SHSY5Y cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA, United States) and was validated by the purchaser cell bank. Cells were grown as monolayer in Eagle’s Minimum Essential Medium (α-MEM) plus HAM’s F12 (1:1), supplemented with 10% heat-inactivated Fetal Bovine Serum (FBS), L-glutamine (2 mM), penicillin (100 IU/ml) and streptomycin (100 μg/ml) and cell cultures were maintained at 37°C, in a humidified atmosphere of 5% CO₂. For passaging, cells were detached from culture flasks with 0.05% trypsin and 0.002% EDTA solution. All media and supplements for cell cultures were acquired from Hyclone (Logan, UT, United States).

For evaluating the responsiveness to Wnt signaling-affecting molecules, exponentially growing SHSY5Y (SHSY5Ywt) cells were seeded at densities ranging from 4 × 10⁴/cm² and 6 × 10⁴/cm² and cultured for 24 h prior to treatments. Cells were treated with 100 nM ANP (PeptaNova, Sandhausen, Germany), a concentration selected as the lowest effective and not toxic dose in preliminary dose-response experiments (Supplementary Figure S1), for times ranging from 3 to 24 h. As positive control of Wnt signaling activation, the treatment with Wnt1a 100 ng/ml (ENZO life Science) was used.

The differentiated phenotype (SHSY5Ydiff) was obtained based on protocols previously reported by other Authors (Lopes et al., 2010, 2017). In detail, SHSY5Y cells were seeded at density of 4 × 10⁴/cm² and maintained in culture for 24 h to obtain monolayer at about 75% confluence. Cells were then treated with 10 μM retinoic acid (RA) in absolute ethanol and maintained in low serum medium (1% FBS) for a total of 9 days. To replenish RA, the differentiating agent was added at day 2, day 5, and day 8 after seeding, by replacing cell culture medium with new fresh medium without detaching the cell monolayer, and at day 10 the differentiated cultures were used for the experiments or analyzed. ANP treatment on SHSY5Ydiff was performed as described above for SHSY5Ywt cells.

For mimicking the neurodegeneration of PD, SHSY5Ywt and SHSY5Ydiff cells were exposed for 24 h to 50 and 100 μM of 6-OHDA (prepared in 0.1% ascorbic acid in DMSO), respectively. These concentrations have been selected as the lowest effective doses in preliminary dose-response experiments (Supplementary Figure S2), in which we tested the susceptibility of the two different cellular systems to the neurotoxin, by using concentration ranging from 25 to 400 μM. For assessing the neuroprotective ability of the natriuretic peptide, SHSY5Ywt and SHSY5Ydiff cells were pre-incubated with 100 nM ANP 30 min or 24 h prior to the addition of 6-OHDA to cell cultures, and analyzed after additional 24 h for morphological changes, and cell viability, and for the expression levels of β-catenin, DA neuron specific markers and survival factors.

Cell morphology were analyzed by phase-contrast microscopy, using the Motic AE31 Trinocular inverted microscope (Motic Asia, Hong Kong). Cell viability was evaluated by Trypan blue dye exclusion method. Mitotic index (MI) evaluation was carried out as previously described (Serafino et al., 2012). Quantitative evaluation of MI was done in a blinded fashion under the LEICA TCS SP5 Confocal Laser Scanning Microscopy (CLSM, Leica Instruments, Mannheim, Germany), by counting a minimum of 300 cells/sample. Results were reported as percentage of cells in mitosis. Cell viability and proliferation analyses were performed on samples in triplicate and results were reported as mean values ± SD. Cell cycle assessment was carried out by cytofluorimetric analysis of DNA content after propidium iodide (PI) staining. Total DNA content was measured using the FACSCalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ, United States)

For confocal microscopic analyses, SHSY5Ywt or SHSY5Ydiff cells were grown or RA-differentiated on the ibiTreat μ-Slide 4 well (Ibidi GmbH, Germany, cod. 80426). Immunocytochemical analyses were performed on cells fixed with 2% paraformaldehyde (Sigma–Aldrich). After permeabilization with 0.2% Triton X-100 (Sigma–Aldrich), immunofluorescence staining was done using the primary antibodies against β-catenin, Nestin, Tubulin-β3, NeuN and TH, described in Table 1. The specificity of each antibody has been preliminary verified by Western blot, by considering the production of a single band matched to the proper molecular weight. Primary antibodies were revealed with Alexa Fluor 488-conjugated anti-mouse or anti-rabbit IgG (Molecular Probes). Samples were analyzed by using the LEICA TCS SP5 confocal microscope.

After washing with ice-cold PBS, cells were lysed using a 50 mM Tris-HCl buffer at pH 8.0. Specifically, the buffer solution contained 150 mM of NaCl, 1% of NP-40, 10% of glycerol, 0.1 mM of EGTA, 0.5 mM of EDTA, 50 mM of NaF, 1 mM of Na3OV4, and a protease inhibitor cocktail (Sigma–Aldrich). Lysates were clarified by centrifugation and proteins were quantified using the Bradford reagent (Bio-Rad, Segrate, Italy). 15–20 μg of each cell extract was separated using SDS/PAGE ranging from 8 to 12%, and then transferred to nitrocellulose membrane (Hybond, Amersham GE Healthcare). Membranes were incubated for 1 h at room temperature with 5% BSA in Tris-buffered saline-Tween (TBS-T; 0.2 M Tris, 1.37 M NaCl, pH 7.6, and 0,05% Tween-20), probed with the specific antibodies reported in Table 1. Primary antibodies were revealed with peroxidase-conjugated secondary antibody (BioRad, Richmond, CA, United States). Densitometric analysis was done using the ImageJ processing program¹. Values, normalized to β-Actin or GAPDH, were reported as uncalibrated Optical Density (OD) or as fold vs untreated control. Data were from at least three independent experiments and presented as the mean ± SD.

Transcription levels of β-catenin (β-cat), LEF1 and TH genes were assessed by qPCR in untreated controls and in SHSY5Ywt cells challenged with 100 nM ANP for times ranging from 3 to 24 h. Total RNA was extracted using TRIzolTM Reagent (Invitrogen, Carlsbad, CA, United States), and treated with RNase-free DNase (Promega), according to the manufacturer’s instructions. After extraction with a mix of Phenol:Chloroform:Isoamyl Alcohol (25:24:1; Invitrogen) and precipitation with ethanol, 1 μg of total RNA was reverse-transcribed using random primers with ImProm II Reverse Transcription system (Promega) according to the instructions of the manufacturer. qPCR was performed with iTaq Universal SYBR Green Supermix (Bio-Rad) using 250 nM of the specific primers, by the CFX Connect Real-Time PCR Detection System (Bio-Rad). Sequences of primers used are reported in Table 2. The amplification conditions were: 1 min at 95°C, 40 cycles of 10 s at 95°C and 40 s at 60°C. Cq values were determined from the system software using ‘single threshold’ mode. The relative expression level for each gene were calculated from these Cqs using experimentally determined amplification efficiencies, and then normalized for the reference gene GAPDH. Results were reported as fold of induction vs untreated control. In all experiments each sample was analyzed in triplicate, and no-template controls and no-reverse transcription controls were inserted.

Statistical analysis was performed using the two-tailed Student’s t-test, and p < 0.05 was considered as statistically significant. All data were from at least three independent experiments and presented as means ± SD.

SHSY5Y cells in basal condition (SHSY5Ywt) have been characterized for the expression and intracellular distribution not only of neuronal markers (Nestin, Tubulin-β3, NeuN) and of the DA neuron specific markers TH and Nurr1, but also of two crucial molecules participating to the Wnt/β-catenin signaling, the Fzd1 receptor and β-catenin (Figure 1). Confocal microscopy (Figure 1A) and Western blot analysis (Figure 1B) showed that SHSY5Ywt cells expressed Tubulin-β3 and NeuN (specific for mature neurons) as well as TH and Nurr1 (specific for DA neurons), together with detectable levels of Nestin (neural Stem/Progenitor cell marker). This confirms that SHSY5Ywt cells possess the features of dopaminergic neurons and are appropriate for studying neurotoxicity or neuroprotection in experimental PD, as also reported by others (Levites et al., 2003; Cheung et al., 2009; Lopes et al., 2010, 2017; Xie et al., 2010). Importantly, Fzd1 receptor and β-catenin are also clearly expressed (Figure 1B), with β-catenin that localized almost exclusively at the cell membrane (Figure 1A, upper panels), and this indicated that SHSY5Ywt cells were potentially responsive to molecules affecting the Wnt pathway.

SHSY5Ydiff cells were obtained as described in Section “Materials and Methods” and schematized in Figure 2A. Even if proliferating and RA-differentiated SHSY5Y cells have long been used for studies in neuroscience as models of DA neurons, data reported in literature about which of the two systems is the most suitable model for investigating the molecular and cellular mechanisms underlying the pathophysiology of PD are often inconsistent (Cheung et al., 2009; Lopes et al., 2010, 2017; Xie et al., 2010). For this reason, both cellular systems have been preliminarily characterized and compared for their morphological and phenotypical features (Figures 2, 3). RA treatment for a total of 9 days gave rise to cells exhibiting neuronal differentiation with neurite outgrowth (Figure 2B), low proliferation rate, as demonstrated by lower levels of c-Myc (Figure 2C), and block in G1 phase, as demonstrated by cytofluorimetric analysis (Figure 2D) and by higher levels of p21 (Figure 2C), than those recorded in the untreated SHSY5Y cells. SHSY5Ydiff cells were then compared with SHSY5Ywt for the neuronal markers (Tubulin-β3, NeuN), DA neuron specific markers (TH, Nurr1), neuronal survival factors involved in neuroprotection against oxidative stress (DJ-1 and pAkt^T308, the phosphorylated active form of AKT), and Wnt pathway related molecules (Fzd1 and β-catenin) (Figure 3). Confocal microscopy (Figure 3A) and Western blot analysis (Figures 3B–D) showed that SHSY5Ydiff cells, expressed higher levels of Tubulin-β3, NeuN, TH and Nurr1 (1.32-fold, 1.23-fold, 1.25-fold, and 1.9-fold of increment, respectively), compared with SHSY5Ywt, indicating that the RA-differentiated cellular system was phenotypically more cognate to mature dopaminergic neurons than the not differentiated one. Moreover, expression levels of the survival factors DJ-1 and active (phosphorylated) AKT were also considerably increased in SHSY5Ydiff as compared with SHSY5Ywt (2.5-fold and 3.7-fold of increment, respectively) (Figure 3C). In both cellular systems β-catenin localized almost exclusively at the cell membrane (Figure 3A), while Fzd1 receptor was expressed at higher levels (Figure 3D) in SHSY5Ydiff compared to SHSY5Ywt (1.6-fold of increment). This indicated that SHSY5Ydiff cells could be potentially responsive to molecules affecting the Wnt pathway, similarly to SHSY5Ywt.

In order to verify whether the treatment with the natriuretic peptide affected the Wnt signaling also in these cellular systems as previously demonstrated for other cells (Serafino et al., 2012; Skelton et al., 2013a,b; Vesely, 2013), SHSY5Ywt neuroblastoma cells were treated for 24 h with ANP in a preliminary dose-response experiment (Supplementary Figure S1), for selecting the lowest effective and not toxic dose to be used. β-catenin nuclear translocation and the concomitant increase in its expression have been used as main markers of Wnt pathway activation. The concentration of 100 nM ANP has been selected as the lowest dose inducing nuclear β-catenin translocation (Supplementary Figure S1A), cell proliferation arrest with a decrement of mitotic index (MI, Supplementary Figure S1B), increased levels of total β-catenin and a concomitant decrease on β-catenin phosphorylation at T41/S45 (decreased β-catenin degradation; Supplementary Figure S1C). The effects of 100 nM ANP on β-catenin intracellular distribution and degradation were then analyzed in time course experiments and compared with those of the Fzd1 ligand Wnt1a, that specifically triggers the Wnt pathway and has been demonstrated to exert protective effects on SHSY5Y cells (Wei et al., 2013). Hence, 100 ng/ml Wnt1a [concentration utilized by other Authors on SHSY5Y cells (Wei et al., 2013)] was used as positive control of Wnt signaling activation.

In SHSY5Ywt cells, WB analysis (Figures 4A,B) showed that ANP, similarly to Wnt1a, induces a significant increase of total β-catenin (p = 0.0098 vs untreated control), already after 3 h of treatment, and a concomitant decrease of β-catenin phosphorylation at Thr⁴¹ (pβ-catenin ^Thr41), one of the molecular event upstream β-catenin ubiquitination and degradation by the proteasome. Moreover, the decrease of pβ-catenin correlated to an increase of GSK phosphorylation at Ser9 (pGSK-3β^Ser9), the inactive form of this enzyme (Stambolic and Woodgett, 1994) deputed to β-catenin phosphorylation, as also recorded in Wnt1a treated cells. CLSM observation (Figure 4C) showed that ANP also caused a redistribution of β-catenin from cell-cell junction sites to cytoplasmic and nuclear compartments, between 6 and 24 h of treatment. Taken together, these results indicate that, in SHSY5Ywt cells, ANP positively affects the Wnt signaling by inducing β-catenin stabilization and nuclear translocation. This is further supported by the up-regulation of the Wnt-signaling dependent transcriptional factor LEF 1, as measured by RT-qPCR analysis after 24 h of ANP treatment, when a significant increase of transcription of the Nurr1-regulated gene TH was also recorded (Figure 4D). Conversely, at 3 and 6 h of ANP treatment, β-catenin mRNA resulted unmodified or down-regulated (Figure 4D), indicating that the high levels of the total protein concomitantly recorded by WB (Figures 4A,B) were actually due to a reduction of β-catenin degradation induced by the natriuretic peptide, rather than to a decrease transcription of β-catenin gene.

SHSY5Ydiff cells were also responsive to treatment with 100 nM ANP, as demonstrate by the β-catenin cytoplasmic accumulation and nuclear translocation observed between 6 and 24 h of treatment, similarly to what observed under treatment with the Fzd1 ligand Wnt1a (Figure 5).

In order to establish the optimal concentration of 6-OHDA to be used in the experiments of neuroprotection, we examined, in a preliminary dose-response test, the susceptibility of both SHSY5Ywt and SHSY5Ydiff to the neurotoxin (Supplementary Figure S2). Proliferative and RA-differentiated SHSY5Y cells were exposed to increasing concentration of 6-OHDA (25, 50, 100, 200, and 400 μM) for 24 h and cytotoxicity was assessed by evaluating cell survival by optical microscopy (Supplementary Figure S2A), cell viability assay (Trypan blue dye exclusion method; (Supplementary Figure S2B) and cell cycle analysis (Supplementary Figure S2C). The concentrations of 50 and 100 μM resulted the lowest effective doses in SHSY5Ywt and SHSY5Ydiff, respectively, that induced detachment of about 50% of cells from the adhering monolayer (Supplementary Figure S2A, red arrows), reduced cell viability by 50–60% (Supplementary Figure S2B) and produced the highest number of apoptotic/necrotic cells by flow cytometry (Supplementary Figure S2A, right panel). The higher dose of neurotoxin selected for SHSY5Ydiff vs SHSY5Ywt cells in the dose-response experiments, demonstrated that the RA-differentiated cells obtained by our experimental conditions, were more resistant to 6-OHDA than the not differentiated phenotype. This could be in part due to the higher levels of the survival factors DJ-1 and active AKT recorded by WB in SHSY5Ydiff compared to SHSY5Ywt (Figure 3C).

Based on results from the preliminary dose-response test, SHSY5Ywt and SHSY5Ydiff were exposed to 50 and 100 μM of 6-OHDA, for times up to 24 h. To verify the ability of ANP in preventing the cytotoxicity induced by 6-OHDA, cells were pre-treated with 100 nM of the natriuretic peptide, 30 min or 24 h before 6-OHDA addition to cell culture medium. Phase contrast microscopy (Figures 6A,B) showed that 6-OHDA treatment drastically decreased the number of adhering cells and destroyed the neurite network in both proliferating and differentiated SHSY5Y cells, while exposure to ANP did not influenced cell adhesion and viability nor neurite integrity. Pre-treatments with 100 nM ANP were able to substantially prevent detachment of cells from the substrate, also partially preserving the neuritic network (Figures 6A,B), and significantly (p < 0.01) reduced the percentage of dead cells induced by 6-OHDA in both cellular models (Figures 6C,D).

In SHSY5Ywt cells, WB analysis (Figure 7) showed that the challenging with 6-OHDA for 24 h dramatically decreased total β-catenin expression (p < 0.001), compared to both untreated control and ANP-treated cells (2.8-fold and 4-fold of decrement, respectively), and concomitantly increased β-catenin phosphorylation and degradation (2.2-fold and 3-fold of increment vs untreated control and ANP treated cells, respectively). Moreover, 6-OHDA significantly (p < 0.01) reduced the expression levels of the neuronal marker Tubulin-β3 and of the DA neuron specific markers TH and Nurr1 (2.4-fold, 2.6-fold, and 1.8-fold of decrement, respectively) as well as of the survival factors involved in neuroprotection against oxidative stress AKT/phospo-AKT and DJ-1 (2-fold/1.5-fold, and 1.7-fold of decrement, respectively). Pre-treatment with ANP, by inducing β-catenin stabilization, partially or completely restored the expression of all markers analyzed to levels near to those recorded in the untreated control at T0 (Figure 7 and Table 3). It is noteworthy that 24 h of treatment with ANP alone significantly up-regulated the expression levels of the survival and anti-apoptotic factor DJ-1 (Figure 7), the product of the PARK3 gene, that has a neuroprotective role against oxidative stress and is dysfunctional or low expressed in PD patients (Bonifati et al., 2003; Canet-Aviles et al., 2004; Martinat et al., 2004; Taira et al., 2004).

Similar results were obtained when we used as model of DA neurons the SHSY5Ydiff cells, where the effects induced on the molecular markers were assessed at 7 and 24 h of 6-OHDA treatment (Figures 8A–D). In particular, already after 7 h of treatment, the neurotoxin was able to significantly (p < 0.001) decrease total β-catenin expression, and concomitantly increase β-catenin phosphorylation and degradation (0.7-fold and 1.24-fold of change vs untreated control, respectively) (Figure 8B), also reducing, between 7 h and 24 h, the DA neuron specific markers TH and Nurr-1 (Figure 8D), while the survival factor and PD-related protein DJ-1 was not significantly modified. Differently to what recorded the proliferating model, in RA-differentiated cells the neurotoxin induces a significant increase in pAKT levels at both times examined (increment: 1.75-fold and 1.77-fold vs untreated control, at 7 and 24 h, respectively) (Figure 8C), and this could additionally explain the higher resistance to the neurotoxin-induced stress exhibited by the SHSY5Ydiff compared to SHSY5Ywt. Nevertheless, also in SHSY5Ydiff pre-treatment with ANP partially or completely reversed the neurotoxin-induced changes of the markers analyzed already after 7 h of treatment (Figure 8 and Table 3). In particular, at this time point, in ANP pre-treated samples we recorded a significant (p < 0.001) increase in total β-catenin and a concomitant decrease in β-catenin phosphorylation and degradation, compared to cells treated with 6-OHDA alone (Figure 8B). A significant (p < 0.01) increase of Nurr-1 expression was also recorded at 24 h in ANP pre-treated samples. At 24 h, TH, which expression is regulated by Nurr-1 (Sakurada et al., 1999), was also up-regulated in ANP pre-treated samples vs 6-OHDA alone, even if the mean values from three independent experiments were not significant (p > 0.05; Figure 8D). Moreover, in samples pre-treated with ANP for 24 h and subjected to the neurotoxin challenging, we recorded an additional increment in AKT phosphorylation levels compared with 6-OHDA alone, and this further sustains the neuroprotective role exerted by the natriuretic peptide (Figure 8C). The neuroprotective efficacy of the natriuretic peptide was also recorded in both SHSY5Ywt and SHSY5Ydiff models subjected to pre-treatment with ANP for 30 min, and this suggests that the molecular events that reinforce cell survival are triggered soon after its addition to culture medium (Figures 6–8 and Table 3).