research Training Program (RTP) Scholarship. S.R. Mattarollo was supported by an NHMRC Career Development Fellowship (APP1061429).

The authors thank Ben Harvie, Kendall Hepple, Kamil Sokolowski, Alistair Zealey, Lauren Windt, Karen Knox, and Rodrick Rupan (Biological Resources Facility) for mouse husbandry and technical assistance, David Sester, Dalia Khalil, and Yitian Ding (Flow Cytometry Facility) for technical assistance, and Ricky Johnstone (Peter MacCallum Cancer Centre) for originally providing the Em-myc 4242 tumor cells.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.