Isoprenaline (ISO, I5627), formoterol (product no. F9552), ICI118,551 (product no. I127), atenolol (product no. A7655) and 3-methyladenine (3-MA, product no. M9281) were procured from Sigma-Aldrich/Merck. SBI-0206965 was procured from Selleck Chemicals. Antibodies for β1-AR (product no. ab3442), β2-AR (product no. ab182136), N-MYC (product no. ab24193), and Beclin-1 (product no. ab109631 for IHC assay) were obtained from Abcam. Antibodies for LC3B (product no. 3868), beclin-1 (product no. 3495), CREB (product no. 48H2), p-CREB (product no. 87G3), p-ULK1 (product no. 5869S) and ULK1 (product no. 8054T) were obtained from Cell Signaling Technology. LC3-II (cat. no. A11923) for IHC staining was purchased from Abclonal. ULK1 for the IHC assay was obtained from Zen BioScience.

The clinical samples of ganglioneuroma (GN), ganglioneuroblastoma (GNB, nodular and intermixed), and NB were obtained from the Guangzhou Women and Children's Medical Center (26, 34 and 71 samples, respectively). The date range of sample collection was from January 2003 to September 2015. These samples were used for constructing tissue microarrays (TMAs) by the Shanghai Outdo Biotech Company. The clinical information of these samples is documented in Table I. For the IHC staining of the TMAs, the clinical tumor samples were fixed and then embedded in paraffin. The tissue slides were treated with endogenous peroxidase blocking solution and normal goat serum to block nonspecific background interactions. Then, the tissues were incubated with anti-β1-AR (dilution 1:200), -β2-AR (dilution 1:200), -ULK1 (dilution 1:200), -beclin-1 (dilution 1:200) and -LC3-II (dilution 1:200) antibodies at 4°C overnight. After incubation with the primary antibodies, the slides were incubated with horseradish peroxidase-labeled anti-rabbit or mouse secondary antibodies for 30 min at room temperature; they were then stained with the chromogenic substrate diaminobenzidine (DAB; Dako; Agilent Technologies, Inc.), and counterstained with hematoxylin. The evaluation of tissue microarray scores was carried out by two independent researchers based on the positive staining intensity and the percentage of positively stained cells. The IHC results were scored as 0 (negative), 1–2 (weakly positive), 3–4 (moderately positive), 5–6 (strongly positive), and high expression of tumor cells were defined as scores >3.

This study was approved by the Human Ethics Committee of the Affiliated Guangzhou Women and Children's Medical Center, Zhongshan School of Medicine. All patients gave informed consent before participation in the present study.

NB cells including SK-N-AS, SK-N-SH, SHEP, SK-N-BE2, SK-N-BE2c and Kelly was used as cell models. Among these cell lines SK-N-AS, SK-N-SH and SHEP were N-MYC non-amplified cells, while SK-N-BE2, SK-N-BE2c and Kelly were N-MYC amplified cells. These cell lines were a kind gift from Professor Li Bo, Sun Yat-Sen Medical School. These cells were cultured in high-glucose Dulbecco's modified Eagle's medium (DMEM; Corning, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) at 37°C and 5% CO₂ conditions in a humidified incubator.

For the CCK-8 assay, SK-N-BE2 and Kelly cells were seeded into 96-well plates at densities of 3×10³ and 5×10³ cells per well, respectively. These cells were treated with selective and non-selective β-AR agonists and antagonists including isoprenaline, ICI118,551 and atenolol for the indicated time periods. Then, the absorbance of the samples was detected at 450 nm according to the manufacturer's instructions.

Proteins from NB cells were harvested using lysis buffer (glycerol, 0.5 M Tris-HCl, 10% SDS, ddH₂O), and the protein concentrations were detected using the BCA kit. Then, the protein samples were treated with 10X loading buffer (0.2% bromophenol blue, β-mercaptoethanol) and 30 µg protein was separated on 10 and 15% SDS-PAGE gels; the resulting protein bands were transferred onto PVDF membranes. The membranes were blocked with 5% skimmed milk for 1 h at room temperature. Then, they were incubated with the indicated antibodies anti-β1-AR (dilution 1:1,000), -β2-AR (dilution 1:1,000), -ULK1 (dilution 1:1,000), -p-ULK1 (dilution 1:1,000), CREB (dilution 1:1,000), p-CREB (dilution 1:1,000), -LC3-II (dilution 1:1,000), N-MYC (dilution 1:1,000) and β-actin (dilution 1:6,000) at 4°C overnight with shaking, followed by incubation with horseradish peroxidase (HRP)-labeled secondary antibodies (dilution 1:3,000) for 2 h at room temperature. The enhanced chemiluminescent (ECL) kit (P90720; Millipore) was used to visualize the western blot results, and ImageJ (×64; National Institutes of Health, Bethesda, MD, USA) software was applied to quantify the densitometry.

For EdU staining, a total number of 2×10⁵ SK-N-BE2 and Kelly cells were mounted onto coverslips in 6-well plates. After being cultured for 24 h, the cells were treated with the indicated reagents for another 24 h. Then, they were incubated with EdU reagents (Nanjing KeyGen Biotech Co., Ltd.) for 2 h, followed by subsequent staining according to the manufacturer's instructions. The images were captured using a fluorescence microscope at a magnification of ×100 (Olympus BX63; Olympus Corp., Tokyo, Japan). A recombinant adenovirus expression vector containing tandem RFP-GFP-LC3 (Hanbio Biotechnology Co.) was used to assess the change of autophagic flux. The MOI was 100, and the cells were incubated with the adenoviruses for 24 h before the subsequent treatments. The images were obtained using a confocal microscope at a magnification of ×630 (LSM 780; Carl Zeiss).

siRNAs were used to knock down the expression of β2-AR and CREB. The siRNAs were synthesized by Guangzhou RiboBio Co., Ltd. (Guangzhou, China). For interfering with the functions of β2-AR and CREB, SK-N-BE2 cells were seeded into 6-well plates at a density of 5×10⁵ cells per well, and transfected with 100 nM of the respective siRNAs using the transfection reagent RNAiMAX (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. After culturing the cells for 48 h, the interference efficiency was verified by western blotting; the cells were subjected to the subsequent treatments based on the above methods. The siRNA sequences of β2-AR and CREB used in our experiments are as follows: β2-AR (CATCCTCCTAAATTGGATA) and CREB (siRNA1, CGTAGAAAGAAGAAAGAAT; siRNA2, GCCACAGATTGCCACATTA; and siRNA3, GCAATACAGCTGGCTAACA).

Student's t-test and one-way ANOVA with post hoc contrasts by a Newman-Keuls test were used to assess the differences between two or three groups by means of GraphPad Prism 5 software (GraphPad Software, Inc.). P-values <0.05 were considered statistically significant.

To investigate the role of β-AR in NB, we first detected the expression of β-AR in the clinical tumor samples. The expression of β-adrenergic receptors was detected in samples of three peripheral neuroblastic tumors (PNTs) including GN, GNB (nodular and intermixed) and NB. Among these subtypes, NB is the most malignant and consistently shows an unfavorable prognosis and low survival rate; the other two types of tumors consistently exhibit a good prognosis compared to NB (9). β-AR includes β1-AR, β2-AR and β3-AR. We mainly tested the expression of β1-AR and β2-AR in our study. IHC results using the TMA samples revealed that the levels of β1-AR and β2-AR were significantly upregulated in the NB samples, compared to the levels noted in the GN and GNB samples (Fig. 1A). The representative images of IHC staining scores of clinical NB samples are shown in Fig. 1B. We further evaluated the expression of β1-AR and β2-AR in different clinical stages of NB, but there was no notable correlation between the β1-AR and β2-AR expression in the different clinical stages of NB (Fig. 1C). This result suggests that the β-AR pathway may play a crucial role in the tumorigenesis of NB.

To further investigate the role of β-AR signaling pathway in NB cell growth, the expression of β1-AR, β2-AR and N-MYC was detected in six different NB cell lines. The amplification status of N-MYC is an important prognostic marker for NB. The expression levels of β1-AR and β2-AR in six NB cell lines are shown in Fig. 2A. We chose two cell lines SK-N-BE2 and Kelly which express high level of β2-AR as cell models. Treatment with the non-selective β-AR agonist isoprenaline (ISO) increased the viability of the NB SK-N-BE2 and Kelly cell lines (Fig. 2B and C). In order to demonstrate which β-AR is responsible for the promotive effect of β-AR activation on cell survival, we utilized specific antagonists of β1-AR (atenolol) and β2-AR (ICI118,551) to ascertain the role of the β-AR pathway in NB cell growth. The results showed that treatment with the β1-AR antagonist atenolol did not affect NB cell survival, while treatment with the β2-AR antagonist ICI118,551 suppressed the cell viability in a concentration-dependent manner (Fig. 2D). These findings demonstrate that the increased proliferation of NB cells induced by β-AR activation occurs through β2-AR signaling. However, the mechanisms underlying the promotive role of the β2-AR pathway on NB cell proliferation have not yet been clarified.

Considerable research has demonstrated the critical role of the autophagy pathway in the initiation and progression of tumors, and autophagy plays a dual role in tumorigenesis and metastasis. To study the role of autophagy in NB, we assessed the expression of the important autophagy markers ULK1, beclin-1 and microtubule-associated protein 1 light chain 3-II (LC3-II) in the clinical tumor samples. ULK1 is a serine/threonine-protein kinase, and the activation of ULK1 can initiate autophagy via the phosphorylation of beclin-1 and the subsequent enhancement of the formation and activity of the autophagic complex Atg14/beclin-1/VPS34 (39). LC3-II is located in the autophagosome and promotes the fusion of autophagosomes with lysosomes, leading to the formation of autolysosomes. The IHC results showed that the levels of ULK1, beclin-1 and LC3-II were significantly increased in the clinical samples of NB, compared to these levels in the GN and GNB samples (Fig. 3). Clinical correlation analysis revealed that the expression of ULK1, beclin-1 and LC3-II correlated positively with that of β2-AR, but not β1-AR (Tables II and III), implying that there is a crosstalk between the β2-AR pathway and autophagy in NB. Thus, we further investigated the role of β2-AR signaling in regulating autophagy in NB cells.

To explore the role of β-AR in autophagy, NB cells were stimulated with the β-AR agonist ISO, and then expression of the autophagy marker LC3 was detected. Upon its synthesis, LC3 is immediately cleaved by Atg4 and is converted to its cytoplasmic form LC3-I. During autophagy activation, cytoplasmic LC3-I is conjugated with phosphatidylethanolamine (PE) by the action of Atg3 and Atg7, and is then translocated into the membranes of autophagosomes, turning into autophagic LC3-II; thus, it promotes the elongation and maturation of autophagosomes. Then, mature autophagosomes fuse with lysosomes to form autolysosomes, degrading the components within themselves. LC3-II has always been used as a molecular marker for the activation of autophagy. Firstly, we used GFP-RFP-labeled adenovirus expressing LC3 to evaluate the change of autophagic flux. During the late stage of the autophagy process, the formation of autolysosomes causes a decrease in pH, which results in the quenching of GFP, as GFP is sensitive to acidic conditions. The results showed that ISO treatment improved the accumulation of yellow (autophagosomes) and red (autolysosomes) vacuoles (Fig. 4A). The western blotting results suggested that the treatment of NB cells with ISO promoted the expression of LC3B-II in the SK-N-BE2 and Kelly cell lines (Fig. 4B), revealing that the activation of β-AR signaling promotes the initiation of autophagy. Treating NB cells with the specific β2-AR agonist formoterol also promoted the expression of LC3B-II (Fig. 4C). These results suggest that the β2-AR pathway promotes the activation of autophagy in NB cells.

To further investigate the role of β2-AR on the autophagy of NB cells, we performed a genetic loss-of-function assay to determine the ISO-induced effects of β2-AR on NB cell autophagy. We transfected NB cells with β2-AR-specific siRNA to inhibit the expression of β2-AR and non-specific scrambled siRNA as a control. The western blot analysis showed that the knockdown of β2-AR suppressed the ISO-induced expression of LC3B-II (Fig. 5A). The results of the GFP-RFP-LC3-labeled adenovirus assay showed that inhibition of β2-AR expression decreased the ISO-induced formation of yellow (autophagosomes) and red (autolysosomes) vacuoles (Fig. 5B), suggesting that β-AR signaling activates autophagy via the β2-AR pathway.

Our cellular functional assays demonstrated that stimulation of NB cells with ISO promoted cell growth and activated autophagy. Further clinical analysis revealed that there was a positive correlation between β2-AR and the autophagy markers ULK1, beclin-1, and LC3-II in the clinical samples of NB. Thus, we proposed that the promotive effect of β2-AR on NB cell proliferation may depend on autophagy. To verify this hypothesis, we inhibited autophagy using 3-MA, which is a pharmacological reagent commonly used to block the formation of the beclin-1/Atg14/VPS34 complex, leading to the prevention of autophagy activation. Our results of the CCK-8 and EdU assays showed that treatment with 3-MA abolished the ISO-induced enhancement of NB proliferation in the Kelly and SK-N-BE2 cell lines (Fig. 6A and B). These results suggest that the β2-AR signaling pathway promotes NB cell growth by activating the autophagy pathway.

In order to further investigate the molecular mechanism of β-AR signaling-stimulated autophagy, we investigated the role of transcriptional factor cAMP-response element binding protein (CREB) in regulating autophagy. CREB is a downstream regulator of β-AR signaling. It has been reported that the activation of CREB under starvation conditions could directly target autophagy-related gene ulk1 to induce autophagy in hepatic cells (40). As CREB is a downstream factor of β-AR signaling, thus, we hypothesized that CREB may also regulate ISO-activated autophagy in neuroblastoma. Above all, we aimed to investigate the regulatory role of CREB in ULK1 expression, which is important to clarify the regulatory mechanism of CREB in autophagy of NB cells. To verify this hypothesis, first of all, we detected the level of p-CREB in clinical samples of GN, GNB and NB; the phosphorylation of CREB was notably upregulated in the NB samples, compared to that noted in the GN and GNB samples (Fig. 7A). Treatment of NB cells with β-AR agonist ISO promoted the phosphorylation of CREB and the autophagy-related protein ULK1 in NB SK-N-BE2 cells (Fig. 7B). Then, we used siRNA to inhibit the expression of CREB; the results of the CREB knockdown assay revealed that CREB knockdown abolished the ISO-induced enhancement of ULK1 and LC3B-II expression (Fig. 7C). As ULK1 expression is downregulated after CREB knockdown, we hypothesized that CREB activates autophagy by targeting ULK1. Further study demonstrated that use of the specific ULK1 inhibitor SBI-0206965 inhibited the ISO-induced expression of LC3B-II (Fig. 7D), suggesting that ISO induces autophagy through CREB-mediated ULK1 expression. These results showed that β-AR activated autophagy via CREB-regulated ULK1 expression.