ADRB3 expression in tumor cells is a poor prognostic factor and promotes proliferation in non-small cell lung carcinoma

DMEM/F12 medium and fetal bovine serum (FBS) were from Invitrogen (Carlsbad, CA). ADRB3 antagonist SR59230A (SR) was from Sigma-Aldrich (St Louis, MO). Antibodies against the following proteins were used: ADRB3, Myeloperoxidase (MPO), IL-6, IFN-γ, CD19, CD68, Nucleolin, which were purchased from Abcam (Cambridge, UK); Ki-67, mTOR, Rictor, p53 and GAPDH, which were purchased from Cell Signaling Technology (Danvers, USA).

A Balb/c mouse was intraperitoneally immunized with human recombinant ADRB3 protein two times at 2-week intervals. Mouse was sacrificed and its spleen cells were collected and fused with myeloma cells by modified hybridoma technique. After the HAT medium (Sigma–Aldrich) selection, the culture supernatants obtained from the hybridoma-containing wells were analyzed for antibody reactivity by Western blotting. A hybridoma cell line was selected and amplified. Balb/c mice received intraperitoneal injection of 10⁶ hybridoma cells, and 8 days after inoculation, the ascites was collected. The saturated ammonium sulfate was concentrated in ascites and purified by Protein A affinity chromatography column.

The lung cancer tissue arrays and patient profiles were from Shanghai Outdo Biotech (Shanghai, China). All the patients were diagnosed with primary non-metastatic lung cancer by pathology and did not receive chemotherapy or radiotherapy before surgery. Tissue arrays were constructed from formalin-fixed paraffin-embedded tissue blocks of pretreatment biopsy specimens. Marrow and blood smear of NSCLC patient and healthy person were from bone marrow test lab in Guangdong General Hospital.

Standard deparaffinization, rehydration and antigen retrieval procedures were performed. Tissue arrays were stained with ADRB3 antibody at dilution of 1:200 for 2 h. HRP activity was visualized using the Liquid DAB Plus Substrate Kit (Thermo Scientific) according to the manufacturer's instructions. At each timepoint, 6 ~ 8 fields were randomly observed and if > 5% of the cells were positive, the section was identified as a case of positive staining. Light yellow brown staining was considered weakly positive (+), brown staining was considered positive (++) and dark brown staining was regarded as strongly positive (+++). The IHC score was interpreted by histopathological evaluation by the authors (MZ and ZX), using the following criteria: 0 = negative; 1 = equivocal/uninterpretable; 2 = weak positive; 3 = strong positive.

The human lung cancer cell lines, A549 were maintained in DMEM/F12 medium containing 10% FBS. Cells were fixed with 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100. Cells were blocked in PBS containing 2% BSA and incubated with mouse anti-Ki67 and rabbit anti-ADRB3 (1:200) overnight at 4℃. Cells incubated with goat anti-mouse secondary antibody conjugated to Alexa Fluor 488 and goat anti-rabbit secondary antibody conjugated to Alexa Fluor 555, followed by DAPI staining. Images were acquired on a scanning confocal microscope (Leica, Germany) and analyzed with Fluorchem 8900 software (Alpha Innotech, CA). ADRB3 expression levels were expressed as the geometric mean fluorescence intensity (MFI).

pcDNA3-ADRB3 were built. A small interfering RNA (siRNA) for the specific inhibition of ADRB3 expression and a negative control siRNA were synthesized by Sigma (Shanghai, China). A549 cells were plated in a 6-well plate before transfection. After 70% confluence was achieved, 2 μg plasmids or 100 pmol siRNA was diluted with 100 μL serum-free medium using the Lipofectamine 3000 transfection reagent (Life Technologies, CA) according to the manufacturer’s instructions.

For the MTT assay, 200 μl of 0.5 mg/ml MTT-tetrazolium salts (Sigma–Aldrich) in DMEM/F12 was added to each well. After 4 h of incubation, the formazan crystals were dissolved by adding DMSO. The absorption of the formazan solution was measured using Multiskan G0 spectrophotometer at a wavelength of 570 nm.

For flow cytometric tests of the cell cycle, propidium iodide (PI) staining was measured at an excitation wavelength of 488 nm and emission wavelength of 670 nm.

Whole-cell lysates were generated using RIPA lysis buffer (Abcam, UK). Total proteins were separated using 10% SDS-PAGE and then transferred onto a nitrocellulose membrane. The membrane was incubated with the primary antibody at 4 °C overnight, followed by a horseradish peroxidase-conjugated secondary antibody the next day for 1 h at room temperature. The immunoreactive bands were visualized using enhanced chemiluminescence reagents.

RNeasy Mini Kit was used (Qiagen, Germany) for RNA isolation. The reverse transcription reaction was performed using High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Applied Biosystems, USA). Changes in the expression level of ADRB3 were tested using ABI 7500 real-time PCR system. The results were standardized, based on the expression of the reference gene of β-actin. The obtained results were shown in the graphs on a logarithmic scale and subjected to statistical analysis. Mouse ADRB3-forward primer: 5′-GACAGCCTCAAATGCATCCT-3′; Mouse ADRB3-reverse primer: 5′-CCCAGTCCACACACACCTTTCT-3′.

FVB/N genetic background ADRB3-null mice were purchased from JAX. C57BL/6 genetic background ADRB3-null mice were generated using C57BL/6 mice backcrossed for nine generations. All mice used in this study were maintained and used at the Guangdong General Hospital mouse facility under pathogen-free conditions according to institutional guidelines and animal study proposals approved by the Institutional Animal Care and Use Committee.

C57BL/6 mice were injected with 2 × 10⁵ LLC cells obtained from the China Type Culture Collection. Tumor dimensions were calculated by caliper measurements and volume was calculated according to the equation V = (π/6) × a × b × c, where a, b and c are diameters in three perpendicular dimensions.

Plasma levels of IL-2, IL-5 and IL-6 were determined using the Mouse Th1/Th2 Array Q1 Kits (RayBiotech, GA). MPO was determined using the Mouse MPO Quantikine ELISA Kits (R&D Systems,) according to manufacturer's instructions.

An EMSA was performed using oligos corresponding to p53 (5′‐TACAGAACATGTCTAAGCATGCTGGGGACT‐3′) as biotin-labeled probe. Nuclear proteins were extracted using Nuclear and Cytoplasmic Protein Extraction Kit (Viagene, China). For each sample, 6 μg nuclear proteins were pre-incubated with the Gel-Shift Binding Buffer to block non-specific binding prior to the addition of the probe (100 pmol) and further incubation for 20 min at 25 °C. For supershifts, nuclear extracts were pre-incubated with 3 μg p53 antibody or non-immune IgG (Santa Cruz Biotechnology) (negative control) for 15 min at 4 °C prior to addition of the probe. Electrophoresis was carried out on non-denaturating polyacrylamide gels (6%) in 0.5 × TBE at 100 V for 90 min and then electrophoretically transferred onto a positively-charged nylon membrane in 0.5 × TBE at 300 mA for 45 min. The DNA–protein complex was visualized with streptavidin-HRP Conjugate.

To examine the distribution of ADRB3 expression in NSCLC, we performed ADRB3 immunostaining on NSCLC tissue microarray (TMA). ADRB3 was only expressed in AMs of paracancerous tissue. In contrast, both cancer cells and AMs expressed ADRB3 at higher level in cancerous tissue (Fig. 1a–d). Moreover, ADRB3 was also highly expressed in squamous dysplasia (Fig. 1e).

ADRB3 was localized in the cytoplasm and nucleus of tumor cells. The positive expression rate of ADRB3 in cancerous tissue (94.6%) was markedly higher than that in paracancerous tissue (7.2%) of 166 NSCLC samples including 87 AD and 79 SCC (Table 1). There is no statistical difference of ADRB3 expression between AD and SCC groups. 57 (34.3%) cases had positive staining in ≥ 25% tumor cells (TC). 37 (22.3%) cases had positive staining in ≥ 50% TC. The higher grade g of malignancy, the higher ADRB3 expression was observed. The differences were statistically significant between G1 versus G2 and G1 versus G3 (Table 1). ADRB3 expression was increased in higher stage (S) of NSCLC. The differences were statistically significant between S1 versus S2 and S1 versus S3 (Table 1). In addition, the expression of ADRB3 in AMs was higher from the patients with inflammatory cell infiltration (ICI) (Fig. 1f, IHC score, 2.7 ± 0.4) compared to patients without ICI (IHC score, 1.8 ± 0.3, P < 0.05). Finally, we evaluated the association between ADRB3 expression in lung cancers and the impact on patients’ survival. Log-rank (Mantel-Cox) analysis showed that patients with weak-positive ADRB3 had longer survival compared to the group with strong-positive (P = 0.038, Fig. 1g).

To explore ADRB3 function in lung cancer cells, pcDNA3-ADRB3 was generated and transfected A549 cells. After 96 h, we performed MTT-assay to determine the viability. We found that pcDNA3-ADRB3 conferred a proliferative advantage to A549 cells suggesting a role of ADRB3 in promoting lung cancer growth (Fig. 2a). Conversely, the selective ADRB antagonist SR59230A and ADRB3 siRNA resulted in significant decreases in cell proliferation (Fig. 2b, c). Anti-human ADRB3 monoclonal antibody (M5D1) developed by our team also inhibited A549 cells viability in a dose-dependent way (Fig. 2d–f). In addition, M5D1-treatment resulted in a dose-dependent increase of apoptosis (Fig. 2g).

Relocation of p53 to the nucleus after cellular stress is desirable to inhibit the growth of malignant cells [21]. So we examined p53 localization in A549 cells cultured in the presence of 0.4 mg/ml M5D1 for 12 h. We found that treatment with M5D1 induced p53 nucleus accumulation in 73.8% of the A549 cells (Fig. 2h–j). To test the activity of p53, EMSA was applied. As shown in Fig. 2k, p53 was activated by M5D1 (0.4 mg/ml), peaking at 2 h and declining at 4 h post-treatment. Furthermore, we tested the effect of M5D1 on modulating cell-cycle progression and found G1 arrest from 22.4 ± 3.5% in Isotype IgG to 65.3 ± 7.1% in M5D1 (0.4 mg/ml) cells (P < 0.01). We then analyzed the effect of ADRB3 on cell cycle regulators in A549 cells. A strong increase in p53 and decrease in mTORC2 was detected in response to the M5D1 compared with the control (Fig. 2l).

Circulating monocytes are recruited to the lung, where they differentiate into monocyte-derived alveolar macrophages (Mo-AMs), so the immunofluorescence staining of bone marrow, blood and tumor tissue for ADRB3 was performed. ADRB3 expression in monocytes and lymphocytes of bone marrow increased significantly in NSCLC compared with healthy control (MFI 11.7 ± 1.6 versus 2.2 ± 0.6, P < 0.01; 2.1 ± 0.2 versus 0.6 ± 0.1, P < 0.01, respectively, Fig. 3a–c). High expression level of ADRB3 was observed when we compared NSCLC patient's peripheral blood monocytes and lymphocytes with cells from healthy individuals (MFI 10.7 ± 1.3 versus 1.2 ± 0.2, P < 0.01; 1.7 ± 0.3 versus 0.4 ± 0.1, P < 0.01, respectively, Fig. 3D-F). Furthermore, we found that Ki-67 expression in ADRB3⁺ monocytes increased significantly in NSCLC compared with healthy control (MFI 15.5 ± 2.3 versus 1.7 ± 0.2, P < 0.01, Fig. 3g–i). These data suggests that ADRB3⁺AMs may be Mo-AMs. These observations suggest that ADRB3 is produced by tumor cells and monocytes and plays a critical role in the recruitment of the proliferative monocytes from the blood into tumor tissues, where they differentiate into inflammatory Mo-AMs which interact with tumor cells and are involving in shaping immunosuppressive microenvironment.

Based on ADRB3 expression in cancer cells, Mo-AMs, monocytes and lymphocytes, we wondered whether ADRB3 would modulate tumor immunity in vivo. The role of ADRB3 in tumor immunity was investigated using mouse tumor models with invasive mouse lung carcinoma (LLC, 2 × 10⁶ cells per mouse) established in C57BL/6 background ADRB3^−/− (A 306 bp genomic fragment containing the sequences encoding the third through the fifth transmembrane domains was replaced with a neomycin selection cassette) and their ADRB3^+/+ littermate. During 2 weeks after subcutaneous inoculation of LLC cells, we observed almost equal tumor growth in both ADRB3^−/− and ADRB3^+/+ mice. However, in the subsequent weeks, significant tumor regression was found in ADRB3^−/− mice, where the tumors became undetectable at week 3 (Fig. 4a). In contrast, the tumors in ADRB3^+/+ mice grew rapidly and invasively, resulting in 100% mortality.

The significant decrease in the levels of serum Interleukin-2 (IL-2), IL-5 and IL-6 (Fig. 4b) and decrease in the number of white blood cells, monocytes (Fig. 4c) were found in ADRB3^−/− mice compared with ADRB3^+/+ mice. There was no difference in the number of lymphocytes. The percentage of monocyte was significantly decreased and the percentage of lymphocytes was increased in ADRB3^−/− mice (Fig. 4d). Biomarkers of inflammation such as neutrophil–lymphocyte ratio (NLR) (Fig. 4e) was decreased in ADRB3^−/− mice.

Macrophages marker CD68, proinflammatory cytokines mainly produced by macrophages such as MPO and IL-6 [22] were all expressed at lower levels in ADRB3^−/− mice lung than ADRB3^+/+ mice (Fig. 4f). The downregulation of CD68, MPO and IL-6 suggests that ADRB3 knockout eliminate recruitment and activation of macrophages in lung. However, the expression of IFN-γ was higher in ADRB3^−/− mice lung, suggesting that ADRB3 knockout might promote naive T cell differentiation into T helper type 1 (Th1) effector cells and increased IFN-γ production. The impaired antitumour effect of CD4 + T cells with their defective Th1 differentiation is restored by IL-6 deficiency [23]. It is speculated that ADRB3 knockout enhances Th1 differentiation with downregulation of IL-6.

We found that lung cancer cells aberrantly express ADRB3 and activation of this receptor promotes tumor cell growth. ADRB3 is also present on Mo-AMs that help tumors evade immune response. ADRB3 is likely to be a novel promising therapy target for cancer, so we've developed anti-ADRB3 monoclonal antibodies. As in Supplementary Fig. 1, the SDS-PAGE analysis of purification of ascites fluid produced by hybridoma clone M5D1. From the analysis, lane 5, 7, 9 showed the molecular weight of IgG, for heavy chain was ~ 55 kDa and its light chain was ~ 25 kDa. After purification process of antibody, the purify monoclonal antibody M5D1 from ascites fluids was test with immunofluorescence technique to identify antigens expressed on A549 cells. Using dilution at 1:100 of purified antibody the experiment was indicated that a bright of fluorescence labeled on the cytoplasm of A549 cells as in Supplementary Fig. 2.

To examine the efficacy of the M5D1 in inhibiting lung tumor growth in vivo, we utilized xenograft mouse model and evaluated the growth of tumors after injection of the antibody into mice. C57BL/6 mice were subcutaneously injected with LLC cancer cells in the dorsal right flank and allowed to grow until the tumor reached 80 mm³. The mice were intravenously injected through the tail vein with 50 μg M5D1 every 5 days or isotype IgG.

When assessing tumor volume, we discovered that mice treated with M5D1 displayed reduced growth of tumors compared to control (Fig. 5a). Furthermore, body weight of the mice was increased by treatment with M5D1. The significant decrease in the level of serum IL-6, MPO (Fig. 5b) and decrease in the number of WBC and monocytes (Fig. 5c) was found in M5D1 group. There was no difference in the number of red blood cells and lymphocytes.

Mice peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll density gradient centrifugation, lymphocytes and monocytes were separated by their adherence to the culture plate. ADRB3 expression was assessed by qPCR analysis. We found that lymphocytes isolated from non-tumor-bearing mice expressed very low to undetectable ADRB3, whereas ADRB3 expression was higher in lymphocytes (4.8-fold; Fig. 5d) and monocytes (11.4-fold; Fig. 5d) from tumor-bearing mice compared with non-tumor-bearing mice. After 15d of M5D1 treatment, the lymphocytes and monocytes from tumor-bearing mice showed an average 1.7- or 1.4-fold increase in ADRB3 expression, respectively, compared with non-tumor-bearing mice.

Immunohistochemistry showed that M5D1 downregulated the expression of ADRB3, MPO, and increased the expression of CD19 in the spleen tissues (Fig. 5e). CD19 regulates B-cell development, activation and differentiation. The upregulation of CD19 suggests that ADRB3 blockade can induce B cells to proliferate and differentiate into CD19^high blasts. ADRB3 might be working as a molecular switch, which turns the differentiation signals off in appropriate time, and its absence may leave constitutively activated differentiation-promoting signals on, at least regarding the mTOR pathway. In the tumor tissues, M5D1 significantly decreased IL-6 and increasd IFN-γ (Fig. 5e). IFN-γ is a cytokine produced both by macrophages and by Th1. It is speculated that ADRB3 blockade enhances Th1 differentiation with downregulation of IL-6 and upregulation of IFN-γ. Overall, these results suggested that ADRB3 blockade can enhance the innate and/or adaptive antitumor immune response by dampening macrophage function.