SH-SY5Y, N2a, HeLa, and RAW264.7 were cultured and maintained in Dubelcco's Modified Eagle Media (DMEM) (Gibcol) supplemented with 10% fetal bovine serum (FBS) and 1% Penicillin/Streptomycin (Hyclone). For treatment with vehicle (DMSO), R1881 or ARN-509 and MDV3100, both obtained from APExBio, TX, USA, cells were grown in phenol red-free DMEM (Gibco) medium supplemented with 10% (or indicated concentration, v/v) dextran-coated charcoal stripped FBS (Biological Industries, Israel) for 72 h prior to treatment. Cells were incubated at 37°C in a humidified incubator containing 5% CO₂. To inhibit AR expression, siRNAs targeting full length AR (siARa, 3′- AAGACGCUUCUACCAGCUCAC -5′; siARb, 3′- AAGAAGGCCAGUUGUAUGGAC -5′; siARc, 3′-GACCUACCGAGGAGCUUU-5′) or siRNA control ordered from GenePharma (Shanghai, China) were transfected using Lipo2000 transfection reagent (Invitrogen) in accordance with the manufacture's protocol.

Cell lysis, protein extraction, and immunoblotting were performed as described previously (Wang et al., 2011).

The proliferative effects of R1881 and the anti-proliferative effects of MDV3100 and ARN509 were measured in vitro by using MTT assay. Briefly, the cells were seeded into 96-well plates at a density of 4,000 cells/well. After treatment, the medium was replaced with fresh culture medium contained 0.5 mg/ml MTT reagent and incubated at 37°C for 4 h. Subsequently, the supernatant was aspirated, and cells were lysed in 200 μl DMSO for 10 min at 37°C. The optical density (OD) was measured at 490 and 570 nm using a plate reader. Each experiment was performed in triplicate.

SH-SY5Y and N2a cells were fixed with 4% paraformaldehyde in 1X PBS at 4°C for 30 min. The slides were incubated with rabbit anti-AR antibody (Santa Cruz) for 1 h, followed by 1 h incubation of TRITC anti-rabbit IgG antibody (Thermo) and stained with DAPI (for the nucleus). Samples were mounted using anti-fade mounting medium, and then visualized with a fluorescent light microscope.

SH-SY5Y and N2a cells were cultured in six-well plates (5 × 10⁵cells/well) and incubated until they reached 90–100% confluence. SH-SY5Y and N2a cells were then maintained in phenol red-free DMEM with 2.5% cFBS (charcoal stripped FBS) or 5% cFBS, respectively, in order to minimize the cell proliferation. A sterile 20 μl tip was used to create scratch wounds of the same width on each monolayer. The plates were then washed twice with phosphate-buffered saline (PBS) to remove the detached cells. Photos were taken at 0, 24, and 48 h, and the distance traveled by the cells enumerated the closure of the wounds. Each experiment was performed in triplicate.

The cells were seeded in the top chamber of Matrigel™-coated inserts (pore size: 8 μm; Falcon) placed in 24-well plates (2 × 10⁴cells/well for SH-SY5Y), while a medium supplemented with 10% cFBS (charcoal stripped FBS) was used as a chemo-attractant in the lower chamber. The wells were coated with 100 μl of Matrigel™ (BD Bioscience) at a dilution of 1:40 (Matrigel: serum-free medium) and air-dried overnight in a biosafety cabinet. The cells were allowed to invade through the Matrigel™ for 48 h at 37°C in a 5% CO₂ incubator. Cells that did not invade were scraped off with a cotton-tip applicator while the invading cells were fixed and stained with 0.005% crystal violet. The number of invading cells was counted under a light microscope (x10 objective) from three fields for each well. Each experiment was performed in triplicate.

Single N2a cell suspension was suspended in Matrigel™/serum free DMEM (1:1) at a concentration of 2 × 10³cells/well or 4 × 10³ in a total volume of 50 μl. The solution was then plated gently individual wells of a 96-well plate and allowed to solidify for 1 h at 37°C. One-hundred microliter of phenol red-free DMEM with 5% cFBS was added gently to the each well and the media (containing the treatment) was changed every 2–3 days. The numbers and morphology of colonies were then counted and observe under a microscope. Each experiment was performed in triplicate.

Warm (37°C, 500 μL per well) base agar solution (1.5% agar) in 1 × DMEM complete culture media) was poured into each well of 24-well plate. The base layer was allowed to solidify at 4°C for 30 min. Then 500 μL of a warm (37°C) top agar solution consisting of 8,000 SH-SY5Ycells in 0.7% agar + 1 × DMEM complete tissue culture media was added over the base layer. The 24-well plates were then incubated at 37°C in the humidified incubator for 3 weeks; 500 μL of fresh media was added every 3 days without disturbing the cells. After 3 weeks, the colonies were stained with 0.005% crystal violet. The numbers and morphology of colonies were then counted and observe under a microscope. Each experiment was performed in triplicate.

To evaluate the efficacy of androgen promoting neuroblastoma cells growing in vivo, 10⁶ N2a cells in 10 μl of PBS were injected into the exposed adrenal glands of immunocompromised mice. Five days prior to N2a cells inoculation, 10 out of 15 of the male mice were castrated to eliminate the androgen impact of the testes. Meanwhile 5 intact male mice were of sham surgery. Fifteen male and 12 intact female 4-week-old mice were anesthetized, the left flanks were prepared in sterile fashion, and transverse incisions were performed to expose the left kidneys and the adrenal glands. The mice were then intraperitoneally injected with testosterone propionate (TP, 0.5 mg/kg body weight) dissolved in sesame oil (vehicle) or merely the vehicle. Thus, all the mice were divided into 5 groups: castrated male TP group (n = 5), intact male vehicle group (n = 5), castrated male vehicle group (n = 5), intact female TP group (n = 6) and intact female vehicle group (n = 6). Each group was treated every 3 days for 2 weeks and then sacrificed. To evaluate the effect of androgen on the survival time of N2a bearing mice, 13 castrated male mice were injected with 2 × 10⁶ N2a cells into the adrenal glands. One week after the inoculation, the mice were randomly separated into two groups and then injected every 3 days intraperitoneally with testosterone propionate (n = 6) or the vehicle (n = 7), respectively. The cumulative survival curve was generated by SPSS 17.0.

All mice experiments were carried out with ethical committee approval and met the standards required by the Dalian Medical University Animal Care and Use Committee guidelines.

Statistical analysis was performed using GraphPad Prism version 5. The significance of the data was analyzed using a Student's t-test, and p-values of p < 0.05 (^*), p < 0.01 (^**) and p < 0.001 (^***) were considered significant and highly significant, respectively.

To determine the expression level of AR protein in neuroblastoma and other cell lines, neuroblastoma cell lines (SH-SY5Y and N2a), human cervical cancer cell line HeLa and mouse leukaemic monocyte macrophage cell line Raw264.7 were analyzed with Western blotting assay. Neuroblastoma cell lines exhibited high level of AR expression while no AR expression was detected in HeLa or Raw264.7 cells (Figure 1A). Western blotting result also showed that siRNA(a) targeting full length AR significantly reduced the AR expression in SH-SY5Y and N2a cells (the siARa target sequences are similar in both human and mouse, Figures 1B,C). Thus, siARa was applied for knockdown of AR in the following experiments.

Androgen-induced AR expression was also analyzed. AR expression and proliferative effect of R1881, a synthetic androgen as the AR agonist, in both cell lines were in a dose-dependent manner (Figures 1D,E). Furthermore, translocation of AR from cytoplasm to nucleus upon ligand-binding can be reduced by AR antagonist (Figure 1F). Intriguingly, with vehicle treatment, AR expressed in both the cytoplasm and nucleus suggesting the constitutive AR activity in neuroblastoma cells.

We first studied the in vitro effect of AR signaling on the cell proliferation of SH-SY5Y and N2A cells lines via MTT assays. The neuroblastoma cells were treated with R1881 (10 nM), a synthetic agonist of AR, or the antagonists of AR, ARN509 (10 μm), and MDV3100 (10 μm) for 72 h. The results showed that R1881 promoted the proliferation of SH-SY5Y and N2a, while the antagonists of AR dramatically inhibit the proliferation of the two cell lines (Figures 2A,B). On the other hand, there was no significant difference of the proliferative activity when AR knockdown cells treated with AR agonist or antagonists in the two neuroblastoma cell lines (Figures 2C,D). Besides, inhibition of AR expression with siRNA led to decreased cell proliferation in both cell lines.

Vehicle (DMSO)-treated cells grew well, but slow, that suggests AR activity is important but not the unique factor to neuroblastoma cells' growth. Besides, it suggested neuroblastoma cells produce hormones to induce the growth of their own, which explained the nuclear expression of AR with vehicle treatment (Figure 1F) as well as administration of AR antagonists can achieve better growth inhibition compared with the vehicle (Figures 2A,B).

A wound-healing assay was used to evaluate the effect of R1881 and the AR antagonists, ARN509, and MDV3100, on migration of both cell lines. R1881 obviously promoted cell migration and the cells were able to almost close the wound at 48 h, while both ARN509 and MDV3100 significantly inhibited the migration of the treated cells (Figures 3A,B). However, the migratory ability of neuroblastoma cells was largely reduced by siARa (Figure 3A vs. Figure 3C and Figure 3B vs. Figure 3D, in vehicle and R1881-treated groups). The result also demonstrated that AR knockdown attenuated the agonist effect of R1881 and the antagonist effect of MDV3100 and ARN509 (Figures 3C,D). These showed R1881 promote the migratory ability of both neuroblastoma cell lines while AR knockdown, ARN509 and MDV3100 suppress the migratory ability.

Next, we use Matrigel™-coated trans-well experiments to detect cell invasion. R1881 obviously increased the amount of invasive cells compared with DMSO or AR antagonists treatment (Figures 4A upper panel, B). Interestingly, depletion of AR by siRNA attenuated the effect of either R1881 or AR antagonist (Figures 4A lower panel, B). These results indicated that the invasive abilities of SH-SY5Y cell line were closely related to AR signaling activation. The assay was not performed on N2a cells since they failed to show an ability to invade when using FBS as a chemo-attractant.

In order to better visualize the sphere-forming capabilities of neuroblastoma cells, single cell suspensions of SH-SY5Y and N2a were cultured in soft agar and Matrigel™ for 21 and 7 days, respectively. The spheres were then visualized under an inverted light microscope and bright field images were acquired. Both cell lines treated with R1881 produced more and relatively larger spheres, while cells treated with AR antagonists produced significantly smaller spheres, compared to those cells treated with vehicle (Figures 5A,C, upper panles). Inhibition of AR expression by siRNA obviously blocked the promoting or suppressing effect of R1881 or AR antagonists, respectively on sphere formation of both cell lines (Figures 5B,D).

To further evaluate the efficacy of androgen on cancer cells growth in vivo, N2a neuroblastoma cells were injected into the adrenal glands of nude mice, and mice were treated with either vehicle alone or testosterone propionate dissolved in the vehicle. Testosterone propionate treatment resulted in increased tumor volume in both male and female mice (Figures 6A,B) as well as shorter survival time in male mice (Figure 6C). These results indicated the proliferative effect of androgen in N2a xenograft models.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer JH and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.