The Sympathetic Nervous System Modulates Cancer Vaccine Activity Through Monocyte-Derived Cells

Abstract

The sympathetic nervous system (SNS) is an important regulator of immune cell function during homeostasis and states of inflammation. Recently, the SNS has been found to bolster tumor growth and impair the development of anti-tumor immunity. However, it is unclear whether the SNS can modulate APC function. Here, we investigated the effects of SNS signaling in murine monocyte-derived macrophages (moMФ) and dendritic cells (DCs) and further combined the nonspecific β-blocker propranolol with a peptide cancer vaccine for the treatment of melanoma in mice. We report the novel finding that norepinephrine treatment dramatically altered moMФ cytokine production, whereas DCs were unresponsive to norepinephrine and critically lack β₂-adrenergic receptor expression. In addition, we show that propranolol plus cancer vaccine enhanced peripheral DC maturation, increased the intratumor proportion of effector CD8⁺ T cells, and decreased the presence of intratumor PD-L1⁺ myeloid-derived suppressor cells. Furthermore, this combination dramatically reduced tumor growth compared to vaccination alone. Taken together, these results offer novel insights into the cell-specific manner by which the SNS regulates the APC immune compartment and provide strong support for the use of propranolol in combination with cancer vaccines to improve patient response rates and survival.

In Brief

• β₂-adrenergic signaling impacts monocyte-derived cells rather than dendritic cells.

• Propranolol improves cancer vaccine treatment of melanoma in mice.

Introduction

Cancer patients experience increased sympathetic nervous system (SNS) tone from both overt autonomic nerve growth at the solid tumor as well as from the emotional stress that accompanies cancer diagnosis.[1] High autonomic nerve density in tumors is associated with worse prognosis and more aggressive characteristics in human prostate, breast, colon, and pancreatic cancers.[2–4] Additionally, inhibition of SNS signaling through β-blockers has been associated with improved prognosis in several types of cancers in human patients.[5–7] The SNS and its primary neurotransmitter norepinephrine stimulate the ubiquitously expressed adrenergic receptors to affect a wide range of biological functions, and SNS regulation of immune cells directly contributes to cancer progression and the effectiveness of cancer immunotherapy. Increased SNS signaling has been associated with decreased tumor infiltrating lymphocytes in several mouse models, and stimulation of β₂-adrenergic receptors (β₂ARs) on tumor-associated macrophages (TAMs) and myeloid-derived suppressor cells (MDSCs) increases their frequencies in tumors and enhances their pro-tumorigenic functions.[8–11] These preclinical findings indicate that combining β-blockers with immunotherapies could subvert tumor-mediated immune suppression to maximize patient immune responses. However, much remains to be explored in the relationship between SNS signaling and tumor immunity.

The SNS is an important regulator of immune cell function both during states of inflammation and under homeostasis.[12] Postganglionic fibers of sympathetic nerves hold a central role in regulation of bone marrow hematopoietic stem cell niches compared to parasympathetic and sensory nerves.[12] Sympathetic nerves additionally innervate secondary lymph organs to regulate circadian patterns of leukocyte recruitment to tissues.[13] Moreover, sympathetic nerves that innervate lymph nodes will release catecholamines during states of inflammation to regulate innate and adaptive immune cell function largely through β₂AR signaling.[12, 14, 15] The β₂AR is a G protein-coupled receptor (GPCR) that induces G protein activation, increased cAMP, protein kinase A activation, and phosphorylation of CREB through one of its major signaling pathways when stimulated.[16] While there has been much study into the immunosuppressive effects of β₂AR signaling on cytokine production in innate immune cells such as macrophages and monocytes,[17, 18] the extent to which catecholamines regulate APC function remains unclear. Conventional dendritic cells (cDCs) and monocyte-derived cells (MCs) are APCs that contribute to T cell priming through their influence on the local inflammatory milieu and antigen presentation, and they exhibit nonoverlapping function depending on the inflammatory trigger.[19–21] During inflammatory events, lymph node resident cDCs, migratory cDCs, and inflammatory MCs will migrate together deep into the T cell-rich paracortex region of lymph nodes to modulate T cell responses.[19, 20, 22] Catecholamine-mediated suppression of APC function could undermine cancer vaccine strategies if it sufficiently affects APC priming of cytotoxic T cells. For these reasons, it is critical to understand how sympathetic signaling regulates these APC subsets in order to overcome it as a potential barrier to cancer vaccine efficacy.

Bone marrow-derived DCs (BMDCs) are commonly used to study DC biology and were originally assumed to represent a homogenous DC population.[23, 24] However, recent transcriptomic and functional analyses have revealed that CD11c⁺MHCII⁺ BMDCs are actually a group of heterogeneous cells arising from common DC precursors and common monocyte precursors that remain distinct cell types after stimulation with LPS.[25, 26] The spontaneously mature CD11c⁺MHCII^hi BMDCs that arise from common DC precursors are functionally and transcriptionally similar to migratory DCs, whereas the semi mature CD11c⁺MHCII^int BMDCs that arise from common monocyte precursors most closely resemble macrophages.[25] Thus, BMDC cultures include a mixture of DCs and monocyte-derived macrophages (moMФ), which makes it an ideal model to study SNS regulation of APC subsets that are present in lymph nodes during states of inflammation and immunization.

Through evaluation of β₂AR agonist-treated BMDC cultures, it has been previously reported that β₂AR signaling impairs DC priming of T_H1 cells through reduced IL-12p70 and increased IL-10 production.[27, 28] However, these studies did not account for BMDC heterogeneity and the presence of moMФ. In this study, we separated BMDC populations based on their expression of the macrophage-colony stimulating factor receptor CD115, which is exclusively expressed on moMФ, and evaluated DC and moMФ responsiveness to norepinephrine to better understand how the SNS regulates the APC immune compartment. We additionally investigated the benefit of combining the nonspecific β-blocker propranolol with a peptide-based cancer vaccine in mounting anti-tumor immune responses. A murine model of B16 melanoma with OVA expression was chosen for this study since it has previously been shown that βAR signaling is relevant both systemically and intratumorally in mice bearing this tumor type.[29, 30] Tumor-bearing mice were vaccinated with a combination of the OVA peptide SIINFEKL and the adjuvant polyinosinic-polycytidylic acid (Poly IC) complexed with poly-L-lysine (Poly ICLC). Poly IC is a synthetic double-stranded homopolymer that stimulates TLR3, which normally recognizes double-stranded RNA to induce Toll/interleukin-1 receptor domain-containing adaptor protein inducing interferon-β-dependent signaling.[31] When complexed with poly-L-lysine to form Poly ICLC, Poly IC has improved stability in serum and enhanced proinflammatory properties.[32] Poly ICLC is commonly used in vaccine design and has proven to be effective in cancer vaccine clinical trials.[31] In summary, our work offers insight into catecholamine-mediated regulation of DCs and moMФ as well as provides supports for the use of β-blockers to subvert tumor-mediated mechanisms of immune suppression. This study also highlights the need to delineate heterogeneous populations in BMDC cultures when studying DC biology.

Materials and Methods

Cell lines

Murine melanoma cancer cell line B16 that stably expresses OVA was a gift from Dr. Kenneth Rock, Dana-Farber Cancer Institute, Boston. B16-OVA cells were cultured in DMEM supplemented with 10% FBS (GenDEPOT) + 1x 2-mercaptoethanol (BME) + 1:100 penicillin and streptomycin (PS, GenDEPOT). Cells were maintained in a cell incubator at 37°C under 5% CO₂.

Generation of BMDCs and cell-based assays

BMDCs were generated by culturing C57BL/6J mouse bone marrow cells in GM-CSF for 10 days. Briefly, four to eight-week-old C57BL/6J mice were sacrificed and femur and tibia bones were dissected and cleaned. Bone marrow cells were flushed and plated on 100 mm² tissue culture dishes at 2 million cells per dish in 10 mL of complete BMDC media consisting of 1640 RPMI + 10% FBS (GenDEPOT) + 1x BME + 1:100 PS (GenDEPOT) + 20 ng/mL GM-CSF (PreproTech). An additional 10 mL of complete BMDC media were added 3 days later, and 10 mL of media were removed and replaced with fresh complete BMDC media every 2 days until day 10. To separate BMDCs into CD115⁺ and CD115⁻ populations, a CD115 microbead kit was used according to the manufacturer’s protocol (Miltenyi Biotec).

Supernatant from BMDCs were collected 18 hours after treatment and IL-10, IL-12p70, and IFN-α were measured using ELISA cytokine measurement kits from ThermoFisher Scientific. Intracellular cAMP was detected using the cAMP-Glo Max Assay kit (Promega). For western blot, antibodies targeting phosphorylated CREB (Ser133), total CREB, and GAPDH were purchased from Cell Signaling Technologies. Quantitative RT-PCR was performed using SYBR^™ Green Master Mix (Thermo Fisher) and StepOnePlus^™ Real-Time PCR system (GE Healthcare). Forward β₂AR primer: 5’-GGGAACGACAGCGACTTCTT-3’; Reverse β₂AR primer: 5’-GCCAGGACGATAACCGACAT-3’; Forward GAPDH primer: 5’-CATGGCCTTCCGTGTTCCTA-3’; Reverse GAPDH primer: 5’-TACTTGGCAGGTTTCTCCAGG-3’. Data was analyzed using the −2ΔΔct method such that β₂AR mRNA expression was normalized to GAPDH expression in each sample, and then fold change of CD115⁻ DC β₂AR was calculated relative to CD115⁺ moMФ within each biological replicate. For flow cytometry, cells were acquired via BD FACS Fortessa. Gating strategies are provided in figure legends and data were analyzed with FlowJo v10.0 software.

Vaccine preparation

Vaccines were composed of 4 mg/kg Poly ICLC, 100 μg of SIINFEKL peptide, and sterile PBS. Poly ICLC was created by mixing Poly IC (Sigma-Aldrich) with poly-l-lysine (Sigma-Aldrich) in 2% carboxymethylcellulose/sterile PBS as previously described.[32]

In vivo treatment

Four to eight-week-old female C57BL/6J mice were inoculated with 1.5×10⁵ B16-OVA cells in 30% Matrigel matrix (Corning) by subcutaneous flank injection. Treatment groups included untreated, propranolol alone, vaccine alone, and propranolol plus vaccine. Propranolol hydrochloride (Sigma-Aldrich, 10 mg/kg, 100 μL) was administered via intraperitoneal injections. Vaccines were administered via 100 μL footpad injections. Mice in each treatment group were co-housed. Tumors were measured with digital calipers, and tumor volumes were calculated as (Length × width × width)/2. For time of flight mass cytometry (CyTOF), tumors were digested in 1640 RPMI + 1:100 PS + 400 U/mL DNase-1 (Sigma) at 37°C for 45 minutes, and immune cells were enriched using Percoll density gradient centrifugation according to manufacturer. Then, cells were stained and acquired via Helios (Fluidigm). Gating strategies are described in Figure 5 legend, and data were analyzed with FlowJo v10.0 software.

Figure 5. Propranolol favorably alters immune cell populations in size-matched B16-OVA tumors following vaccination:

C57BL/6J mice were subcutaneously inoculated with B16-OVA on days −3, −1, or 0 and were treated according to the schematic presented in figure 6A. Mice were sacrificed when tumors reached approximately 1200 mm³. Tumors were dissected, digested into single cells suspensions, and analyzed by CyTOF. B) Representative t-SNE plots gated on CD45⁺ cells with color-coded cell populations overlaid. C) Fold change of indicated T cell population frequencies among total CD45⁺ cells over untreated control group. Percentage of D) CD8⁺ T cells and E) CD4⁺ T cells that expressed PD-1, CTLA-4, or Foxp3 are shown. F) Fold change of indicated myeloid cell population frequencies among total CD45⁺ cells over untreated control group. Percentage of G) PMN-MDSCs and H) M-MDSCs that expressed PD-L1, NOS2, and ArgI are shown. Gating for each cell type are as follows: CD3⁺CD8⁺ for CD8⁺ T cells; CD3⁺CD8⁺CD44⁺CD62L⁻ for CD8⁺ T_EM; CD3⁺CD8⁺CD44⁺CD62L⁺ for CD8⁺ T_CM; CD3⁺CD4⁺ for CD4⁺ T cells; CD3⁺CD4⁺CD44⁺CD62L⁻ for CD4⁺ T_EM; CD3⁺CD4⁺CD44⁺CD62L⁺ for CD4⁺ T_CM; CD11b⁺Ly6C⁺ for M-MDSCs; CD11b⁺Ly6G⁺ for PMN-MDSCs; CD11b⁺Ly6C⁻F4/80⁺ for TAMs; CD11b⁺Ly6C⁻F4/80⁺CD206⁺ for M2 TAMS; CD11b⁺Ly6C⁻F4/80⁺CD86⁺ for M1 TAMs; CD11C⁺MHCII⁺Ly6C⁻CD64⁻ for DCs; B220⁺ for B cells. n = 5 mice/group, total = 20, and ANOVA followed by Tukey HSD post hoc tests were performed.

T cell proliferation assay

OVA_257–264-specific CD8⁺ T cells (OT-Is) were isolated from transgenic mice using a CD8⁺ T cell negative isolation kit (Stemcell) and stained with CFSE (ThermoFisher). OT-I cells were then cocultured with CD115⁺ or CD115⁻ cells that were previously treated with 100 nM SIINFEKL peptide and 60 μg/mL Poly ICLC with or without 1 μM norepinephrine. OT-Is and BMDCs were cocultured at a ratio of 5:1 for 48 hours at 37°C. OT-I total cell number and CFSE status were evaluated via flow cytometry and cell counting beads mixed into each sample (Thermofisher). Dividing cells were quantified via flow cytometry by gating all OT-I cells with reduced CFSE fluorescent intensity compared to the parent OT-I CFSE fluorescent peak.

Statistical analysis

Data were expressed as mean ± standard deviation. Statistical analysis for experiments with multiple groups was assessed using ANOVA followed by post hoc Tukey multiple comparison tests in GraphPad Prism 8 version 1.2. Two-tailed student T test was used for Figure 2D. No animals were excluded from analysis. Statistically significant differences are denoted by asterisks such that * p < 0.05, ** p < 0.01, and *** p < 0.001.

Figure 2. CD115^± moMФ, but not CD115⁻ DCs in BMDC cultures express functional β₂AR:

A) Flow cytometry analysis of CD11c⁺MHCII⁺ BMDCs with CD115 expression overlaid. CD115⁺ and CD115⁻ cells were separated from BMDCs using MACS microbeads. B) Molecular markers were evaluated by flow cytometry. C) Cell morphology was evaluated by confocal microscopy (representative 60x images shown, scale bars = 30 μm). D) Relative β₂AR mRNA expression via RT-PCR, E) surface β₂AR by flow cytometry, and F) cAMP production G) and CREB phosphorylation following norepinephrine treatment via cAMP bioluminescent assay and western blot, respectively. H) β₂AR on surfaces of macrophages (CD45⁺CD11b⁺CD64⁺Ly6C⁻), moDCs (CD11b⁺CD11c⁺MHCII⁺CD64⁺), and DCs (CD45⁺CD11c⁺MHCII⁺CD64⁻Ly6C⁻) from C57BL/6J mouse popliteal lymph nodes and spleens was evaluated via flow cytometry. n = 3 replicates/group for A-H Representative groups shown for A-C, E, G, and H. ANOVA followed by Tukey HSD post hoc tests were performed for F, and two-tailed student’s T test performed for D.

Study approval

Animals were purchased from Charles River Laboratories and maintained in barrier animal facilities approved by the American Association for Accreditation of Laboratory Animal Care (AAALAC) and in accordance with current regulations and standards of the United States Department of Agriculture, Department of Health and Human Services, and National Institutes of Health. All studies related to animals including housing, tumor inoculation and growth, and treatment were conducted under the approval of Houston Methodist affiliated IACUC.

Results

IL-10 contributes to norepinephrine suppression of BMDC pro-inflammatory cytokine production and maturation

Similar to previous studies,[27, 28] we found that norepinephrine pretreatment of BMDCs resulted in large, dose-dependent increases in IL-10 following Poly ICLC treatment as well as concomitant decreases in IL-12p70 and IFN-α (Figure 1a, b, c). IL-10 is known to independently control DC inflammation responses in skin, lung, and gut tissues as well as in in vitro cultures.[33–36] Additionally, IL-10 is known to directly regulate transcription of the IL-12p40 subunit,[37] so we hypothesized that the increase in IL-10 following norepinephrine treatment contributes to reduced BMDC inflammatory responses. To test this hypothesis, we treated BMDCs with norepinephrine, TLR agonists, and increasing doses of a neutralizing IL-10 antibody. Depletion of active IL-10 abolished norepinephrine reduction of IL-12p70 and IFN-α secretion following Poly ICLC treatment (Figure 1b, c). Inclusion of the neutralizing IL-10 antibody also completely relieved norepinephrine suppression of IL-12p70 after treatment with the TLR agonists LPS and CpG (Figure 1d). Additionally, we found that inclusion of the neutralizing IL-10 antibody allowed BMDCs to maximally express CD86 and CD80 costimulatory molecules after Poly ICLC treatment even in the presence of norepinephrine (Figure 1e, f). Taken together, these results reveal that IL-10 greatly contributes to norepinephrine control of BMDC function.

Figure 1. IL-10 contributes to norepinephrine suppression of BMDC pro-inflammatory cytokine production and maturation:

BMDCs were pretreated with 1 μM (or indicated concentration) norepinephrine (NE) for 1 hour, then treated with 60 μg/mL poly ICLC, 100 ng/mL LPS, 500 ng/mL CpG, and/or 2.5 μg/mL (or indicated μg/mL concentration) neutralizing IL-10 antibody (αIL-10 ab) and culture supernatant were evaluated for A) IL-10, B&D) IL-12p70, and C) IFN-α 18 hours after treatment via ELISAs. E&F) BMDC expression of CD80 and CD86 costimulatory surface markers on CD11c⁺MHCII⁺ BMDCs were evaluated by flow cytometry 18 hours after the indicated treatments. n = 3 replicates/group for A-F, and ANOVA followed by Tukey HSD post hoc tests were performed.

CD115⁺ moMФ in BMDC cultures, but not CD115⁻ DCs express functional β₂AR

Since IL-10 contributes to norepinephrine control of BMDC inflammation, and β₂AR signaling is known to promote IL-10 production from macrophages and monocytes,[18, 38] we sought to delineate the individual cytokine contributions from DCs and moMФ in BMDC cultures following norepinephrine treatment. Consistent with previous reports describing heterogeneity in BMDC cultures,[25] we found that CD11c⁺MHCII^int BMDCs express CD115, whereas CD11c⁺MHCII^hi BMDCs lack CD115 (Figure 2a). Using a CD115 magnetic microbead kit, we separated BMDCs into CD115⁻ and CD115⁺ populations that correspond with DC and moMФ, respectively. After separation, CD115⁻ DCs were loosely adhered, round cells that commonly presented fine dendritic protrusions that are characteristic of DCs (Figure 2c). On the other hand, CD115⁺ moMФ firmly attached to their surfaces and exhibited wide leading-edge lamellar protrusions that are similar to macrophage morphology (Figure 2c). CD115⁻ DCs also expressed higher levels of the DC specific molecules CD135, PD-L2, Zbtb64, and CD117 than CD115⁺ moMФ (Figure 2b) and were stronger inducers of OT-I T cell proliferation than CD115⁺ moMФ (Figure 3f, g). Based on these differences in cell morphology, DC-specific molecule expression, and T cell stimulation, magnetic bead separation of BMDCs based on CD115 expression is an effective method to isolate authentic DCs and moMФ. To evaluate whether norepinephrine differentially impacts CD115⁻ DCs and CD115⁺ moMФ, we examined β₂AR expression and β₂AR downstream signaling after incubation with norepinephrine. We discovered that CD115⁺ moMФ express 5-fold higher amounts of β₂AR mRNA than CD115⁻ DCs (Figure 2d). Flow cytometry analysis revealed that CD115⁺ moMФ express β₂AR on their surface, whereas β₂AR is undetectable on CD115⁻ DCs (Figure 2e). To test β₂AR signal transduction, we examined cAMP production and CREB phosphorylation following norepinephrine treatment. After 30 minutes of norepinephrine treatment, CD115⁺ moMФ produced approximately 19 nM cAMP, whereas CD115⁻ DCs did not produce detectable amounts of cAMP after norepinephrine treatment (Figure 2f). Additionally, norepinephrine treatment resulted in much greater CREB phosphorylation in CD115⁺ moMФ compared to CD115⁻ DCs (Figure 2g). To determine if mouse DCs lack β₂AR similar to our in vitro observations, we evaluated β₂AR expression on the surfaces of DCs, monocyte-derived DCs (moDCs), and macrophages from C57BL/6J mouse popliteal lymph nodes and spleens via flow cytometry. Indeed, we found that macrophages and moDCs exhibited a distinct fluorescence shift after incubation with a β₂AR antibody, whereas β₂AR was undetectable on CD8⁺ and CD103⁺ cDCs, and pDCs and CD11b⁺ DCs exhibited minimal expression of β₂AR (Figure 2h). In summary, these results show that CD115⁺ moMФ and CD115⁻ DCs in BMDC cultures have different expression patterns of the β₂AR and that these patterns of β₂AR expression correlate with in vivo populations of DCs, macrophages, and moDCs.

Figure 3. Norepinephrine alters cytokine secretion from CD115⁺ moMФ rather than CD115⁻ DCs:

CD115⁺ moMФ and CD115⁻ DC were treated with 1 μM norepinephrine (NE), 60 μg/mL Poly ICLC, 100 ng/mL LPS, and/or 500 ng/mL CpG. 18 hours after the indicated treatments, A&D) IL-10, B&E) IL-12p70, and C) IFN-α from cell supernatant were evaluated via ELISAs. F) Representative histograms of OT-I CFSE fluorescence after 48-hour cocultures. G) Total live OT-I cell number after 48-hour coculture normalized to OT-I + CD115⁻ DC without norepinephrine group. H) Percent dividing OT-I cells. n = 3 replicates/group for A-G, and ANOVA followed by Tukey HSD post hoc tests were performed.

Norepinephrine alters cytokine secretion from CD115⁺ moMФ rather than from CD115⁻ DCs

Following our discovery that CD115⁺ moMФ, but not CD115⁻ DCs, express functional β₂AR, we sought to delineate the individual contributions of CD115⁺ moMФ and CD115⁻ DCs to cytokine production following norepinephrine treatment. After treatment with norepinephrine and Poly ICLC, CD115⁺ moMФ experienced an approximately 3.5-fold increase in IL-10 production, whereas CD115⁻ DCs did not significantly increase IL-10 following norepinephrine treatment (Figure 3a). Norepinephrine also led to dramatically reduced IL-12p70 and IFN-α production from CD115⁺ moMФ following Poly ICLC treatment but did not reduce IL-12p70 and IFN-α production from CD115⁻ DCs (Figure 3b, c). This trend persisted after treatment with CpG and LPS. Only CD115⁺ moMФ experienced significantly increased IL-10 and significantly decreased IL-12p70 after treatment with norepinephrine and either CpG or LPS (Figure 3d, e). Additionally, treatment of CD115⁻ DCs with norepinephrine did not affect their ability to stimulate OT-I T cell proliferation (Figure 3f–h). Norepinephrine treatment did not alter CD115⁺ moMФ induction of OT-I proliferation, but we did observe reduced OT-I number after coculture with norepinephrine treated CD115⁺ moMФ, which suggests some undetermined effect on OT-I viability (Figure 3f–h). These results demonstrate that CD115⁺ moMФ are responsible for the changes to cytokine production in BMDC cultures and exhibit altered T cell interaction following norepinephrine treatment, whereas CD115⁻ DCs in BMDC cultures are unaffected by norepinephrine.

Propranolol enhances lymph node DC maturation following vaccination

Despite our finding that DCs do not express surface β₂AR in vitro or in vivo, it has been reported that sympathetic signaling can diminish DC function in vivo.[39, 40] For this reason, we aimed to examine if inclusion of a β-blocker can improve DC responses to vaccination. We performed a therapeutic vaccine experiment according to the schematic in Figure 4a. Mice were sacrificed on day 15, and lymph node, tumor, and spleen tissues were processed for flow cytometry analysis of DC surface costimulatory markers. Combination of propranolol and cancer vaccine resulted in increased proportions of DCs that expressed CD80 and CD86 costimulatory markers two days after the final vaccination in both popliteal and inguinal lymph nodes compared to untreated controls and monotherapy controls (Figure 4b–e). The combination treatment only minimally benefited DC maturation status in spleens, and no changes to maturation status were observed for DCs in tumors (Supplementary Figure 1). These results demonstrate that combination of propranolol and a cancer vaccine enhanced DC maturation at lymph nodes proximal to the vaccine injection site.

Figure 4. Propranolol enhances DC maturation following vaccination:

A) C57BL/6J mice were subcutaneously inoculated with B16-OVA on day 0, given daily IP propranolol injections, and given foot pad injections of the poly ICLC/SIINFEKL vaccine according to the schematic. Mice were sacrificed on day 15. Representative flow cytometry histograms of CD80 and CD86 surface expression on CD45⁺CD11c⁺MHCII⁺CD64⁻Ly6C⁻ DCs from B) inguinal lymph nodes and D) popliteal lymph nodes. C, E) Quantification of their respective percentage positive CD80 or CD86 are shown on the right. n = 5 mice/group, total = 20, and ANOVA followed by Tukey HSD post hoc tests were performed.

Propranolol favorably alters immune cell populations in size-matched B16-OVA tumors following vaccination

Since we found that sympathetic signaling affects the maturation of lymph node DCs, we aimed to evaluate how propranolol impacts immune cell populations in growing B16-OVA tumors. We have previously observed that tumor size has a great effect on MDSC infiltration into tumors leading to enhanced immunosuppressive cytokine production (unpublished data). Since these changes could skew immunoprofiling of the tumor microenvironment, we chose to utilize size-matched tumors for the following experiment. To obtain size-matched tumors, we inoculated B16-OVA cells in the combination group 3 days prior (day −3) and the vaccine alone group 1 day prior (day −1) to the propranolol and no treatment groups. Mice were treated according to the schematic in Figure 5a, and intratumor immune cell populations were analyzed via CyTOF once tumors reached 1200 mm³. We found that both the vaccine alone group and the combination group experienced increases in intratumor CD8⁺ T cells among CD45⁺ cells as well as increases in the frequencies of CD8⁺ effector memory T cells (T_EM) and PD-1⁺CD8⁺ T cells (Figure 5b, c). However, the combination of propranolol plus vaccine group experienced significantly greater increases in CD8⁺ T cells, CD8⁺ T_EM cells, and PD-1⁺CD8⁺ T cells compared to vaccination alone with approximately 4.8-fold, 13-fold, and 5.8-fold increases over no treatment, respectively (Figure 5c). Vaccination alone only resulted in approximately 2.3-fold, 4.6-fold, and 2.5-fold increases in CD8⁺ T cells, CD8⁺ T_EM cells, and PD-1⁺CD8⁺ T cells over no treatment, respectively (Figure 5c). Additionally, combination of propranolol plus vaccine resulted in a significant 2.4-fold increase in intratumor CD4⁺ T cells among CD45⁺ cells, whereas vaccination alone did not result in increased CD4⁺ T cell number (Figure 5c). Furthermore, propranolol alone resulted in significantly decreased CD4⁺ T cell expression of Foxp3 compared to each other group, but decreased Foxp3⁺CD4⁺ T cells was not observed in the combination group (Figure 5e). It has been previously reported that denervation of sympathetic nerves will reduce expression of the immune checkpoint molecules PD-1 and CTLA-4 on intratumor CD8⁺ and CD4⁺ T cells.[41] However, we did not observe decreased expression of either of these molecules in either the propranolol alone group or the combination group (Figure 5d, e).

Combination of propranolol plus vaccine also resulted in significant changes to immunosuppressive myeloid cell populations that was not observed in the vaccine alone group. The combination group exhibited significant decreases in the frequencies of polymorphonuclear MDSCs (PMN-MDSCs), PD-L1⁺ PMN-MDSCs, and PD-L1⁺ monocytic MDSCs (M-MDSCs) among CD45⁺ cells compared to the no treatment group (Figure 5b, f). Additionally, both propranolol alone and combination groups experienced significant decreases in PD-L1 expression on PMN-MDSCs and M-MDSCs compared to the no treatment group (Figure 5g, h). Combination of propranolol and vaccine also affected the MDSC enzymes inducible nitric oxide synthase (NOS2) and arginase-I (ArgI), which are both associated with tumor-promoting and immunosuppressive functions.[42] The combination group experienced significant decreases in PMN-MDSC expression of NOS2 and ArgI compared to the no treatment group (Figure 5g). Propranolol alone, vaccine alone, and the combination group each exhibited reduced Arg-I⁺ M-MDSCs compared to the no treatment group (Figure 5h). Lastly, the combination propranolol plus vaccine group exhibited significantly reduced TAMs compared to the vaccine group (Figure 5b, f).

Taken together, these results reveal that combination of propranolol plus a Poly ICLC/peptide cancer vaccine favorably alters intratumor immune cell populations by increasing the proportion of effector CD8⁺ T cells and decreasing regulatory cell types including PD-L1⁺ MDSCs and TAMs.

Propranolol improves peptide vaccine treatment of B16-OVA murine melanoma

Based the observed benefits to DC maturation and intratumor immune cells in propranolol plus vaccine treated mice, we hypothesized that this combination would improve cancer vaccine treatment of melanoma in mice. To test this combination, we performed another therapeutic vaccine study according to the schematic in Figure 6a.

Figure 6. Propranolol improves therapeutic vaccine treatment of B16-OVA melanoma:

A) C57BL/6J mice were subcutaneously inoculated with B16-OVA on day 0 and treated according to the schematic. B) Tumor volumes were calculated starting day 11 and repeated every 2 days until each group was sacrificed on day 20. C) Splenocytes were harvested from each mouse on day 20 and incubated with 10 μg/ml SIINFEKL peptide on IFN-γ antibody-coated ELISpot plates for 48 hours. After color development, number of spots per well were counted with a digital ELISPOT plate reader. n = 10 mice/group, total = 40, and ANOVA followed by Tukey HSD post hoc tests were performed on tumor volumes at day 20.

The propranolol plus vaccine combination group experienced dramatically reduced tumor volumes (approximately 300 mm³) by day 20 compared to the untreated control group (approximately 1700 mm³) as well as compared to the monotherapy treatment groups (Figure 6b). The vaccine monotherapy group had slightly reduced tumor volumes (approximately 1200 mm³) compared to untreated mice by day 20, but this difference did not reach statistical significance. The propranolol monotherapy group did not differ in tumor growth compared to the untreated control group. In addition, we isolated splenocytes from these groups to test for systemic immune responses to the vaccine peptide SIINFEKL. After incubating splenocytes from each group with SIINFEKL peptide on IFN-γ coated ELISPOT plates for 48 hours, we found that combination of propranolol and the vaccine led to an approximately 2.5-fold increase in IFN-γ producing splenocytes compared to vaccine monotherapy (Figure 6c). These results show that combination of propranolol with a peptide cancer vaccine improved treatment of melanoma in mice as well as improved systemic IFN-γ production.

Discussion

In this study, we report the novel finding that CD115⁺ moMФ in BMDC cultures are responsible for the observed changes to cytokine production following norepinephrine treatment, whereas CD115⁻ DCs do not express surface β₂AR and are unresponsive to norepinephrine treatment. This relationship appears to hold true in vivo since we demonstrated that DCs in mouse lymph nodes and spleens do not express surface β₂AR, but macrophages and moDCs in these tissues do express β₂AR. GM-CSF grown BMDC cultures have been previously used to demonstrate that β₂AR signaling impairs DC production of IL-12p70 and subsequent priming of T_H1 cells.[27, 43, 44] However, these studies did not account for heterogeneity within CD11c⁺MHCII⁺ BMDC cells.[25] We found that CD115 alone is an adequate marker to separate DCs and moMФ in BMDC cultures, evident by unique cell morphology, expression of DC markers, and ability to stimulate CD8⁺ T cell proliferation. We then demonstrated that β₂AR signaling does not directly impact DCs as previously reported, but instead directly affects β₂ARs on moMФ. Importantly, this in vitro observation correlates with our finding that mouse lymph node and spleen cDCs lack surface β₂ARs. cDCs receive much attention as mediators of adaptive immunity, but circulating monocytes and peripheral moMФ can migrate to lymph nodes and adopt APC characteristics under inflammatory conditions to modulate T cell responses.[19, 20] These MCs contribute to T cell priming and exhibit distinct and nonredundant functionality compared to DCs.[19–21] Our results indicate that moMФ possess a unique role compared to cDCs as key mediators of SNS regulation of the APC immune compartment.

We additionally explored the impact of sympathetic signaling on cancer vaccine activity by evaluating DC function, intratumor immune cells, and tumor growth in melanoma-bearing mice that were treated with a combination of the non-specific β-blocker propranolol and a peptide cancer vaccine. Combination of propranolol and a cancer vaccine improved DC maturation in lymph nodes near the vaccination site evident by increased expression of CD80 and CD86 costimulatory markers. This finding is consistent with previous reports that increased SNS tone will impair DC maturation and that blockade of SNS signaling will improve APC function and subsequent T cell priming.[39, 45, 40] Despite our finding that mouse DCs lack β₂AR, propranolol improved DC maturation following vaccination, possibly by altering the local inflammatory milieu via MCs.

In size matched B16-OVA tumors, we found that combination of propranolol with a peptide vaccine increased the proportion of total and effector CD8⁺ T cells among CD45⁺ cells compared to vaccination alone. Interestingly, propranolol alone resulted in significantly reduced expression of Foxp3⁺ in CD4⁺ T cells that was not observed in the combination group. This change in Foxp3⁺ expression is consistent with a previous study that reported decreased Foxp3⁺ expression in intratumor CD4⁺ T cells following sympathetic denervation.[41] Additionally, the combination of propranolol and a cancer vaccine favorably altered several myeloid cell populations. Both propranolol alone and combination groups exhibited decreased PD-L1 expression on MDSCs, which is consistent with a previous study that reported β₂AR signaling greatly supports MDSC survival and immunosuppressive functions.[10] It was reported that β₂AR signaling in MDSCs was associated with increased PD-L1 expression, increased STAT3 phosphorylation, and decreased apoptosis gene signatures.[10] However, it is unclear how β₂AR signaling mediates STAT3 phosphorylation, which regulates many immunosuppressive functions in MDSCs.[46] Based on our findings, β₂AR-mediated increases in IL-10 could potentially explain increased STAT3 phosphorylation and associated immunosuppressive functions in MDSCs, but further studies into this relationship are needed. Furthermore, the combination group exhibited significantly reduced TAMs compared to the vaccine group. This finding is consistent with two previous studies that found βAR signaling contributes to TAM infiltration into tumors and cancer progression.[47, 11]

Another study that combined propranolol and a cancer vaccine was recently published and similarly showed dramatic improvement in anti-cancer efficacy.[48] The authors demonstrated that naïve CD8⁺ T cells are responsive to β₂AR signaling but limited their evaluation of DC function to in vitro analyses and did not evaluate changes to the intratumor myeloid environment. In contrast, we show that the combination of propranolol and a vaccine resulted in improved DC maturation in lymph nodes, which likely improves naïve T cell priming. In addition, the combination treatment in our study dramatically minimized immune regulatory elements through changes to MDSC and TAM populations. Thus, we present novel findings into SNS modulation of the myeloid cell compartment in lymph nodes and the tumor environment that impair cancer vaccine activity.

In conclusion, our results show that SNS signaling directly impacts moMФ rather than DCs and that combination of a β-blocker with a cancer vaccine dramatically improves anti-tumor efficacy in the treatment of melanoma in mice. This combination treatment resulted in improved lymph node DC maturation, increased intratumor total and effector CD8⁺ T cells, and minimized immunoregulatory MDSCs and TAMs in the tumor microenvironment. Based on these benefits, our study provides strong support for the combination of β-blockers with cancer vaccines in a clinical setting. β-blockers such as propranolol are widely prescribed and rarely contraindicated, so they are poised to be rapidly repurposed for use in combination with cancer immunotherapeutics. Additionally, it is understood that aging leads to reduced cellular immunity and that sympathetic regulation of immunity is intact in advanced age.[49, 50] Therefore, combination of β-blocker agents and cancer vaccine could provide much needed benefit in elderly patients. Our study also demonstrates the need to delineate myeloid subpopulations in BMDC cultures when studying DC biology.

Abbreviations:

SNS

Sympathetic nervous system

moMФ

monocyte derived macrophages

DCs

dendritic cells

β₂AR

β₂-adrenergic receptor

TAMs

tumor associated macrophages

MDSCs

myeloid-derived suppressor cells

GPCR

G-protein coupled receptor

cDCs

conventional dendritic cells

MCs

monocyte derived cells

BMDCs

bone marrow-derived DCs

Poly IC

polyinosinic-polycytidylic acid

Poly ICLC

polyinosinic-polycytidylic acid complexed with poly-L-lysine

BME

2-mercaptoethanol

PS

penicillin and streptomycin

OT-I

OVA_257–264-specific CD8⁺ T cells

PMN-MDSCs

polymorphonuclear-MDSCs

M-MDSCs

monocytic-MDSCs

T_EM

effector memory T cells

NOS2

inducible nitric oxide synthase

ArgI

arginase-I