Cooperative Cross-Talk between Neuroblastoma Subtypes Confers Resistance to Anaplastic Lymphoma Kinase Inhibition

Abstract

Neuroblastoma is a pediatric solid tumor that can be stratified into stroma-rich and stroma-poor histological subgroups. The stromal compartment of neuroblastoma is composed mostly of Schwann cells, and they play critical roles in the differentiation, survival, and angiogenic responses of tumor cells. In certain neuroblastoma cell lines, the coexistence of neuroblastic N-type and substrate-adherent S-type is frequently observed. One such cell line, SK-N-SH, harbors a F1174L oncogenic mutation in the anaplastic lymphoma kinase (ALK) gene. Treatment of SK-N-SH with an ALK chemical inhibitor, TAE684, resulted in the outgrowth of S-type cells that expressed the Schwann cell marker, S100α6. Nucleotide sequencing analysis of these TAE684-resistant (TR) sublines revealed the presence of the ALK F1174L mutation, suggesting their tumor origin, although ALK protein was not detected. Consistent with these findings, TR cells displayed approximately 9-fold higher IC₅₀ values than N-type cells. Also, unlike N-type cells, TR cells have readily detectable phosphorylated STAT3 but weaker phosphorylated AKT. Under coculture conditions, TR cells conferred survival to N-type cells against the apoptotic effect of TAE684. Cocultivation also greatly enhanced the overall phosphorylation of STAT3 and its transcriptional activity in N-type cells. Finally, conditioned medium from TR clones enhanced cell viability of N-type cells, and this effect was phosphatidylinositol 3-kinase dependent. Taken together, these results demonstrate the ability of tumor-derived S-type cells in protecting N-type cells against the apoptotic effect of an ALK kinase inhibitor through upregulating prosurvival signaling.

Keywords

neuroblastoma ALK TAE684 Schwann cells stromal

Introduction

Neuroblastoma (NB) is a childhood malignancy of the sympathetic nervous system characterized by cellular heterogeneity and histological diversity.^1,2 Observations from tumor histology and cell lines derived from primary tumor specimens have identified 3 major NB subtypes: N (neuroblast), S (substrate adherent), and I (intermediate).³ The SK-N-SH cell line has been used to isolate multiple NB subtypes. For example, the N-type subline, SH-SY5Y, is characterized by neurite extension, moderate tumorigenicity, and expressing neuronal markers such as neurofilaments 70/160 and dopamine β-hydroxylase.⁴ The S-type subline, SH-EP1, is epithelial like and nontumorigenic and expresses EGF receptor and fibronectin. The I-type cell line, SH-IN, is stem cell like and can be differentiated into either N- or S-type cells with retinoic acid (RA) and 5-bromo-2′-deoxyuridine (BrdU), respectively.⁵

The role of tumor stroma in the pathogenesis of NB still remains controversial. Histologically, stroma-rich NB is associated with favorable prognosis.⁶ This may be related to the ability of Schwann cells in promoting differentiation in neuroblastic tumor cells.⁷ Alternatively, Schwann cells secrete antiangiogenic factors that can limit the formation of extensive tumor vasculatures.⁸ This is clearly demonstrated by the engraftment of SMS-KCNR cells in the sciatic nerve of mice.⁹ In this model, the infiltration of mouse Schwann cells is correlated with reduced angiogenesis, increased neuroblastic differentiation, and apoptosis. The origin of Schwann cells in NB has been traced to infiltrating reactive cells from the normal stroma.¹⁰ However, the finding of the same genomic aberrations in stromal cells and neuroblastic tumor cells would indicate that they are from a common progenitor.¹¹ However, still unknown is the role of tumor-derived Schwann cells in stroma-poor NB. Also, the signaling properties stemming from the cross-talk between stromal and tumor cells have not been reported.

Genome-wide linkage analysis of kindred with high-risk NB has identified a susceptibility locus on the short arm of chromosome 2 at band p23-24. The gene involved, anaplastic large cell lymphoma kinase (ALK), is a bona fide oncogene with both activating mutations and gene amplification being observed in a significant fraction of NB. The frequency of mutation in primary tumor samples is estimated to be 8%.^12-17 Interestingly, almost all mutations are within the catalytic loop or the C-helix of the kinase domain. Furthermore, 15% (14 of 94) of NB with MYCN amplification also have gene amplification in ALK.¹³ In addition, increased ALK expression without mutation is observed in 77% of high-grade (stage 3-4) tumors.¹⁸ One of the most common mutations (4.3%, 4 of 93) is in exon 23, which involves a cytosine-to-adenine substitution, altering the phenylalanine at codon 1174 to a leucine (F1174L).¹³ Meta-analysis of NB uncovers a significant association between F1174 mutation and MYCN amplification. In fact, the coexistence of both genetic aberrations is a poor prognosticator of disease.¹⁹

The ALK proto-oncogene encodes a protein of 1,620 amino acids with a predicted molecular weight of 176.4 kDa.^20,21 ALK is a class I receptor tyrosine kinase with the identity of the ligand still remaining controversial.²² ALK was originally identified as a fusion partner of the NPM-ALK chimeric oncoprotein found in anaplastic large cell lymphoma.²² Similar translocations are observed in multiple malignancies including inflammatory myofibroblastic tumors, squamous cell carcinomas, and non–small cell lung cancer (NSCLC).²² Invariably, ALK fusion oncoproteins are constitutively active and possess transforming activities.

There are intense efforts in developing small molecule inhibitors of ALK. NVP-TAE684, a 5-chloro-2,4-diaminophenylpyrimidine, was first identified as a potent ATP-competitive inhibitor of ALK with an IC₅₀ of 3 nM.²³ TAE684 induces growth arrest and apoptosis in several ALK-positive NB cell lines. Inhibition correlates with the suppression of Akt, STAT3, and Erk-dependent signaling. Similarly, PF-0234066 (crizotinib), another selective ATP-competitive inhibitor for both ALK and Met, displays an IC₅₀ of 23 nM in the Karpas299 cell line, which expresses the NPM-ALK oncoprotein.²⁴ This small molecule inhibitor is in phase III clinical trials of ALK-positive NSCLC.²⁵ As demonstrated for imatinib in the treatment of chronic myelogenous leukemia, the acquisition of drug resistance due to secondary mutations poses considerable challenges to achieve long-term remission.²⁶ Currently, data related to acquired resistance to ALK inhibitors are limited. Also, the responsiveness of different NB subtypes to ALK inhibition is not known. We describe in this report that different NB subtypes display differential responsiveness to the ALK inhibitor. Furthermore, the interactions between the N and S subtypes confer drug resistance towards ALK inhibition.

Results

Isolation of NB sublines insensitive to ALK inhibitor

In an attempt to isolate cell populations that were resistant to the ALK inhibitor, uncloned SK-N-SH was passaged in an escalating dose of TAE684 from 30 to 600 nM. SK-N-SH cell line was originally isolated from bone marrow metastases of an NB patient who possessed an admixture of N- and S-type cells in an approximately 30:1 ratio.²⁷ This cell line has a F1174L mutation in the ALK gene, rendering the mutant protein constitutively tyrosine phosphorylated.¹³ At the highest dose of 600 nM of TAE684, numerous resistant colonies with flat, epithelial-like morphology, resembling previously described S-type cells, were observed. These sublines were referred to as SK-N-TRs (TR hereafter) due to their TAE684-resistant properties. TR sublines could be distinguished morphologically from N-type cells that accounted for over 90% of the parental SK-N-SH cultures. Four TR clones (TR1-4) were isolated for further analysis (Fig. 1A).

As a comparison, a subline with flat cell morphology, SK-N-S1 (S1), was isolated from the untreated parental culture. However, S1 cells were smaller in size and with less prominent cell nuclei (Fig. 1A). In addition, 4 N-type sublines, SK-N-N1 to SK-N-N4 (N1, N2, N3, and N4), were randomly picked from the untreated parental culture, and they resembled the SH-SY5Y neuroblastic subclone described previously (Fig. 1A).⁴ Direct nucleotide sequencing analysis of the genomic DNA from N1, S1, and TR1 revealed the C to A mutation at codon 1174 of the ALK gene (Fig. 1B). This finding provided evidence that these sublines were tumor derived. This conclusion was further validated by short tandem repeat (STR) polymorphism analysis of N1 and TR3 genomic DNA. As expected, both sublines displayed identical polymorphic alleles in all 9 loci previously established for SK-N-SH. These data confirmed that TR sublines were derived from the same individual (Suppl. Fig. S1). Furthermore, karyotype analysis of TR1 and TR3 sublines revealed several marker chromosomes found in the SK-N-SH parental line. These included trisomy 7 [+7, del(7)(q34q36)], an aberrant chromosome 9 [add(9)(q34)], an interstitial deletion of chromosome 14 [del(14)(q13q22)], a derivative chromosome 22 with an unbalanced translocation with chromosome 17 [der(22)t(17,22)(q21.3;q13)], and an aberrant chromosome X [add(X)(p22.3)] (Suppl. Fig. S2). However, 4 of the 20 TR1 cells analyzed have additional materials of unknown origin translocated to the short arm of chromosome 2 [add(2)(p13)]. In addition, while all parental cells have an additional copy of chromosome 1, this aberrant event was only observed in 5 of 20 TR2 cells and none in TR1 cells analyzed (see Suppl. Table S1 for details).

Figure 1.

Characterization of NB subtypes. (A) Cell morphology of parental SK-N-SH cell line and representative S (S1, TR1, TR2) and N (N1, N2) sublines (upper panels). Magnification, 10x. Bar = 100 µm. (B) Genomic DNAs isolated from the indicated sublines were amplified by PCR, and direct sequencing analysis was performed to demonstrate the presence of a C-to-A substitution at codon 1174 (red asterisk).

Signaling divergence between N- and S-type sublines

To determine the activation states of ALK and its associated downstream signaling pathways in TR sublines, Western blotting analysis was performed. While both 220- and 140-kDa ALK protein species (c-ALK) were readily observed in SK-N-SH and all N-type sublines, they were almost undetectable in S1 and all the TR sublines (Fig. 2A). Similar to a previous report, the 220-kDa protein species, which represented the glycosylated full-length version of ALK, was the predominant isoform of the ALK F1174L mutant.²⁸ The lack of ALK expression in TR sublines was most likely due to transcriptional silencing, as RT-PCR analysis detected only faint signals in TR sublines while being readily detectable in all N-type sublines as well as SK-N-SH (Fig. 2B). Intriguingly, in the S1 cells, ALK was transcribed, but the lack of ALK protein would suggest a posttranslational block in its expression. Using an anti-phospho-Tyr1604 ALK antibody, the parental SK-N-SH and all N-type cells possessed robust phosphorylation of ALK (Fig. 2A and 2C). There were faint bands at 130 and 170 kDa in all TR lines. However, the complete absence of c-ALK in TR sublines tends to indicate that they were nonspecific signals.

For downstream signaling events, S-type cells, including TR sublines and S1 cells, all have robust phosphorylation of STAT3. On the contrary, all N-type sublines have lower levels of c-STAT3 and a 2- to 15-fold decrease in specific STAT3 phosphorylation (Fig. 2A and 2D). In stark contrast, N-type cells, especially for N1 and N2, possessed greater phosphorylation of the survival kinase, AKT, than all the TR lines (Fig. 2A and 2E). Finally, all TR sublines displayed marginally higher (~2- to 4-fold) levels of p-ERK1/2 when compared to the N-type cells (Fig. 2A and 2F). It is important to note that while N-type and S-type cells have distinct signaling profiles, the parental SK-N-SH, which has an admixture of N- and S-type cells, maintained robust phosphorylation of both STAT3 and AKT (Fig. 2A). Intriguingly, ERK1/2 phosphorylation was significantly inhibited in SK-N-SH (Fig. 2A). These observations would indicate the existence of cross-talk between these 2 distinct cell types (see section below).

Figure 2.

Signaling profiles of NB subtypes. (A) Total cell lysates were prepared from parental SK-N-SH, SK-N-S1 (S1), and 4 N- and S-type sublines. Approximately 50 µg of protein was loaded, and Western blotting analysis was carried out using the antibodies indicated. The 220-kDa glycosylated and 140-kDa processed isoforms of ALK are indicated by arrowheads. Unknown background bands are indicated by an asterisk. (B) Total RNA isolated from the indicated cells was subjected to RT-PCR analysis using 2 sets of ALK-specific primers; C = no RNA added to the RT-PCR reactions. Actin was used as a control. Amplified products were resolved on 1.5% agarose gel. N1a and N1b are different passages of the N1 subline. (C-F) The relative phosphorylation of the indicated signaling molecules was quantified from 3 independent lysates using the parental SK-N-SH cell line as a reference. Bars = SEM.

To determine the relative sensitivity of these cell lines to TAE684, dose response analysis revealed an IC₅₀ of 32 and 16 nM for N1 and N2, respectively. In contrast, both TR1 and TR2 were at least 27-fold more resistant, with an IC₅₀ of 870 nM (Fig. 3A). The ability of TAE684 in inhibiting p-ALK was clearly demonstrated in N1 cells, with >90% suppression observed between 30 to 100 nM (Fig. 3B). More importantly, TAE684 induced very similar kinetics of inhibition of p-AKT, reaffirming the role of ALK in activating this critical downstream kinase. Furthermore, when N1 and N2 cells were exposed to 30 nM TAE684, there was a 62.5% reduction in cell proliferation after 9 days in cultures (Fig. 3C and 3D). On the contrary, the proliferative capacity of 2 TR clones, TR1 and TR2, was not affected by TAE684 treatment under similar experimental conditions (Fig. 3E and 3F). Of note, most TR sublines isolated in this study have a much lower proliferative rate than N-type sublines. This may explain the fact that S-type cells are a minor population in the parental SK-N-SH cell line.

Figure 3.

Differential sensitivity of NB subtypes to TAE684. (A) Approximately 0.3 to 0.8 × 10⁴ cells were plated per well of a 96-well plate in triplicates and treated with an increasing concentration of TAE684 (Axon Medchem BV, Groningen, the Netherlands) for 72 hours. Cell viability was measured by the MTS assay (Promega, Madison, WI). Estimated IC₅₀s are shown. Bars = SD. (B) N1 cells were treated with the indicated doses of TAE684 for 6 hours. Total cell extracts were subjected to Western blotting analysis using the antibodies shown. * = unknown background band. (C-F) Approximately 1.5 × 10⁴ cells of the indicated sublines were plated per well of a 12-well plate and allowed to propagate in the absence (open circle) or presence (closed square) of TAE684 (30 nM). Growth curves were obtained by counting cell numbers at the indicated time points. Results from a single experiment performed in triplicates. Bars = SD.

Differentiation potential of TR sublines

To define the cellular characteristics of these sublines, lineage-specific markers were used to track neurogenesis in response to differentiation agents. For this, NF68, a 68-kDa core protein of neurofilaments presence in neurites, was used as a neuronal marker. In addition, S100α6, an EF hand–containing calcium binding protein, was used as a marker for Schwann cells.³ Both neuroblastic sublines, N1 and N2, responded to retinoic acid treatment by a drastic increase in neurite-like extensions (Suppl. Fig. S3A). This was correlated with a 20-fold increase in the level of NF68 (Suppl. Fig. S3B). On the contrary, BrdU almost completely abrogated the neurite-like extension and greatly increased cell spreading (Suppl. Fig. S3A). However, this transition to an S-type–like morphology did not result in an upregulation of S100α6 or any changes in the levels of NF68. Instead, α-SMA and vimentin, both markers of mesenchymal cells, were increased (Suppl. Fig. S3B).

As expected for S-type cells, both TR1 and TR2 expressed high levels of S100α6 (Suppl. Fig. S3B). By immunofluorescence analysis, N1 cells did not show positive staining for S100α6 (Suppl. Fig. S3C). In contrast, TR1 displayed strong staining for S100α6 in the cytosol, which was unlike the predominantly nuclear localization previously reported for other S-type cells (Suppl. Fig. S3C).²⁹ Furthermore, both TR1 and TR2 expressed high levels of α-SMA and vimentin. Treatment of TR1 and TR2 with RA failed to induce significant changes in cell morphology (Suppl. Fig. S3A), although S100α6 levels were drastically reduced to almost undetectable levels (Suppl. Fig. S3B). Finally, BrdU did not appear to cause any appreciable changes in cell shape (Suppl. Fig. S3A). Also, the levels of S100α6 were not greatly affected by this treatment (Suppl. Fig. S3B). Thus, the neuroblastic N-type cells still retain the capacity to undergo morphological differentiation, while the S-type sublines are relatively refractory to the induction of differentiation.

Survival-promoting properties of TR sublines

While the ability of S-type cells in promoting the survival of neuroblastic tumor cells has been reported,³⁰ how this cross-talk could affect the therapeutic efficacy of the ALK inhibitor was not studied. Taking advantage of the selective inhibition of N-type cells by TAE684, the ability of S-type cells in conferring survival on N-type cells was tested. Cocultivation experiments were carried out by mixing N1 and TR1 at a ratio of 3:1. To selectively analyze the N1 population, TR1 was labeled with CFSE. Next, cultures were treated with 30 nM TAE684 for 48 hours, a dosage that was substantially below the IC₅₀ for the TR sublines. By annexin V (AxV) staining, N1 has a basal apoptotic fraction of 10.2% ± 1.1%. Coculturing with TR1 effectively reduced this to 4.2% ± 0.9% (Fig. 4A and 4B). The addition of TAE684 increased the AxV⁺ fraction to 20.3% ± 3.6%. Strikingly, the presence of TR1 significantly reduced this to 4.3% ± 0.7%. Similarly, TAE684-induced late apoptotic events (AxV⁺PI⁺ cells) in N1 cells were also greatly attenuated from 32.5% ± 6.8% to 9.7% ± 2.2% when TR1 was present (Fig. 4A and 4B).

To test if similar prosurvival activities could be observed in additional S-type cells, TR2 was tested. In this experiment, a higher dose (50 nM) of TAE684 and a longer treatment duration of 72 hours were selected. Under this condition, TAE684 induced an AxV⁺ fraction of 62.5% in N1 cells cultured alone. Coculturing with TR2 drastically attenuated this fraction to 12.9% (Fig. 4C). To determine if S-type cells isolated from parental SK-N-SH could also promote survival, the S1 subline was examined. Coculturing TAE684-treated N1 with S1 reduced the combined fractions of nonviable cells (AxV⁺ and AxV⁺PI⁺) from 33.5% to 15.1% (Fig. 4D). Taken together, these results show that S-type cells exert chemoprotective effects on N-type cells.

Figure 4.

Survival-promoting effect of TR cells. (A) Approximately 3 × 10⁵ N1 cells were comixed with or without 1 × 10⁵ CFSE (0.75 µM)–labeled TR1 on a 100-mm plate. Cultures were treated with or without 30 nM TAE684 for 48 hours. Cells were collected and stained with APC-annexin V and PI. CFSE-negative populations were gated for analysis, and the fractions of AnV⁺, PI⁺, and AnV⁺PI⁺ were indicated in the FACS plots. (B) Results from 3 independent experiments were quantified. Error bars = SEM. *P < 0.05; **P < 0.005. (C) A similar experiment was performed by coculturing N1 and CFSE-labeled TR2. Cells were treated with or without 50 nM TAE684 for 72 hours. Cell death assay was performed, as described (A). (D) N1 and CFSE-labeled S1 were cocultured and treated with or without 30 nM TAE684 for 48 hours, as described (A). Samples were analyzed for nonviable cells by FACS analysis.

Synergistic activation of STAT3 and AKT

To investigate if the observed prosurvival effects were correlated with the activation of specific signaling pathways, N- and S-type sublines were comixed at a higher ratio (27:1) to facilitate the detection of activation events in N-type cells. On average, a single S-type cell could establish direct contacts with 10 to 20 N-type cells. As shown in Figure 5A, both N1 and N2 cells have almost undetectable p-STAT3. Although STAT3 was phosphorylated in TR1 (Fig. 2A), the low number of cells plated rendered it below the threshold of detection (Fig. 5A). Comixing drastically elevated the levels of p-STAT3 in the cocultures. Similarly, a 2- to 3-fold increase in p-AKT was also observed. However, p-ERK1/2 levels were not altered, affirming the specificity of this signaling synergism. To demonstrate that these cooperative signaling events could be mediated by other S-type cells, we comixed N1 with either S1 or TR2. In this experiment, cocultivation induced a striking synergistic upregulation of both p-STAT3 and p-AKT by at least 20-fold (Fig. 5B). Next, whether similar signaling cross-talk could be observed in an independent pair of N- and S-type cells was investigated. For this, LA1-55n and LA1-5S, which were, respectively, N- and S-type cells derived from the same patient, were cocultured.^31,32 As expected, comixing LA1-55n with LA1-5S resulted in a synergistic activation of STAT3, although the effect on AKT was more additive in nature, while ERK1/2 was not upregulated (Fig. 5C).

Although S-type cells were underrepresented in the mixed cultures, it was plausible that N-type cells could have altered signaling events in S-type cells. To exclude this possibility, conditioned medium from N1 was added to TR2 and S1 cells for 2 days. As predicted, N1-conditioned medium failed to induce phosphorylation of STAT3, AKT, or ERK1/2 (Fig. 5D). To gain further insights into the cell types responsible for the striking synergism in STAT3 activation, we sought to determine its transcriptional response by a luciferase-based reporter assay. N1 and TR1 were separately transfected with the pTATA-tk-Luc reporter construct, followed by comixing with untransfected counterparts. As shown in Figure 5E, both N1 and TR1 transfected with the control or pTATA-tk-Luc reporter plasmids elicited modest responses. However, comixing N1 with TR1 induced a 4.6-fold stimulation of STAT3 reporter readout in N1, while it was 2.3-fold in TR1 cells. These differences were unlikely to be the result of variation in transfection efficiency, as the data were normalized with the control Renilla luciferase reporter. We conclude from these data that the juxtaposing of N- and S-type cells stimulates STAT3 transcriptional activity mostly in N1 cells.

Figure 5.

Synergistic activation of signaling pathways. (A-C) Sublines were comixed at the indicated combinations in 6-well plates for 48 hours. The cell numbers used are shown in the bottom sections. Individual wells were solubilized in 135 µL of 2x Laemmli buffer with β-mercaptoethanol. Total cell extracts were sonicated, and equal volumes were subjected to Western blotting analysis with the antibodies indicated. (D) Conditioned medium (CM) at the indicated dilutions from N1 was added to TR2 and S1 for 48 hours. Cultures were solubilized for Western blotting analysis, as described (A-C). (E) 2 × 10⁴ N1 and 5 × 10³ TR1 cells were plated per well in triplicate of 96-well plates and transfected with 0.2 µg of either pGL2 or pTATA reporter plasmids. pRL (5 ng) control plasmid was cotransfected in all wells. Untransfected N1 and TR1 cells were seeded into selected wells as indicated. After 48 hours, relative luciferase activity was measured using a dual luciferase assay kit (Promega, Madison, WI). Error bars = SD. Similar results were obtained in a separate experiment.

Conditioned media from S-type cells promote cell viability

The ability of tumor stroma in modulating cancer cell behavior through paracrine mechanisms has been widely documented. For instance, primary human Schwann cells have been shown to secrete factors that inhibit endothelial cell proliferation and migration.³³ To test if S-type sublines isolated in this study could produce growth- or survival-promoting factors, supernatants from various cultures were collected after 2 days of conditioning. Conditioned media (CM) from TR1, TR2, and S1 cell lines all stimulated an increase in cell numbers by approximately 2-fold when tested on N1 cells. This effect could be titrated down with serial dilutions. Treatment with TAE684 resulted in a 50% reduction in the number of viable cells. The addition of CM, especially from TR2 and S1, restored cell viability to levels of the untreated cultures (Fig. 6A). Consistent with these findings, the fraction of AxV⁺ apoptotic cells was drastically diminished from 25.4% to 9.36% with the addition of CM in TAE684-treated N1 cells (Fig. 6B).

A high fraction of trophic factors, including pleiotrophin, the putative ligand for ALK, are heparin-binding proteins.³⁴ The importance of heparin-binding factors in conferring cell viability was tested. For this, CM was preincubated with control or heparin sepharose. As demonstrated in Figure 6C, CM incubated with control beads caused an approximately 1.5-fold increase in the number of N1 and N2 cells. However, preclearing CM of heparin-binding proteins reduced this survival-promoting effect by 50% and 88% in N1 and N2 cells, respectively.

Figure 6.

Conditioned media from S-type cells promote cell viability. (A) Supernatants of TR1, TR2, and S1 cells conditioned for 2 days in 10% FBS were collected. As a negative control, medium was incubated under the same condition on a blank plate. Conditioned media (CM) were added at the indicated dilutions to 1 × 10⁵ N1 cells in the presence or absence of 30 nM TAE684 for 48 hours. The number of viable cells was counted by a hematocytometer. Results were generated from triplicated wells and repeated at least once. Error bars = SD. *P < 0.05; **P < 0.005; ***P < 0.0005. (B) N1 cells were treated without or with 50 mM of TAE684 for 72 hours in the absence or presence of CM from TR3. The fraction of apoptotic cells was analyzed by flow cytometry as described in Figure 4. (C) Similar proliferation experiments were performed, as described (A), except that TR3 CM were preincubated with blank (Blk) beads or heparin (Hep) sepharose for an hour. Results from a single experiment performed in triplicate. Error bars = SD. *P < 0.05.

To examine if enhanced cell viability correlated with the activation of key signaling pathways, N1 cells treated with CMs were subjected to Western blotting analysis. As shown in Figure 7A, CM from TR1, TR2, and S1 all induced a robust activation of AKT in a dose-dependent manner. However, none of the CMs tested was able to activate either STAT3 or ERK1/2 (Fig. 7A). To evaluate the relevance of AKT activation, the ability of CM in promoting cell viability was tested in the presence of pharmacological inhibitors. The addition of either LY294004 or UO126, which inhibited PI3-K and ERK1/2, respectively, to N1 cells at 5 µM did not cause a significant decrease in cell viability. However, LY294004 significantly reduced the ability of CM from TR1 in promoting cell viability. On the contrary, UO126 failed to cause any significant perturbation (Fig. 7B). Western blotting analysis of cell extracts from control and treated cultures clearly demonstrated the specificity of these inhibitors (Fig. 7C). Thus, these results indicate that the ability of secretory factors from S-type cells in stimulating cell viability of N-type tumor cells is dependent on PI3-K.

Figure 7.

Conditioned media from S-type cells stimulate cell viability through PI3-K. (A) Conditioned media (CM) from TR1, TR2, and S1 were added to N1 cells for 2 days. Western blotting analysis was performed using the antibodies indicated. (B) 3 × 10⁴ N1 cells were plated per well in 6-well plates. CM was added in the presence or absence of LY294002 (LY, 5 µM) and UO126 (UO1, 5 µM) as indicated. After 72 hours, cell numbers were counted by a hematocytometer. Results represent data from 3 independent experiments. Error bars = SEM. **P < 0.005; ***P < 0.0005. (C) Cells were plated, as described (B), and were solubilized 30 minutes after the addition of LY and UO1. Western blotting analysis was carried out using the indicated antibodies.

Discussion

The present study identifies a subpopulation of NB cells that are insensitive to the ALK kinase inhibitor, TAE684. These NB sublines have Schwann cell–like features, possess unique signaling profiles, and confer prosurvival properties to N-type tumor cells. The S1 subline, which harbors features of the TR cells, would suggest the pre-existence of TAE68-insensitive cells in the parental culture. Treatment of inflammatory myofibroblastic tumor with the ALK kinase inhibitor, crizotinib (PF-02341066), has resulted in the development of drug resistance due to the acquisition of a F1174L mutation in the ALK gene.³⁵ Our results prove against this type of drug-resistant mechanism. Instead, the lack of sensitivity towards TAE684 in the TR sublines is the result of a lack of ALK expression. The silenced expression of ALK in TR cells may reflect its Schwann cell origin. This is consistent with the finding that ALK is expressed in neurons but not in satellite Schwann cells of E18 dorsal root ganglion.³⁶

The present study has shed light on the origin of S-type cells in SK-N-SH. While there are controversies concerning the origin of tumor-associated Schwann cells in NB, the finding of the ALK F1174L mutation in TR1 cells would indicate that they are tumor derived. This assertion is consistent with the finding of the same 3 marker chromosomes in both SH-SY5Y and SH-EP cells, which are N- and S-type cells, respectively.²⁷ In addition, analysis of paired N- and S-type cells from primary NB for allelic imbalance and loss of heterozygosity has arrived at the same conclusion.¹¹ More interestingly, because SK-N-SH was isolated from bone marrow metastases, we speculate that tumor cells seeded in the bone marrow must have retained the capacity to differentiate into neuroblastic and Schwann cell types. However, it is not clear if the diverse cell types found in cell cultures are phenocopies of the cellular heterogeneity found in primary NB. Finally, we cannot exclude the possibility that the ALK F1174L mutation in SK-N-SH is in fact a germline mutation associated with familial NB syndrome. In this regard, the lack of archived pathological specimens hampers efforts in addressing this possibility.

Our findings have clinical implications in treating NB with ALK inhibitors. It is envisaged that anti-ALK therapy would eradicate most ALK-positive neuroblastic tumor cells while sparing tumor-derived S-type cells. Given the propensity to interconvert between different NB subtypes,²⁷ it is plausible that residual stromal cells may transdifferentiate to N-type tumor cells and contribute to tumor relapse. It is also noteworthy that S-type cells only account for a minor percentage (~5%) of the parental SK-N-SH cell line. From a histological standpoint, SK-N-SH resembles a stroma-poor NB that is refractory to conventional chemotherapies with a high risk of recurrence. It is tempting to speculate that tumor-derived stromal cells could form a niche to promote survival of N-type tumor cells. The formation of such a chemoresistant niche has been reported in the thymus following the chemotherapy of Burkitt lymphoma.³⁷

The roles of STAT3 in the pathogenesis and chemoresistance of NB have been extensively reported.^38-41 Also, the ability of oncogenic ALK mutants in activating STAT3 has been demonstrated in lymphoma cells.^42,43 Therefore, it is unexpected that ALK-positive N-type cells would have much weaker STAT3 phosphorylation than ALK-negative TR cells. Paradoxically, parental SK-N-SH with predominant N-type cells has robust STAT3 phosphorylation. Using transcriptional reporter assays, we have provided evidence that cocultivation leads to STAT3 activation predominantly in N-type cells. Thus, these results highlight the importance of juxtaposing N- and S-type cells for optimal STAT3 activation. This signaling mechanism may be crucial for SK-N-SH in establishing metastatic growth in the bone marrow. In other words, the coevolution of N- and S-type cells in close proximity may provide a prosurvival milieu in an otherwise hostile bone marrow microenvironment.

Soluble factors produced by Schwann cells capable of promoting survival and differentiation of N-type cells have been reported.³⁰ In addition, pigment epithelial derived factor (PEDF)⁴⁴ and secreted protein acidic and rich in cysteine (SPARC)³³ isolated from Schwann cell–conditioned media have been shown to possess antiangiogenic activity towards endothelial cells. Although we have not tested conditioned media for antiangiogenic activity, their ability in stimulating cell viability is reproducibly observed. Also, conditioned media from S-type cells are effective in protecting N-type cells from the apoptotic effects of TAE684. As for the signaling events, while conditioned media from S-type cells only activate AKT in N-type cells, comixing of these 2 cell types results in the activation of both AKT and STAT3. These results suggest that stromal cells can exert both short- and long-range instructive effects on N-type tumor cells by activating distinct signaling pathways. The biological significance of this is not clear, but one speculation is that a gradient of survival signaling may restrict the size or differentiation states of neuroblastic tumor cells in a newly colonized metastatic site.

From a therapeutic standpoint, blocking cross-talk between N- and S-type cells may disrupt signaling events essential for tumor maintenance and survival. Alternatively, eradicating tumor-derived S-type cells with selective pathway inhibitors following anti-ALK therapy may limit tumor recurrence. Overall, results presented here are generated from a single NB cell line and are clearly not representative of all NB. However, to our knowledge, similar studies characterizing signaling cross-talk between N- and S-type cells have not been reported. Therefore, the present work provides a solid framework for future research aiming at deciphering the signaling synergism between stromal and neuroblastic tumor cells.

Materials and Methods

Chemicals

NVP-TAE684 (TAE684) was purchased from Axon Medchem (Groningen, the Netherlands) and dissolved in dimethylsulphoxide (DMSO) as a 3 mM stock. LY294002 and UO126 were obtained from EMD Biosciences (Gibbstown, NJ). Retinoic acid (RA) and 5-bromo-2-deoxyuridine (BrdU) were purchased from Sigma-Aldrich (St. Louis, MO).

Cell lines

SK-N-SH was obtained from Dr. Stuart Aaronson (Mount Sinai Medical Center, New York, NY). LA1-55n and LA1-5S cell lines were provided by Dr. Robert Ross and Barbara Spengler (Fordham University, New York, NY). Neuroblastic sublines, N1 and N2, and the substrate-adherent subline, S1, were isolated based on morphological criteria using cloning cylinders. To isolate TR1, TR2, TR3, and TR4 sublines, the parental SK-N-SH cell line was exposed sequentially to an increasing concentration (30, 100, and 600 nM) of TAE684 for a 7-day period. Prior to dose escalation, cells were allowed to expand and replate at low cell density. Individual colonies that survived the selection were isolated by cloning cylinders. All cultures were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA).

Karyotype analysis

Chromosome analysis was conducted by the Wisconsin State Laboratory of Hygiene (University of Wisconsin, Madison, WI). G-banded chromosome analysis was carried out on SK-N-SH, TR1, and TR2. A total of 20 metaphase cells were used for each subline.

Antibodies

Antibodies for p-ALK Tyr1604 (#3341), ALK (#3342), p-STAT3 Tyr705 (#9138), STAT3 (#9132), p-Akt Ser473 (#4051), Akt (#9272), p-Erk1/2 (#9106), and Erk1/2 Thr202/Tyr204 (#9102) were obtained from Cell Signaling Technology (Danvers, MA). Antibodies for NF68 (Ab72995), S100α6 (Ab55680), and vimentin (Ab7783) were purchased from Abcam (Cambridge, MA). Mouse monoclonal antibody against α-smooth muscle actin (#A5228) was obtained from Sigma-Aldrich. Horseradish peroxidase (HRP)–conjugated anti-actin (sc-1616), HRP anti-rabbit IgG (sc-2313), and HRP anti-mouse IgG (sc-2314) were from Santa Cruz Biotechnology (Santa Cruz, CA).

Western blotting

Total cell extracts were prepared by solubilizing cells in Laemmli buffer (LB) with 10% β-mercaptoethanol (BME) and boiled for 5 minutes and sonicated to reduce viscosity. Equal volume of lysate was subjected to electrophoresis on 8% SDS-PAGE gels, and proteins resolved were transferred onto nitrocellulose membrane (Bio-Rad, Hercules, CA). Filter membranes were blocked in 5% milk in TTBS for 30 minutes and incubated with primary antibodies at 1:100 to 1:5,000 dilutions. Bound antibodies were incubated with either HRP-conjugated anti-mouse or anti-rabbit secondary antibodies at 1:5,000 dilutions followed by electrogenerated chemiluminescence (ECL) (Pierce, Rockford, IL) detection on X-ray films (Denville, Metuchen, NJ).

Mutation analysis

Genomic DNAs were extracted by DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA) from selected SK-N-SH sublines. Exon 23 of the human ALK gene was amplified using 2 primers, ALK7: 5′tgcctttatacatt gtagctgctg3′ and ALK8: 5′aactgcagcaaagactggttctca3′. The 274–base pair PCR fragment was purified and sequenced with a nested primer, ALK9: 5′ctgttcctcccagtttaagatttg3′, using a commercial source (Eton, San Diego, CA).

RT-PCR

Around 1 µg of total RNA was transcribed to cDNA with a SuperScript III First Strand Kit (Invitrogen) according to the manufacturer’s instructions. Approximately 1 µL of cDNA that was amplified with 2 sets of ALK primer pairs: ALK16, 5′aaacatcaccctcattcggggtct3′ (sense), and ALK18, 5′caatgttctggtggtttaatttgctgatgatcagg3′ (antisense); and ALK25, 5′gccacacctgccactctcgctgat3′ (sense), and ALK23, 5′ttccgcggcacctccttcaggtca3′ (antisense). For normalization, actin cDNA was amplified using a pair of actin primers: actinF1, 5′cccacactgtgcccatctacg3′ (sense), and actinR1, 5′gcttctccttaatgtcacgc3′ (antisense).

Proliferation assay

Approximately 5 × 10⁴ N-type and 1 × 10⁴ TR-type cells were seeded in each well of a 12-well plate in triplicates. Wells were treated without or with either conditioned media or TAE684. Cell numbers were determined by a hematocytometer at selected time points.

Cell viability assay

Cell number estimation was determined indirectly by the [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt] (MTS) assay (Promega, Madison, WI). Approximately 8 × 10³ N-type cells or 3 × 10³ S-type cells were plated per well of a 96-well plate. Two to 3 days after treatments, cells were incubated with 20 µL of MTS/PMS reagents in a final volume of 120 µL of phenol red–free DMEM. Absorbance at 490 nm was measured after 1 to 2 hours with an iMark microplate reader (Bio-Rad).

Apoptosis assay

Approximately 1 × 10⁵ S-type cells were labeled with 0.75 µM of carboxyfluorescein succinimidyl ester (CFSE) for 15 minutes at 37°C followed by 3 washes with DMEM + 10% FBS. Labeled cells were cocultured with 3 × 10⁵ N-type cells on 100-mm tissue culture dishes (BD, Franklin Lakes, NJ) in the presence or absence of TAE684 (30-50 nM). After 2 to 3 days, cells were harvested and costained with propidium iodide (PI) and APC-conjugated annexin V (AxV) (BD). Non-CFSE N-type cells were gated by flow cytometry for analysis for apoptotic and necrotic populations.

Transcriptional reporter assay

Approximately 2 × 10⁴ N1 cells or 5 × 10³ TR1 cells were seeded per well of a 96-well plate in triplicates. Individual wells were transfected with 0.2 µg of pGL2 control or STAT3 reporter plasmid, pTATA-tk-Luc, a gift from Dr. Jacqueline Bromberg (Memorial Sloan-Kettering, New York, NY). All wells were cotransfected with 5 ng of pRL using 0.5 µL of Lipofectamine 2000 (Invitrogen). After 24 hours, same numbers of untransfected N1 and TR1 cells were added to comix with the heterologous cell types. Following incubation for an additional 48 hours, cells were solubilized in 20 µL of passive lysis buffer, and the relative luciferase activity in each well was measured using the dual luciferase assay kit (Promega). Data were normalized to the signals emitted from the Renilla luciferase.

Conditioned medium

TR sublines were cultured in 100-mm tissue culture plates to near confluence in 7 mL of 10% FBS/DMEM. Equal volume of 10% FBS in DMEM was placed in a blank plate to serve as a negative control. After 48 hours, conditioned media were collected, centrifuged at 3,500 rpm for 15 minutes, and used in proliferation assays. Alternatively, 5 mL of conditioned media were incubated with either 500 µL of CL-4B 200 control beads or heparin sepharose (Sigma-Aldrich) for 1 hour at 4°C to preclear heparin-binding proteins.

Footnotes

Acknowledgements

The authors thank Dr. Stuart Aaronson (Mount Sinai, NY), Dr. Robert Ross, Dr. Barbara Spengler (Fordham University, NY), and Dr. Jason Jarzembowski (Medical College of Wisconsin, WI) for the gifts of neuroblastoma cell lines. They also appreciate the assistance of Dr. Suresh Kumar at the imaging core at the Medical College of Wisconsin and Dr. Guan Chen (Medical College of Wisconsin, WI) for critical reading of the article.

Supplementary material for this article is available on the Genes & Cancer website at .

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

This work was supported by the MACC Fund, Advancing a Healthier Wisconsin, Wisconsin Breast Cancer Showhouse (O.W.), National Institutes of Health (NIH) [grant number CA133669] (H.G. and A.M.C.), and the Children’s Research Institutes of the Children’s Hospital of Wisconsin.

References

Ambros

Brodeur

. International consensus for neuroblastoma molecular diagnostics: report from the International Neuroblastoma Risk Group (INRG) Biology Committee. Br J Cancer. 2009;100:1471-82.

Maris

Hogarty

Bagatell

Cohn

. Neuroblastoma. Lancet. 2007;369:2106-20.

Ciccarone

Spengler

Meyers

Biedler

Ross

. Phenotypic diversification in human neuroblastoma cells: expression of distinct neural crest lineages. Cancer Res. 1989;49:219-25.

Biedler

Roffler-Tarlov

Schachner

Freedman

. Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones. Cancer Res. 1978;38:3751-7.

Ross

Spengler

Domenech

Porubcin

Rettig

Biedler

. Human neuroblastoma I-type cells are malignant neural crest stem cells. Cell Growth Differ. 1995;6:449-56.

Nagoshi

Tsuneyoshi

Enjoji

. S-100 positive undifferentiated neuroblastomas with a special reference to the tumor stroma related to favorable prognosis. Pathol Res Pract. 1992;188:273-83.

Liu

Tian

Chlenski

Yang

Salwen

Cohn

. “Cross-talk” between Schwannian stroma and neuroblasts promotes neuroblastoma tumor differentiation and inhibits angiogenesis. Cancer Lett. 2005;228:125-31.

Chlenski

Guerrero

Peddinti

. Anti-angiogenic SPARC peptides inhibit progression of neuroblastoma tumors. Mol Cancer. 2010;9:138.

Liu

Tian

Chlenski

. Cross-talk between Schwann cells and neuroblasts influences the biology of neuroblastoma xenografts. Am J Pathol. 2005;166:891-900.

10.

Katsetos

Karkavelas

Frankfurter

. The stromal Schwann cell during maturation of peripheral neuroblastomas: immunohistochemical observations with antibodies to the neuronal class III beta-tubulin isotype (beta III) and S-100 protein. Clin Neuropathol. 1994;13:171-80.

11.

Mora

Cheung

Juan

. Neuroblastic and Schwannian stromal cells of neuroblastoma are derived from a tumoral progenitor cell. Cancer Res. 2001;61:6892-8.

12.

Chen

Takita

Choi

. Oncogenic mutations of ALK kinase in neuroblastoma. Nature. 2008;455:971-4.

13.

George

Sanda

Hanna

. Activating mutations in ALK provide a therapeutic target in neuroblastoma. Nature. 2008;455:975-8.

14.

Janoueix-Lerosey

Lequin

Brugieres

. Somatic and germline activating mutations of the ALK kinase receptor in neuroblastoma. Nature. 2008;455:967-70.

15.

Mosse

Laudenslager

Longo

. Identification of ALK as a major familial neuroblastoma predisposition gene. Nature. 2008;455:930-5.

16.

Caren

Abel

Kogner

Martinsson

. High incidence of DNA mutations and gene amplifications of the ALK gene in advanced sporadic neuroblastoma tumours. Biochem J. 2008;416:153-9.

17.

De Brouwer

De Preter

Kumps

. Meta-analysis of neuroblastomas reveals a skewed ALK mutation spectrum in tumors with MYCN amplification. Clin Cancer Res. 2010;16:4353-62.

18.

Passoni

Longo

Collini

. Mutation-independent anaplastic lymphoma kinase overexpression in poor prognosis neuroblastoma patients. Cancer Res. 2009;69:7338-46.

19.

De Brouwer

De Preter

Kumps

. Meta-analysis of neuroblastomas reveals a skewed ALK mutation spectrum in tumors with MYCN amplification. Clin Cancer Res. 2010;16:4353-62.

20.

Iwahara

Fujimoto

Wen

. Molecular characterization of ALK, a receptor tyrosine kinase expressed specifically in the nervous system. Oncogene. 1997;14:439-49.

21.

Morris

Naeve

Mathew

. ALK, the chromosome 2 gene locus altered by the t(2;5) in non-Hodgkin’s lymphoma, encodes a novel neural receptor tyrosine kinase that is highly related to leukocyte tyrosine kinase (LTK). Oncogene. 1997;14:2175-88.

22.

Palmer

Vernersson

Grabbe

Hallberg

. Anaplastic lymphoma kinase: signalling in development and disease. Biochem J. 2009;420:345-61.

23.

Galkin

Melnick

Kim

. Identification of NVP-TAE684, a potent, selective, and efficacious inhibitor of NPM-ALK. Proc Natl Acad Sci U S A. 2007;104:270-5.

24.

Christensen

Zou

Arango

. Cytoreductive antitumor activity of PF-2341066, a novel inhibitor of anaplastic lymphoma kinase and c-Met, in experimental models of anaplastic large-cell lymphoma. Mol Cancer Ther. 2007;6:3314-22.

25.

Rodig

Shapiro

. Crizotinib, a small-molecule dual inhibitor of the c-Met and ALK receptor tyrosine kinases. Curr Opin Investig Drugs. 2010;11:1477-90.

26.

Gorre

Mohammed

Ellwood

. Clinical resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification. Science. 2001;293:876-80.

27.

Ross

Spengler

Biedler

. Coordinate morphological and biochemical interconversion of human neuroblastoma cells. J Natl Cancer Inst. 1983;71:741-7.

28.

Mazot

Cazes

Boutterin

. The constitutive activity of the ALK mutated at positions F1174 or R1275 impairs receptor trafficking. Oncogene. 2011;30:2017-25.

29.

Acosta

Lavarino

Paris

. Comprehensive characterization of neuroblastoma cell line subtypes reveals bilineage potential similar to neural crest stem cells. BMC Dev Biol. 2009;9:12.

30.

Kwiatkowski

Rutkowski

Yamashiro

Tennekoon

Brodeur

. Schwann cell-conditioned medium promotes neuroblastoma survival and differentiation. Cancer Res. 1998;58:4602-6.

31.

Spengler

Lazarova

Ross

Biedler

. Cell lineage and differentiation state are primary determinants of MYCN gene expression and malignant potential in human neuroblastoma cells. Oncol Res. 1997;9:467-76.

32.

Reynolds

Biedler

Spengler

. Characterization of human neuroblastoma cell lines established before and after therapy. J Natl Cancer Inst. 1986;76:375-87.

33.

Chlenski

Liu

Crawford

. SPARC is a key Schwannian-derived inhibitor controlling neuroblastoma tumor angiogenesis. Cancer Res. 2002;62:7357-63.

34.

Stoica

Kuo

Powers

. Midkine binds to anaplastic lymphoma kinase (ALK) and acts as a growth factor for different cell types. J Biol Chem. 2002;277:35990-8.

35.

Sasaki

Okuda

Zheng

. The neuroblastoma-associated F1174L ALK mutation causes resistance to an ALK kinase inhibitor in ALK-translocated cancers. Cancer Res. 2010;70:10038-43.

36.

Degoutin

Brunet-de Carvalho

Cifuentes-Diaz

Vigny

. ALK (anaplastic lymphoma kinase) expression in DRG neurons and its involvement in neuron-Schwann cells interaction. Eur J Neurosci. 2009;29:275-86.

37.

Gilbert

Hemann

. DNA damage-mediated induction of a chemoresistant niche. Cell. 2010;143:355-66.

38.

Ara

Song

Shimada

. Interleukin-6 in the bone marrow microenvironment promotes the growth and survival of neuroblastoma cells. Cancer Res. 2009;69:329-37.

39.

Song

Ara

. Oncogene MYCN regulates localization of NKT cells to the site of disease in neuroblastoma. J Clin Invest. 2007;117:2702-12.

40.

Hecht

Papoutsi

Tran

Wilting

Schweigerer

. Hepatocyte growth factor/c-Met signaling promotes the progression of experimental human neuroblastomas. Cancer Res. 2004;64:6109-18.

41.

Rebbaa

Chou

Mirkin

. Factors secreted by human neuroblastoma mediated doxorubicin resistance by activating STAT3 and inhibiting apoptosis. Mol Med. 2001;7:393-400.

42.

Zamo

Chiarle

Piva

. Anaplastic lymphoma kinase (ALK) activates Stat3 and protects hematopoietic cells from cell death. Oncogene. 2002;21:1038-47.

43.

Khoury

Medeiros

Rassidakis

. Differential expression and clinical significance of tyrosine-phosphorylated STAT3 in ALK+ and ALK- anaplastic large cell lymphoma. Clin Cancer Res. 2003;9:3692-9.

44.

Crawford

Stellmach

Ranalli

. Pigment epithelium- derived factor (PEDF) in neuroblastoma: a multifunctional mediator of Schwann cell antitumor activity. J Cell Sci. 2001;114:4421-8.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

4.01 MB