New therapeutic strategies in neuroblastoma: combined targeting of a novel tyrosine kinase inhibitor and liposomal siRNAs against ALK

Many different aberrations in the Anaplastic Lymphoma Kinase (ALK) were found to be oncogenic drivers in several cancers including neuroblastoma (NB), therefore ALK is now considered a critical player in NB oncogenesis and a promising therapeutic target. The ALK-inhibitor crizotinib has a limited activity against the various ALK mutations identified in NB patients. We tested: the activity of the novel ALK-inhibitor X-396 administered alone or in combination with Targeted Liposomes carrying ALK-siRNAs (TL[ALK-siRNA]) that are active irrespective of ALK gene mutational status; the pharmacokinetic profiles and the biodistribution of X-396; the efficacy of X-396 versus crizotinib treatment in NB xenografts; whether the combination of X-396 with the TL[ALK-siRNA] could promote long-term survival in NB mouse models. X-396 revealed good bioavailability, moderate half-life, high mean plasma and tumor concentrations. X-396 was more effective than crizotinib in inhibiting in vitro cell proliferation of NB cells and in reducing tumor volume in subcutaneous NB models in a dose-dependent manner. In orthotopic NB xenografts, X-396 significantly increased life span independently of the ALK mutation status. In combination studies, all effects were significantly improved in the mice treated with TL[ALK-siRNA] and X-396 compared to mice receiving the single agents. Our findings provide a rational basis to design innovative molecular-based treatment combinations for clinical application in ALK-driven NB tumors.


INTRODUCTION
Neuroblastoma (NB) is an embryonic tumor of the sympathetic nervous system, and is the most commonly solid malignancy diagnosed in the first year of life, accounting for about 9-10% of paediatric cancer mortality. NB is characterized by broad clinical phenotypes from spontaneous regression to fatal outcome in advanced stages with metastatic disease. Despite aggressive and multi-modal therapies (i.e. aggressive chemotherapy, surgery, radiation therapy, stem cell transplantation, immunotherapy) the effective treatment of advanced NB is still a challenge for paediatric oncology [1,2].
Over the years, many genes have been encountered to have potential roles in NB pathogenesis, but the NB causing genes remained unidentified. The identification of germline and somatic activating mutations in the tyrosine kinase domain of the Anaplastic Lymphoma Kinase (ALK) gene as well as amplification, rearrangments and/or over expression of either mutated or wild-type ALK allelesrevealed that such aberrations are oncogenic drivers in NB and correlate with worse patients' outcome or unfavourable aggressive NB phenotype [3][4][5][6][7]. Notably, as both ALK and MYCN oncogene are located in close proximity to each other on chromosome 2p, amplification of MYCN can also involve amplification of the ALK locus [3,4,7,8]. When expressed together, ALK and MYCN promote NIH3T3 cell transformation and both wild-type and activated mutant forms of ALK stimulate transcription of MYCN [9].
Therefore, ALK has emerged as a critical player in NB and tractable 'oncogene' opening new perspectives into the design and development of gene-targeted therapies for NB and other ALK-driven cancers [8,10,11], such as Non-Small-Cell Lung Cancers (NSCLC) [12] and Anaplastic Large-Cell Lymphomas (ALCL) [13] characterized by oncogenic ALK fusion proteins arising from chromosomal translocations.
The success of various tyrosine kinase inhibitors in the treatment of different cancers as well as the increased number of human malignancies involving aberrant ALK activity, both in children and adults, encouraged the search and motivated the development of ALK-selective small-molecule inhibitors. Crizotinib, the only ALK-inhibitor approved by Food and Drug Administration (FDA, USA), showed limited activity against ALK-driven NB [14]. The mutated variants also showed differential in vitro crizotinib sensitivities [8].
The most aggressive ALK-mutation in NB, ALK F1174L , has been described as a relapse-specific mutation correlating with unresponsiveness to therapy [15]. ALK F1174L also arises as a secondary event after an initial response to crizotinib [16] and is able to potentiate the oncogene activity of MYCN [17]. To overcome acquired resistance to crizotinib, different categories of next-generation ALK-inhibitors have been developed and are subject of various pre-clinical studies or earlyphase trials [18]. However, there is no certainty that these compounds will be more potent than crizotinib against ALK-mutants. In this view, Xcovery (West Palm Beach, FL) has developed a novel and powerful small molecule ALK-inhibitor named X-396 (http://www.xcovery.com). The potency and the selectivity of X-396 were validated by comparison with other ALK-inhibitors available and tested in animal models of Non-Small-Cell Lung Cancer (NSCLC). Noteworthy, X-396 is active against multiple ALK variants founds in NSCLC, including ALK mutations associated with acquired resistance to crizotinib [10]. Indeed, in June 2012, X-396 entered a Phase 1 safety trial in patients with solid tumors [11] (for details see NCT01625234 at http://www.clinicaltrials. gov). Preliminary clinical data have shown that X-396 is generally well-tolerated and has anti-tumor activity in patients with NSCLC bearing an ALK fusion protein [19].
Based on these findings, we hypothesize that X-396 could be more effective than crizotinib on NB cells bearing in either of the two more common ALK-mutations, ALK F1174L or ALK R1275Q [20]. In addition, we believe that novel combination strategies that increase the efficacy of ALK-inhibitors like X-396 without increasing toxicities in children are potentially relevant.
To this regard, we have successfully developed and validated by in vitro and in vivo studies a RNAi-mediated therapeutic approach to selectively knockdown ALK expression by using NB targeted nanoliposomes [21,22]. Since our formulation is a safe and powerful siRNA-based therapeutic tool for NB, we thought it may be ideal to combine with an ALK kinase inhibitor. Here we present results aimed at testing whether a combined therapeutic approach using the novel inhibitor X-396 working on ALK at protein level, and the NB targeted liposomal siRNAs against ALK working at mRNA level, could represent an improved strategy with additive and/or synergistic effects to promote long-term survival in NB xenografts.

X-396 is a kinase inhibitor with higher potency against ALK-mutated neuroblastoma-cell lines than crizotinib in vitro
Pre-clinical studies have shown that NB cell lines harbouring the F1174L mutation, the second most common ALK-mutation seen in NB tumors, are significantly more resistant to crizotinib than those harbouring the most common mutation, R1275Q [16,20]. For this reason, we first measured 3 H-thymidine uptake to assess the dose-dependent effect of X-396 and crizotinib on the growth of two NB cell lines, each one harbouring one of the hot spot-activating ALK-mutations: LAN-5 carrying R1275Q ALK mutation (ALK R1275Q ) and SH-SY5Y carrying F1174L ALK mutation (ALK F1174L ).
Although the slope of the curves between X-396 and crizotinib are similar, the differences between the elevations were significantly different in both cell lines. Specifically, in LAN-5 cells X-396 was 3.7-fold more potent than crizotinib (IC 50 : crizotinib 328 nM, X-396 88 nM) to decease cell growth ( Figure 1A).   A. and SH-SY5Y B. cells were seeded in 96-well plates. The day after, cells were treated with various concentrations (1-2000 nM) of crizotinib or X-396. Results, derived from three different experiments, are expressed as mean percentage of 3 H-thymidine incorporation 72 hours after initiation of treatment as compared to that of control cells (culture medium containing 0.01% dimethylsulfoxide). Error bars represent 95% confidence interval. Inset, indicate the concentration of drugs causing 50% inhibition of cell proliferation (IC 50 ) for each cell line, evaluated by nonlinear regression (curve fit). LAN-5 C. and SH-SY5Y D. cells viability was measured by AlamarBlue staining 72 hours after treatment with 10-100-1000 nM of X-396 or crizotinib. Bars represent the cell viability, derived from three different experiments, as mean percentage of sextuplicate wells, considering the control level of cells (treated as in panels A and B) to be 100%. Error bars represent 95% confidence interval. P value (two-tailed) were calculated using the Student's t test with Welch's correction. *P < 0.05, **P < 0.01, ***P < 0.001. LAN-5 E. and SH-SY5Y F. cells were treated with various concentrations (20-1000 nM) of crizotinib or X-396 or 0.01% DMSO (Ctr) for 72 hours. Lysates were subjected to immunoblotting with the specific antibodies.
We next examined the activity of X-396 on the cell viability of cultured NB cells by AlamarBlue staining. Treatment with X-396 induced a statistically significant dose-dependent decrease in cell viability compared with the same dose of crizotinib ( Figure 1C, 1D).
To confirm the target specificity of X-396, we assessed the ability of the compound to reduce the endogenous ALK phosphorylation in SH-SY5Y and LAN-5 NB cells. Compared to crizotinib, X-396 inhibited ALK phosphorylation at lower concentrations of drug ( Figure  1E, 1F and Supplementary Figure S1).
The above results indicated that X-396 is an ALKinhibitor more powerful and gene-specific than crizotinib in accordance with previous findings obtained by Lovly et al [10].

Pharmacokinetic profile, biodistribution properties and anti-tumor activity of X-396 in subcutaneous NB mouse model
We next investigated the effects of X-396 in vivo, in SH-SY5Y xenografts. Pharmacokinetic profiles revealed that X-396, after multiple oral administrations by gavage in NB xenograft mice, showed substantial bioavailability, moderate half-life and slow absorption in vivo (time of the maximum concentration (T MAX : 2 h). At the low dose of 25 mg/kg, the mean plasma concentration 2 hours after the last dosing was 1284 ng/mL or about 2.3 μM which is 15x that of the IC 50 of inhibiting the SH-SY5Y cell growth, in vitro ( Figure 2A and Table 1).
The tissue concentration of X-396 was also examined in these mice. It clearly indicated that the X-396 concentrations were higher in the tumor site than in the plasma in each schedule of treatment performed, reaching 32 fold-ratio increase 8 h after the last treatment with 50 mg/kg BID (Table 2). Since the plasma and tumor concentrations of X-396 are well above the IC 50 of inhibiting cell growth, we expected that X-396 should be able to inhibit tumor growth.
Indeed, X-396 was able to significantly reduce the tumor volume in a subcutaneous SH-SY5Y NB model in a dose-dependent manner ( Figure 2B) and its efficacy was more pronounced than that obtained with crizotinib when administered at the same dose ( Figure 2C). These findings suggest that the same anti-tumor activity may be achieved by administration of lower doses of X-396 with respect to crizotinib, in keeping with results obtained in other animal models by Lovly et al [10].
The Maximum Tolerated Dose (MTD) was not identified with any given regimens and no signs of adverse effect were observed (i.e. dehydration, severe weight loss, Supplementary Figure S2) during the treatment.

In vivo anti-tumor activity of X-396 against human NB orthotopic xenografts
We next asked whether the above anti-tumor results could be recapitulated in a more clinically relevant mouse model. To this purpose, we explored the effects of X-396 in biologically relevant orthotopic mouse models [23], obtained by implanting of Luciferase-stably-transduced NB cells, SH-SY5Y-Luc and LAN-5-Luc, into the adrenal gland of nu/nu mice.
To avoid the possible stressful mice conditions, due to the repeated oral gavage in the same day, BID, we decided to administrate 50 mg/kg and 100 mg/kg of X-396 in two mice groups only once a day (QD), starting 7 days post cell implantation.
Treatments with X-396 did not revealed any sign of toxicities (i.e. severe weight loss, dehydration or abdominal dilatation). As clearly shown by in vivo BLI ( Figure 3A and 4A) capturing the Luciferase signal intensity at day 22, X-396 treatments were able to slow down the primary tumor growth. All data were confirmed and quantified at day 29 by the fold increase of photon counts over the time in the tumor region ( Figure 3B and 4B), indicating a significant dose-dependent anti-tumor activity of X-396 in both mouse models. In SH-SY5Y xenografts the lower dose of X-396 (50 mg/kg QD) did not significantly increase the mice life span compared to control group, while a significant improvement of the anti-tumor effect was obtained with the higher dose (100 mg/kg QD) ( Figure 3C). This is consistent with the high brain tissue concentration of X-396 when dosed at 100 mg/kg, where the mean is 138 ng/g 8 hours after dosing. This concentration (~246 nM) is 1.7x the IC 50 , while that at 50 mg/kg is only 34.6 ng/mL or ~61 nM which is below the IC 50 of 149 nM (Table 2). Noteworthy, either 50 mg/kg QD or 100 mg/kg QD doses of X-396 exerted a significant increased life span in LAN-5-bearing mice ( Figure 4C).

Combined anti-tumor activity of X-396 and NBtargeted nanoparticles entrapping siRNA against orthotopic and pseudometastatic human NB xenografts
To determine whether the silencing of ALK, independently of its mutational status, could improve the anti-tumor efficacy of the ALK-inhibitor X-396, we decided to introduce our well established NB-targeted nanoparticles entrapping siRNA for specific ALKknockdown [21,22] in the treatment schedule.
Seven days after orthotopic implantation of SH-SY5Y-Luc and LAN-5-Luc-transduced cells, mice were treated for 21 days with 50 mg/kg QD of X-396 or with 1 mg/Kg of targeted liposomes entrapped siRNA against ALK (TL[ALK-siRNA]) administered either individually or in combination (COMBO). As clearly summarized by the ROI values computed at day 22 with respect to the control group at T 0 , TL[ALK-siRNA] Results are expressed as mean plasma concentration of X-396 ± Standard Deviation (SD). (B) Tumors were measured at fixed times with a calliper, and volume calculated. Error bars ± SD. C) Comparison of X-396 and crizotinib administered at the same dose. NB-bearing mice were OG treated with 50 mg/kg BID of X-396 or crizotinib and tumor volume determinated over time. Error bars ± SD. The statistical significance of differential findings between experimental groups and controls was determined by one-way analysis of variance (ANOVA) with the Tukey's multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001  determined a statistically significant decrease of the signal in both cell lines ( Figure 5A and 5C). Noteworthy, the combination of X-396 plus TL[ALK-siRNA] treatment led to further significant reduction of ROI signals respect to the single agents. These results were supported by a significant increase of the mice life span, as showed in Figure 5B and 5D. Based on these encouraging results, we decided to verify whether the increase of X-396 dose to 100 mg/Kg QD was able to further ameliorate tumor growth in the mice. As shown in Figure 6 Noteworthy, the combination of X-396 plus TL[ALK-siRNA] significantly enhanced the anti-tumor activity of the single agents ( Figure 6A, 6B, 6C). The combined treatment revealed values of median survival equal to 51 and 98 days for SH-SY5Y and LAN-5 orthotopic implant, respectively, and 114.5 days for the pseudometastic model. implanted orthotopically into the left adrenal gland of immunodeficient mice. NB-bearing mice were randomly assigned to three groups and treated 7 days after the cells transplantation (considered T 0 ) via OG QD with X-396 at the indicated doses or vehicle alone (Control) for 21 days. Body weight and general physical status were recorded daily, and mice were sacrificed by cervical dislocation after the administration of xilezine when they showed signs of poor health. A. Lateral (cell implantation side) images of BLI intensity from five NB-bearing mice implanted with SH-SY5Y-Luc cells evaluated at day 22. The relative levels of bioluminescence are shown as a pseudocolor display, with red and violet representing the strongest and weakest photon fluxes, respectively. B. Quantification of the in vivo Luciferase signal of the tumor area expressed as fold increase of photon counts over time in the region of interest (ROI) computed at time 29 day respect to T 0 . Error bars represent 95% confidence intervals. P value (two-tailed) were calculated using ANOVA with the Tukey's multiple comparison test. *P < 0.05, **P < 0.01. C. Survival curves of nu/nu NB-bearing mice (n = 6 for each group) in response to treatments. Kaplan-Meyer method & Log Rank (Mantel-Cox) test was applied to build up and compare the survival curves between analyzed groups. www.impactjournals.com/oncotarget DISCUSSION Despite intensive multimodal treatments (i.e. combination of isotretinoin, myeloablative therapy with stem-cell rescue and immunotherapy), the outcome of high-risk NB-patients remain very poor and the 2-years progression-free survival was improved only by 20% [25]. Since a considerable cohort of children are refractory to conventional approaches and/or can face relapse [26], new improved treatment options are needed.
Many studies on familial and sporadic cases with advanced NB have shown either germline or somatic activating mutations in the ALK gene as well as amplification or rearrangements and/or over expression independently of its genetic status, providing evidence of its pivotal role in NB oncogenesis, growth and survival [3,15,27,28]. Therefore, patients presenting with tumors exhibiting ALK aberrations, may potentially benefit from ALK-targeted therapy.
The use of ALK kinase inhibitors have been shown to be promising in treatment of Non-Small-Cell Lung Cancer (NSCLC), Anaplastic Large Lymphoma (ALL) and other cancers with an ALK-driven component, and for these reasons novel small-molecule inhibitors suitable for clinical applications are in development or in early-phase trials [29][30][31].
Furthermore, since ALK expression in normal adult tissues is found in very low levels, selective ALK inhibitors would exhibit sufficiently wide therapeutic windows in patients with ALK-activated cancers [31]. Thus, ALK represents a bona fide target and its inhibition is not predicted to result in undesirable systemic side effects [32].
Crizotinib, the first small-molecule kinase inhibitor, timely approved by Food and Drug Administration for the treatment of NSCLC patients (for details see FDA-US website), has limited activity against the various ALK-mutations identified in children with NB [14]. As  observed with other kinase inhibitors, ALK-positive NSCLC patients eventually relapse on crizotinib. Several distinct mechanisms of crizotinib resistance have been identified but most of them as yet remain unknown and to be elucidated [33][34][35][36][37]. The crizotinib-resistance of NB cells with ALK F1174L mutation should be surmountable with very high doses of the drug and/or with new generation of inhibitors [20]. Importantly, ALK F1174L may secondarily arise as a mechanism of resistance after an initial response to crizotinib in patients with ALK-rearranged cancers [16].
Recently, Xcovery developed X-396, a secondgeneration highly specific ALK small molecule tyrosine kinase inhibitor. X-396 has proved to be more selective and up to 10 times more potent than crizotinib in NSCLC [10].
Preliminary clinical results showed that X-396 had efficacy in crizotinib resistant patients [19]. Accordingly, in the present study, X-396 proved to be more effective in inhibiting in vitro NB cell growth than crizotinib (IC 50 2.6-3.7 fold lower).
Importantly, we confirmed and extended the ability of X-396 to inhibit ALK phosphorylation in NB cell lines harbouring the two most common ALK mutations, F1174L and R1275Q [10]. Indeed, one of the most recent study undertaken in 1,596 diagnosed NB patients identified 8% mutations at three "hot spot" mutation sites that display  the highest oncogenic potential in NB cells and account for 85% of mutations (R1275 43%, F1174 30%, and F1245 12%) [8], the most common of which, F1174L and R1275Q (73%), are present in some NB cell lines such as SH-SY5Y and LAN-5 that we have chosen for therapeutic approaches in vivo of the present study Pharmacokinetic profiles performed in nude mice indicated that X-396 is slowly absorbed in vivo as indicated by T MAX value of 2 h, followed by a moderate terminal plasma elimination half-life, thus indicate a favourable behaviour for clinical translation. The plasma and tumor concentrations were well above the IC 50 required to inhibit the NB cell growth even at the low doses of 25 mg/kg or 50 mg/kg, but higher dose of 100 mg/kg is required to reach high enough concentration in the brain to achieve prolonged inhibition, if brain metastases were present. However, the determination of the brain concentration is mainly relevant from a toxicology perspective.
The development of acquired resistance to targeted therapies is considered a largely inevitable hurdle that has a substantial impact on patients [15,38]. At least one mechanism of acquired resistance is the emergence of new ALK mutations at relapse of NB [27]. Therefore, alternative strategies that may overcome acquired resistance to therapeutic agents and increase the efficacy of small molecule inhibitors in ALK-mutated NB, are of particular relevance. One promising strategy could be the combination of kinase inhibitors with gene-specific silencing molecules, such as siRNAs, which are able to knockdown ALK independently of its mutational status by promoting degradation of the gene mRNA. In this context, we have recently generated targeted nanoliposomal formulations carrying ALK-directed siRNAs (TL[ALK-siRNA]), which are highly specific for the gene target [21,22] and efficiently delivered to NB cells thanks to a selective tumor targeting provided by the carrier [24,39].
Indeed, compared to free ALK-siRNA, TL[ALK-siRNA] formulations have low plasma clearance, increased siRNA stability, improved binding to NB cells, and are effective for ALK-silencing and induction of NB cell death [22]. Our liposomal formulation showed a strong ALK knockdown in mice-bearing NB tumors, which resulted in cell growth inhibition and prolonged survival, as described [21].
In the current study, we sought to combine the TL[ALK-siRNA] with the novel ALK-inhibitor X-396 showing, for the first time to our knowledge, that this drug combination exhibits a greater anti-tumor effectiveness in biologically and clinically relevant NB mouse models when compared with the single agents. Moreover, the combined treatment was also proven to be well tolerated, with no obvious toxicities (i.e. weight loss, skin rush), further underlining a potentially promising therapeutic translation.
In summary, this study provides solid evidence that a more sensitive and specific therapeutic approach for NB may be achieved by blocking the ALK signaling pathway simultaneously through the directed ALK gene knockdown by ALK-siRNA nanoliposomal formulation in combination with the pharmacological inhibition of the ALK kinase activity by X-396.
Although ALK mutations are found in a small percentage of neuroblastomas, other ALK activating aberrations may occur at genomic, or post-transcriptional or post-translational level (gene amplification, over expression, epigenetic regulation, phosphorylation). Therefore, the ALK role in the activation of downstream pathways involved in the tumorigenesis of NB might be far greater than expected.
We think that our data suggest that novel small molecule ALK inhibitors combined with RNA interferingbased nanoliposomes could ultimately subdue many ALKdriven cancers into manageable diseases. All other reagents were of analytical grade purity or the highest available purity and purchased from Sigma-Aldrich.

Compounds
X-396 were provided by Xcovery (West Palm Beach, FL) through a materials transfer agreement, as a lyophilized white cake in 10 mL glass vials and stored at 4°C. Crizotinib were purchased from Selleckchem (Houston, TX 77054 USA; licensed by Pfizer). For the in vitro experiments, the appropriate amount of powder was dissolved in dimethyl sulfoxide (DMSO) to a final concentration of 10 mM and stored at -20°C until use. Reconstituted drugs were thawed and diluted in culture medium (to a final DMSO concentration less than 0.01%) immediately before use.
The vehicle solution, obtained by mixing 0.5% Hydroxy Propyl-Methyl-Cellulose (HPMC) and 0.4% Tween-80 in 99.1% sterile water (w\w\v), was used to resuspend X-396 and crizotinib for the in vivo experiments.

Cell lines and culture conditions
Two human Neuroblastoma (NB) cell lines carrying different ALK-mutations, SH-SY5Y (ALK F1174L ) and LAN-5 (ALK R1275Q ), were grown in Dulbecco's-modified-EM and RPMI 1640 medium (Sigma), respectively [21,22]. Moreover, cells were tested for mycoplasma contamination, cell proliferation, morphology evaluation, and multiplex short tandem repeat profiling test, both after thawing and within six passages in culture.

Cell proliferation and cell viability assay
NB cell lines were seeded in 96-well plates (at 1 × 10 3 -8 × 10 3 cells per well) in complete medium and cultured for 24 hours. The medium was removed and replaced with fresh complete medium supplemented with four different concentrations of crizotinib or X-396 (10 -2000 nM) or 0.01% DMSO (control). Six replicates of cells for each condition were then incubated for an additional 72 hours. Eighteen hours before the end of the treatment cells were then incubated with 0.5 μCi (0.0185 MBq) of 3 H-thymidine (PerkinElmer, Waltham, Ma, USA) and processed for liquid scintillation counting (Packard Instruments Company, Downers Grove, IL) [40,41].
For viability assay, two hours before the end of the treatment cells were incubated with AlamarBlue staining according to the manufacture's instructions (Life Technologies).

Western blot analysis
NB cell lines were plated in 10 cm 2 -dishes and 24 hours after were treated with four different concentrations of crizotinib or X-396 (10 -1000 nM) or 0.01% DMSO (Control) for in total 72 hours.
Total cell lysates were prepared and analyzed by western blot analysis [40,42]. Briefly, cells lines were lysed with Cell Extraction Buffer (Life Technologies) plus protease-inhibitor cocktail (Sigma). Protein lysates (50 μg per lane) were resolved on sodium dodecyl sulphate (SDS) 8% polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were then incubated with antibodies against ALK and phospho-ALK tyrosine 1586 (Cat N°3333 and 3343 respectively, Cell Signaling Technology, Danvers, MA). Peroxidase-conjugated goat anti-mouse and anti-rabbit antibodies were used as secondary antibodies (Cell Signaling Technology). Immune complexes were visualized with the use of a Supersignal West Pico Chemiluminescent Substrate (Thermo-Scientific, Rockford, IL, USA) or Clarity Western ECL Substrate (Bio-Rad Laboratories, Milan, Italy) according to the manufacturer's instructions, and normalized to internal controls (a mouse monoclonal antibody against β-actin (Sigma)). Bands were quantified by using of Image J software (http://imagej.nih.gov/ij/index.html).

Pharmacokinetic and therapeutic studies in subcutaneous NB model
Six-week-old male Balb/c nude mice (Beijing HFK Bio-Technology co. Ltd, Beijing, China) were subcutaneously injected with 4 × 10 6 SH-SY5Y cells in matrigel. Once tumors reached an average volume of 250 mm 3 , 24 mice were randomly splitted in three groups and treated via oral gavage (OG) with either 25 or 50 mg/Kg of X-396 administered bis in die (BID) or 100 mg/kg of X-396 administered quaque die (QD) for 14 days. A uniform volume per administration (200 μL) was used for each group. Mice were monitored daily throughout the treatment period for sign of morbidity/mortality. On the last day of study, 0.1 -0.2 mL of blood was collected from the X-396 treated mice as follows: 3 mice in each group were bled at pre-dose, 1, 4 hours post dose, and sacrificed at 4 hours.
Another 3 mice in each group were bled at 0.5, 2, and 8 hours post-dose, sacrificed at 8 hours. Tumor and brain tissues were explanted. Drug concentrations in plasma and tissues were analyzed by Liquid Chromatography-Mass Spectrometry (LC-MS)/MS [10].
In other set of experiments, Balb/c nude mice were xenografted with SH-SY5Y cells as earlier. Once tumors reached an average volume of 250 mm 3 , mice were randomized and treated as above or with either X-396 or crizotinib via OG, with 50 mg/Kg BID, or control vehicle. Tumors were measured twice weekly using callipers, and volume was calculated using the formula: Length × Width 2 × 0.5. Body weight was also assessed twice weekly. The experiment was terminated after 3 weeks of treatment.
All studies were done in accordance with the current International Conference on Harmonization harmonized tripartite guidelines, and the protocol was approved by the Sundia Institutional Animal Care and Use Committee (Shangai, China).

Preparation of Targeted Liposomes [ALK-siRNA]
Briefly, ALK-directed siRNA were complexed with cationic lipids and then coated with neutral lipids [22]. Successively, Fab' fragment of anti-GD 2 monoclonal antibody was coupled to the malemide terminus of DSPE-PEG 2000 -MAL of the liposomes [24] to provide a formulation of Targeted Liposomes entrapping siRNA against ALK (TL[ALK-siRNA]). Liposomes size, polidispersity, and zeta-potential were analysed by dynamic light scattering using zeta-sizer Nano-S ZS90 particle sizer at a fixed angle (90°) (Malvern Instruments, Malvern UK). www.impactjournals.com/oncotarget In vivo therapeutic studies in orthotopic and pseudometastic NB mouse models All experiments involving these models have been reviewed and approved by the licensing and ethical committee of the IRCCS Azienda Ospedaliera Universitaria San Martino-IST Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy and by Italian Ministry of Health. All the in vivo experiments were performed using the least number possible in accordance with the 3 Rs policy.
No mice died as result of this treatment.
In the pseudometastatic model, 3 × 10 6 SH-SY5Y cells were injected intravenously (i.v.) in the tail vein in four-weekold female nu/nu mice [24,44]. Mice were then randomly assigned to the different groups and treated with either the vehicle alone (control mice) or with X-396 and/or TL[ALK-siRNA] (administered individually or in combination). Depending on the experimental schedules, mice received 50 or 100 mg/Kg of X-396 via oral gavage QD. An amount of 1 mg/Kg (20 μg/injection per mouse) of TL[ALK-siRNA] were i.v. injected twice a week, for a total of 3 weeks treatment, with a 3-day interval between injections.
Body weight and general physical status of the animals were recorded daily until they were judged to be in discomfort by animal caretakers. Specifically, once showing signs of poor health (i.e., abdominal dilatation, dehydration, paraplegia, severe weight loss), mice were euthanized by anaesthesia with xilezine (Xilor 2%, Bio98 Srl, Milan, Italy).

Bioluminescent imaging for evaluation of primary therapeutic responses
A Bioluminescence Imaging System (BLI), able to capture signals released by Luciferase (Luc)-stablytransduced NB cells [45,46], named SH-SY5Y-Luc and LAN-5-Luc, was used to monitor disease progression, metastasis and therapeutic effects in orthotopic-xenografted mice. Briefly, SH-SY5Y-Luc and LAN-5-Luc tumor cells, were orthotopically implanted into the left adrenal gland of nu/nu mice [45,47]. The initial in vivo trafficking of the malignant cells, organ-specific homing, orthotopic expansion over time, and response to X-396 treatment were all readily visualized and quantified by a highly sensitive, cooled Charge-Coupled Device (CCD) camera mounted in a light-tight specimen box (IVIS Caliper Life Sciences, Hopkinton, MA) after 10 min incubation with 150 μg/mL of D-Luciferin (Caliper Life Sciences) [45].
Quantification of the in vivo Luciferase signal of the tumor area, expressed as fold increase of photon counts over time in the Region Of Interest (ROI), was computed at time 22 and 29 day with respect to the starting day of the treatment (T 0 ).

Statistical analysis
Results are expressed as mean values with 95% confidence intervals. All in vitro data are from at least three independent experiments. The statistical significance of differential findings between experimental and control groups was determined by Student's t test with Welch's correction or by one-way analysis of variance (ANOVA) with the Tukey's or with the Bonferroni's multiple comparison test. IC 50 values were evaluated by non-linear regression (curvefit: log inhibitor vs normalized response-variable slope) and significance between slopes were evaluated by linear regression, Best-fit values and Spearman r (two-tailed) in GraphPad Prism 5.03 software (GraphPad Software, Inc., La Jolla, CA). These findings were considered significant if twotailed P values were < 0.05.
All in vivo experiments were repeated at least two times with similar results. Kaplan-Meier method & Log Rank (Mantel-Cox) test was applied to build up and compare the survival curves between analyzed groups.

ACKNOWLEDGMENTS
D. Di Paolo is partially supported by Fondazione Umberto Veronesi (FUV) and Associazione Italiana per la Ricerca sul Cancro (AIRC). F. Pastorino is supported by FUV and Istituto G. Gaslini (IGG) Excellence Contract 2013-14. A. Di Fiore was supported by AIRC. F. Piaggio was supported by FUV. C. Brignole is partially supported by FUV and AIRC. P. Perri is supported by IGG Excellence Contracts 2012-13 and 2014-2015. We thank Monica Loi for helpful technical support.