TRAIL induces pro-apoptotic crosstalk between the TRAIL-receptor signaling pathway and TrkAIII in SH-SY5Y cells, unveiling a potential therapeutic “Achilles heel” for the TrkAIII oncoprotein in neuroblastoma

TrkAIII expression in neuroblastoma (NB) associates with advanced stage disease, worse prognosis, post therapeutic relapse, and in NB models TrkAIII exhibits oncogenic activity and promotes chemotherapeutic-resistance. Here, we report a potential therapeutic “Achilles heel” for the TrkAIII oncoprotein in a SH-SY5Y NB model that is characterised by one-way TRAIL-induced, pro-apoptotic crosstalk between the TRAIL receptor signaling pathway and TrkAIII that results in the delayed induction of apoptosis. In TrkAIII SH-SY5Y cells, blocked in the intrinsic apoptosis pathway by elevated constitutive Bcl-2, Bcl-xL and Mcl-1 expression, TRAIL induced delayed caspase-dependent apoptosis via the extrinsic pathway and completely abrogated tumourigenic capacity in vitro. This effect was initiated by TRAIL-induced SHP-dependent c-Src activation, the induction of TrkAIII/SHP-1/c-Src complexing leading to SHP-mediated TrkAIII de-phosphorylation, subsequent induction of complexing between de-phosphorylated TrkAIII and cFLIP associated with a time-dependent increase the caspase-8 to cFLIP ratio at activated death receptors, resulting in delayed caspase cleavage and caspase-dependent apoptosis. We also confirm rate-limiting roles for c-FLIP and Mcl-1 in regulating the sensitivity of TrkAIII SH-SY5Y cells to TRAIL-induced apoptosis via the extrinsic and intrinsic pathways, respectively. Our study unveils a novel mechanism for the TRAIL-induced apoptosis of TrkAIII expressing NB cells that depends upon SHP/Src-mediated crosstalk between the TRAIL-receptor signaling pathway and TrkAIII, and supports a novel potential pro-apoptotic therapeutic use for TRAIL in TrkAIII expressing NB.


INTRODUCTION
The developmental and stress-regulated alternative TrkAIII splice variant of the neurotrophin receptor tropomyosin related kinase A (TrkA) is expressed by advanced stage human neuroblastomas (NB) and NB cell lines, associates with metastatic disease, worse prognosis and post-therapeutic disease-relapse in high TrkA expressing unfavourable tumours and exhibits oncogenic activity in NB models [1][2][3][4][5]. It is characterised by exon 6-7 skipping and is expressed as an immature N-glycosylated 100 kDa protein that is devoid of the extracellular D4 Ig-like domain and several N-glycosylation sites that prevent spontaneous receptor activation and facilitate cell surface expression. In contrast to fully-spliced TrkA receptors, TrkAIII is miss-localised to intracellular membranes within which it undergoes ligand-independent activation that promotes self-perpetuating re-cycling between the endoplasmic reticulum (ER) and ERGIC and recruitment to the centrosome. The result is chronic oncogenic signaling through IP3K/Akt but not Ras/ MAPK, a pro-survival ER-stress response characterised
In the present study, we report a potential therapeutic "Achilles heel" for TrkAIII in a human SH-SY5Y NB model. We show that TRAIL induces one way, pro-apoptotic crosstalk between the TRAIL receptor signaling pathway and TrkAIII, resulting in the induction of delayed apoptosis through the extrinsic pathway and the complete abrogation of tumourigenic activity in vitro. This TRAIL-induced pro-apoptotic cross talk is mediated by SHP protein tyrosine phosphatase(s) and the nonreceptor tyrosine kinase c-Src, depends upon TrkAIII de-phosphorylation, and TrkAIII binding and sequester of cFLIP, which increases the caspase-8 to cFLIP ratio at activated death receptors, inducing delayed caspasedependent apoptosis.
In substrate-independent tumourigenesis assays in vitro, TRAIL (200 ng/ml) completely abrogated tumourigenic growth of TrkAIII SH-SY5Y clone 1 and 2 but did not reduce the tumourigenic growth of either pcDNA SH-SY5Y clone 1 and clone 2 or NT SH-SY5Y cells over 14 days (Figure 2A-2C, data displayed for NT SH-SY5Y, pcDNA SH-SY5Y clone 1 and TrkAIII SH-SY5Y clone 1, only). SiRNA knockdown of Mcl-1 in NT-SH-SY5Y or pcDNA SH-SY5Y cells did not reduce tumorigenic activity in vitro in the presence of TRAIL (200 ng/ml) over 14 days (Figure 2A) nor sensitize to NT-SH-SY5Y or pcDNA SH-SY5Y cells to TRAILinduced apoptosis (not shown). The clear difference in NT-SH-SY5Y, pcDNA SH-SY5Y and TrkAIII SH-SY5Y tumourigenic activity in vitro in the presence of TRAIL (200 ng/ml) is demonstrated in Figure 2C, at higher magnification. TrkAIII SH-SY5Y clone 1 and pcDNA SH-SY5Y clone 1 were selected for further study.
Together, these data show that TrkAIII sensitizes SH-SY5Y cells to TRAIL-induced apoptosis, resulting in the abrogation of tumorigenic activity in vitro.

NT, pcDNA and TrkAIII SH-SY5Y cells express all components required for TRAIL-induced apoptosis
Western blot and RT-PCR comparisons of NT SH-SY5Y, pcDNA SH-SY5Y and TrkAIII SH-SY5Y cells revealed similar levels of functional (DR4 and DR5) and decoy (DcR1 and DcR2) TRAIL receptor mRNA and protein expression ( Figure 3A and 3B), with cell surface DR4 and DR5 expression detected in TrkAIII SH-SY5Y cells by indirect IF of non-permeabilized cells ( Figure 3C), Oncotarget 80822 www.impactjournals.com/oncotarget adding to our previous report of cell surface DR4 and DR5 expression in NT SH-SY5Y and pcDNA SH-SY5Y cells [49]. NT, pcDNA and TrkAIII SH-SY5Y cell lines also exhibited similar levels of constitutive caspase-3, caspase-8, caspase-9 mRNA and protein expression and cBID protein expression ( Figure 3A and 3B). In contrast to NT and pcDNA SH-SY5Y cells, TrkAIII SH-SY5Y cells also exhibited higher constitutive mRNA and protein expression of the intrinsic apoptosis pathway inhibitors Bcl-2, Bcl-xL and Mcl-1, and exhibited lower levels of cFLIP protein but not mRNA expression ( Figure 3A and 3B).
The possibility that degradation at the proteasome was responsible for reducing cFLIP protein levels in TrkAIII SH-SY5Y cells was confirmed using the proteasome inhibitor MG132 (10 µM), which increased cFLIP protein levels in TrkAIII SH-SY5Y cells ( Figure 4E).
Together, these data show that resistance to TRAILinduced apoptosis does not result from the lack of TRAILinduced apoptosis component expression or the expression of Bcl2 family inhibitors but is associated with higher levels of cFLIP expression. Conversely, TrkAIII SH-SY5Y sensitivity to TRAIL-induced apoptosis is associated with reduced cFLIP levels that result from TrkAIII activity and cFLIP degradation at the proteasome and occurs despite increased Bcl2 family expression.  Oncotarget 80826 www.impactjournals.com/oncotarget TRAIL-induced apoptosis is delayed, caspasedependent and cFLIP and Mcl-1-regulated TRAIL (200 ng/ml)-induced TrkAIII SH-SY5Y apoptosis was not detected prior to 6 hours. At 6 hours, TRAIL-induced a mean (±SD) of 18.3 ± 9.8% cell death (p = 0.0264, n = 6), 30.4 ± 13.8% cell death at 12 hours (p = 0.0033, n = 6) and 68.4 ± 23.4% cell death at 24 hours (p < 0.0001, n = 6) ( Figure 5A and 5B).
In addition to inducing TrkAIII/SHP-1/c-Src complexing, TRAIL also induced de-phosphorylation of TrkAIII Y674/675 within 6 hours ( Figure 10A and 10B), prevented by both PP1 and NSC87877 ( Figure 10A and 10B)  Oncotarget 80831 www.impactjournals.com/oncotarget in association with significant inhibition of TRAIL-induced apoptosis, confirming roles for both c-Src and SHP in TRAIL-induced TrkAIII de-phosphorylation and subsequent apoptosis. PP1 and NSC87877 did not reduce constitutive TrkAIII Y674/675 phosphorylation in the absence of TRAIL (data not shown). Association between TRAIL-induced TrkAIII Y674/675 de-phosphorylation and apoptosis was confirmed in Western blots by exclusive detection of TrkAIII Y674/675 de-phosphorylation in apoptotic but not surviving TrkAIII SH-SY5Y cells (see Figure 11B, lower panel).
Together, these data implicate both SHP and c-Src in TRAIL-induced pro-apoptotic crosstalk with TrkAIII, and implicate SHP in c-Src activation and TrkAIII de-phosphorylation.

TRAIL promotes TrkAIII/cFLIP complexing and alters the caspase-8 to cFLIP ratio recruited to death receptor complexes
In co-immunoprecipitation experiments, cFLIP immunoprecipitated from untreated whole TrkAIII SH-SY5Y cell extracts pulled down very low levels of TrkAIII, that were not reduced further by GW441756 (1 µM for 24 hours) ( Figure 11A). C-FLIP immunoprecipitates pulled down nonphosphorylated TrkAIII from TRAIL-treated (200 ng/ml for 24 hours) apoptotic but not surviving TrkAIII SH-SY5Y cells ( Figure 11A and 11B). TRAIL-induced TrkAIII Y674/675 de-phosphorylation was confirmed in input samples from apoptotic but not surviving cells ( Figure 11B, INPUT). In ligand-precipitation/Western blots, streptavidin precipitation of biotin-labelled TRAIL-ligated DR4 positive death receptor complexes from TrkAIII SH-SY5Y cells treated for 1 and 6 hours with biotinylated TRAIL, revealed an increase in the caspase 8 to cFLIP ratio from 0.66 following 1 hour TRAIL-treatment to 1.53 following 6 hour TRAIL-treatment. In contrast, the caspase-8 to cFLIP ratio of 0.68 pcDNA SH-SY5Y cells treated for 1 hour with biotinylated TRAIL remained at the same level in 6 hour-treated cells ( Figure 12A and 12B).
Together, these data link TRAIL-induced SHP/c-Src-mediated TrkAIII de-phosphorylation and complexing with cFLIP to a pro-apoptotic increase in caspase-8 to cFLIP ratio at TRAIL-activated death receptors.

DISCUSSION
In the present study, we report a potential therapeutic "Achilles heel" for the TrkAIII oncoprotein in a SH-SY5Y NB model. In this model, TRAIL induced delayed caspase-dependent apoptosis of SH-SY5Y cells engineered to express TrkAIII, resulting in the complete abrogation of tumorigenic growth capacity in vitro. Consistent with the TRAIL-resistant phenotype of parental and control-transfected SH-SY5Y cells (this study and 49), TRAIL-induced apoptosis of TrkAIII SH-SY5Y cells resulted from one-way SHP/c-Src-mediated pro-apoptotic crosstalk between the TRAIL receptor signaling pathway and TrkAIII. This results from SHP/c-Src-dependent binding and de-phosphorylation of TrkAIII, inducing cFLIP complexing with de-phosphorylated TrkAIII, increasing the caspase-8 to cFLIP ratio at TRAIL-activated death receptors, which explains the delayed induction of caspase cleavage and apoptosis, with rate-limiting roles confirmed for both c-FLIP and Mcl-1.
All the cell lines used in this study exhibited constitutive expression of components required for TRAIL-induced apoptosis (this study and 49), a proapoptotic expression equilibrium between functional and decoy TRAIL receptors and cell-surface functional DR4 and DR5 TRAIL receptor expression (this study and 49). In contrast to some NB cell lines [55], all of our cell lines expressed caspase 8 mRNA and protein. However, inhibitory caspase-8 Y380 phosphorylation [54] was also detect in all cell lines, providing an additional mechanism for reducing SH-SY5Y sensitivity to TRAIL-induced apoptosis (this study and 49).
TrkAIII SH-SY5Y cells expressed lower levels of cFLIP but higher levels of Bcl-2, Bcl-xL and Mcl-1 than either NT SH-SY5Y or pcDNA SH-SY5Y cells. Inhibitor (GW441756 and MG132) studies indicated that lower cFLIP protein expression by TrkAIII SH-SY5Y cells was post-transcriptional and dependent upon enhanced cFLIP degradation at the proteasome and TrkAIII tyrosine kinase activity. We are currently investigating whether this may depend upon Nedd-4 family E3 ligases, which bind activated TrkA receptors and promote cFLIP degradation at the proteasome [56,57].
TRAIL-induced TrkAIII SH-SY5Y apoptosis was delayed and not immediate, was not detected prior to 6 hours and increased to a maximum post 12 hours. Delayed apoptosis was associated with delayed cleavage of caspase 8 and delayed cleavage of the caspase 8 substrates c-BID [61] and caspase 3 [62], confirming a delay in caspase 8 activation. Therefore, low cFLIP levels combined with inhibitory caspase 8 Y380 phosphorylation were sufficient to prevent immediate apoptosis, implicating additional TRAIL-induced time-dependent pro apoptotic changes.
The central role of cFLIP in delaying TRAILinduced TrkAIII SH-SY5Y apoptosis was confirmed by siRNA knockdown, which accelerated and augmented apoptosis to > 90% within 6 hours, and by cFLIP over expression that significantly inhibited TRAIL-induced TrkAIII SH-SY5Y apoptosis. This suggests that TRAILinduced reduction in cFLIP function resulting from proapoptotic crosstalk with TrkAIII may underpin the delay in TRAIL-induced apoptosis. However, since TRAIL did not reduce cFLIP protein expression in TrkAIII SH-SY5Y cells any further reduction in cFLIP function would have to be post-translational.
Inhibitor studies confirmed that TRAIL-induced proapoptotic crosstalk between the TRAIL-receptor pathway and TrkAIII involved SHP, c-Src and TrkAIII activity. TrkA (GW441756), c-Src (PP1) and SHP1/2 (NSC-87877) inhibitors significantly reduced TRAIL-induced TrkAIII SH-SY5Y apoptosis, which was associated with timedependent changes in c-Src and TrkAIII phosphorylation. These changes were characterised by initial TRAILinduced de-phosphorylation of c-Src inhibitory tyrosine Y527 combined with increased phosphorylation of c-Src active site tyrosine Y416, consistent with activation [63], which was accompanied by the induction of complexing between phosphorylated TrkAIII, SHP-1 and c-Src, followed by TrkAIII Y674/675 de-phosphorylation, consistent with inactivation [64], implicating SHP-1 and c-Src in TRAIL-induced pro-apoptotic crosstalk with TrkAIII. It remains unclear, however, whether both SHP-1 and c-Src, which directly bind activated TrkA receptors [65,66], interact directly with TrkAIII under these conditions.
The SHP inhibitor NSC-87877 prevented both TRAIL-induced Src Y527 de-phosphorylation and increased Y416 phosphorylation. The c-Src inhibitor PP1 prevented TRAIL-induced c-Src Y416 phosphorylation but not Y527 de-phosphorylation, and both inhibitors prevented TRAIL-induced TrkAIII Y674/675 de-phosphorylation. This suggest that TRAIL first activates SHP, which dephosphorylates c-Src Y527 within constitutive complexes leading to c-Src activation and Y416 auto phosphorylation, which in turn promotes interaction with phosphorylated TrkAIII, resulting in SHP-mediated TrkAIII Y674/675 dephosphorylation and subsequent apoptosis. This possibility is supported by reports that SHP is recruited and activated at death receptors [67], TRAIL activates c-Src [68], c-Src and SHP bind activated TrkA receptors [65,66] and SHP de-phosphorylates TrkA Y674 and Y675 residues [66].
The association between TRAIL-induced TrkAIII dephosphorylation and apoptosis was confirmed by detection of TrkAIII Y674/675 phosphorylation in TRAIL-treated www.impactjournals.com/oncotarget surviving but not apoptotic cells. TrkAIII de-phosphorylation alone, however, would appear to be insufficient to sensitize SH-SY5Y cells to TRAIL-induced apoptosis, since TrkAIII inactivated by GW441756 (this study and 8) significantly inhibited TRAIL-induced TrkAIII SH-SY5Y apoptosis. This suggests that a more complex interaction between c-Src/ SHP and phosphorylated TrkAIII, resulting in TrkAIII dephosphorylation is required for a pro-apoptotic outcome. Furthermore, a pro-apoptotic outcome may also reflect changes in the c-Src/SHP equilibrium, since activated c-Src promotes apoptotic resistance and TrkA phosphorylation [65,68], whereas activated SHP promotes death receptormediated apoptosis and TrkA de-phosphorylation [66,67], and both exhibit reciprocal regulation [69]. However, since inhibitors of both SHP and c-Src reduced TRAILinduced TrkAIII SH-SY5Y apoptosis in association with maintenance of TrkAIII phosphorylation, both SHP and c-Src are required for TRAIL-induced pro-apoptotic cross talk between the TRAIL receptor pathway and TrkAIII.
In a direct comparison between TRAIL-treated apoptotic and surviving TrkAIII SH-SY5Y cells, increased complexing between cFLIP and de-phosphorylated TrkAIII was detected in apoptotic but not in surviving TrkAIII SH-SY5Y cells. This suggests that TRAIL may induce cFLIP sequester by de-phosphorylated TrkAIII with intracellular membranes [1,8], reducing its recruitment to activated cell surface TRAIL receptors, which was confirmed by the time-dependent increase in caspase-8 to c-FLIP ratio recruited to DR4 positive receptors in TRAILtreated TrkAIII SH-SY5Y but not NT-SH-SY5Y cells. This extend our previous reports that c-FLIP complexes with NGF-activated TrkA receptors [49,70], sensitizing TrkA SH-SY5Y cells to TRAIL-induced apoptosis [49] to include a role for TrkAIII/cFLIP complexing in TRAILinduced TrkAIII SH-SY5Y apoptosis. cFLIP, therefore, can interact with both phosphorylated TrkA and nonphosphorylated TrkAIII receptors depending upon the circumstance, which may reflect not only differences in receptor structure but also localisation, post receptor signaling and/or complex composition.
SiRNA Mcl-1 knockdown accelerated and augmented TRAIL-induced TrkAIII SH-SY5Y apoptosis, characterising Mcl-1 as a major regulator of TrkAIII SH-SY5Y sensitivity to TRAIL-induced apoptosis. In contrast to siRNA cFLIP knockdown, which accelerated and augmented TRAILinduced apoptosis through the extrinsic pathway, z-LEHDfmk caspase-9 inhibitor prevented accelerated and augmented TRAIL-induced apoptosis following siRNA Mcl-1 knockdown, confirming intrinsic apoptosis pathway involvement [23][24][25]. Interestingly, Mcl-1 expression in TrkAIII SH-SY5Y cells was inhibited by the PERK inhibitor GSK2656157 [71] but not by the TrkA inhibitor GW441756, whereas Bcl2 and Bcl-xL expression was reduced by GW441756 but not GSK2656157, implicating TrkAIII activity in Bcl-2 and Bcl-xL but not Mcl-1 expression and the PERK arm of the ER-stress response in the expression of Mcl-1 but not Bcl-2 and Bcl-xL. This supports reports of ER-stress regulated Mcl-1 expression [72] and the activation an ER-stress response survival adaptation in TrkAIII SH-SY5Y cells [11,12], adding PERK up-regulated expression of Mcl-1 as a novel potential "non-oncogene addiction" mechanism to oncogenic TrkAIII repertoire.
Constitutive inhibitory caspase 8 Y380phosphorylation [73] was detected at similar levels in all SH-SY5Y cell lines, providing an additional mechanism for reducing sensitivity to TRAIL-induced apoptosis. However, neither GW441756 nor TRAIL reduced caspase 8 Y380-phosphorylation in TrkAIII SH-SY5Y cells, indicating that caspase 8 Y380-phosphorylation does not depend upon TrkAIII and its de-phosphorylation was not involved in sensitizing TrkAIII SH-SY5Y cells to TRAILinduced apoptosis.
In conclusion, we have identified a potential therapeutic "Achilles heel" for the TrkAIII oncoprotein, characterised by TRAIL-induced one-way pro-apoptotic crosstalk between the TRAIL receptor signaling pathway and TrkAIII, in SH-SY5Y NB cells. We propose that TRAIL-induced apoptosis initiates with TRAIL activation of SHP, followed by SHP-mediated activation of c-Src. This promotes complexing between phosphorylated TrkAIII, SHP and c-Src, leading to SHP-mediated TrkAIII de-phosphorylation and the sequester cFLIP, which increases the caspase-8 to cFLIP ratio recruited to activated TRAIL-receptors, explaining the delay in caspase cleavage and apoptosis via the extrinsic pathway. Our observations add significantly to a previous report of pro-apoptotic cross talk between TNF-α signaling networks and tyrosine kinase receptors [74] and provide a novel rational for the therapeutic use of TRAIL, combined with siRNA cFLIP and Mcl-1 inhibitors, in TrkAIII expressing NBs. Such an approach may work best in firstline therapy, as TRAIL promotes KRAS-driven metastasis [75] and KRAS mutations are detected in relapsed but not primary NBs [76][77][78].

Transient Bcl-xL and cFLIP expression
Bcl-xL and flag-tagged cFLIP were overexpressed in TrkAIII SH-SY5Y cells by transient transfection of mammalian expression vectors bearing Bcl-xL, FLAG-tagged cFLIP [79] or empty pcDNA. Briefly, 0.6 µg/ml of each vector were transfected into sub-confluent cell cultures using FUGENE transfection reagent as directed by the manufacturer (Promega, Madison, Wi). At 6 hours transfection medium was replaced with fresh growth medium (RPMI/10%FCS/ glutamine plus antibiotic) and the cells left for 48 hours, at which time they were utilised for experimentation. Overexpression of Bcl-xL and Flag-tagged cFLIP was verified by Western blotting of total cell extracts (20 µg).

Cell extraction, immunoprecipitation and Western blotting
Cells were extracted in lysis buffer (PBS containing 0.5% sodium deoxycholate, 1% NP40, 0.1% SDS, 1 mM sodium orthovanadate, 1 mM PMSF, 1 μg/ml of pepstatin A and Aprotinin) and protein concentrations were calculated by Bradford protein concentration assay (Sigma-Aldrich). Samples were mixed with reducing SDS-PAGE sample buffer and subjected to reducing SDS-PAGE/Western blotting. Briefly, proteins separated by reducing SDS-PAGE, were trans-blotted onto Hybond C+ nitrocellulose membranes by electrophoresis (Amersham Int. UK) and the membranes subsequently air-dried. Non-specific protein binding-site on membranes were blocked by incubation for 2 hours in 5% non-fat milk in TBS-T prior to incubation with primary antibodies, at recommended dilutions, for 2-16 hours at 4°C. Membranes were then washed in TBS-T, incubated with secondary HRP-conjugated antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) diluted in blocking solution and immunoreactive species detected by chemiluminescence reaction, as directed by the manufacturer (Amersham Int).

Cell death assay
TRAIL-induced cell death was routinely assayed using a modification of previously described methods [81,82]. Briefly, cells were washed once in Ca 2+ free PBS, detached with ice cold PBS containing 1 mM EDTA, transferred to sterile 15 mls tubes, centrifuged for 5 minutes at 1,000 × g at 4°C, washed with ice cold PBS and re-pelleted by centrifugation at 1,000 × g for 5 minutes, at 4°C. Cell pellets were re-suspended in 25 μl of PBS containing 2 μl of acridine orange/ethidium bromide solution (100 μg/ml acradine orange and 100 μg/ml ethidium bromide in PBS) plated onto glass slides and examined immediately under a Zeiss "Axioplan-2" fluorescence microscope. Representative fields were digitally photographed under identical exposure conditions and the number of dead cells (orange/red nuclei) and live cells (green nuclei) counted. In addition, phase contrast micrographs of parallel cultures were used to confirm changes in the relative percentage of adherent (surviving) and suspension (apoptotic) cells, following TRAIL-treatment. TRAIL-induced TrkAIII SH-SY5Y cell apoptosis was confirmed by its complete inhibition by z-VAD-fmk pan caspase inhibitor.

Tumour growth in soft agar
For substrate-independent tumour growth assays [49], 1 × 10 4 cells in single-cell suspension (passed through a gauge × 18 syringe needle) were mixed in a 33% solution of agar (BiTec; Difco) in RPMI containing 5% FCS at 37°C, with or without TRAIL (200 ng/ml) and layered onto a solid 0.6% agarose substrate also with or without TRAIL, prepared in the same growth medium. Following top phase agar solidification, complete medium was added and re-placed every 2 days. Tumour spheroid growth was monitored over a 14-day period by phase contrast microscopy. Tumour spheroids were counted in 10 random fields at ×10 magnification.

Isolation of TRAIL-activated death receptor complexes
TRAIL-ligated death receptor complexes were purified from biotin-TRAIL treated pcDNA SH-SY5Y and TrkAIII SH-SY5Y cells by ligand affinity precipitation, as previously described [83]. Briefly, biotinylated TRAIL was prepared by incubating TRAIL (1 mg/ml) with Sulfo-NHS-LC-Biotin (1 mg/ml) (Pierce) for 1 hour on ice. The reaction was stopped by adding 1/10 volume of 1M Tris-HCl [pH. 7.5] and unincorporated biotin removed by buffer exchange into 150 mM NaCl, 30 mM HEPES (pH 7.5) using PD-10 columns (Amersham Pharmacia Biotech). For ligand affinity precipitation, cells (5 × 10 6 cells per sample) were washed twice in RPMI at 37°C and incubated with 1 mg/ml biotinylated TRAIL for 1 and 6 hours. Death receptor complex formation was stopped by the addition of 15 volumes of ice cold PBS and cells were then lysed in 4.5 mls of lysis buffer (30 mM TRIS-HCl [pH. 7.5], 150 mM NaCl, 10% glycerol, 1% Triton X-100, supplemented with complete protease inhibitor cocktail) (Roche Diagnostics, Mannheim, GE). TRAIL receptor protein complexes were precipitated from lysates by coincubation with 20 μl of Streptavidin beads (Pierce) for 3 hours at 4°C with rotation. Ligand affinity precipitates were washed four times in lysis buffer, eluted from beads in reducing SDS-PAGE sample buffer and subjected to SDS-PAGE Western blotting.

Indirect immunofluorescence
Cells grown on Nunc glass chamber slides (Sigma-Aldrich) were washed in PBS, fixed in 4% paraformaldehyde, washed in PBS then processed for indirect immunofluorescence (IF). Fixed, non-permeabilised cells were incubated for 1 h in blocking solution (1% bovine serum albumin in PBS-0.03% TX100) and then incubated for 2 hours with primary antibody in blocking solution at room Oncotarget 80838 www.impactjournals.com/oncotarget temperature. Slides were washed three times in PBS-0.03% TX100, incubated with secondary fluorochrome-conjugated antibody diluted in blocking solution for 1 h at room temperature, washed in PBS-0.03% TX100 and mounted using VectorMount. IF images were obtained using a Zeiss Axioplan 2 fluorescence microscope with a digital camera and Leica M500 Image Manager software.

Statistical analysis
Data were analysed statistically by Student's t-test and statistical significance was associated with probabilities of < 0.05.

ACKNOWLEDGMENTS AND FUNDING
The first two authors contributed equally to this work. This work was supported by the Maugeri Foundation.