Constitutively Active NFκB Is Required for the Survival of S-type Neuroblastoma*

The NFκB transcription factors can both promote cell survival and induce apoptosis depending on cell type and context. Neuroblastoma (NB) cells display two predominant culture phenotypes identified as N- and S-types. Malignant S-type cells express neither high levels of MYCN nor Bcl-2, suggesting that other survival mechanisms are important. We characterized NFκB activity in S-type cells and determined its role in their survival. S-type lines (SH-EP1 and SK-N-AS) were treated with pyrrolidine dithiocarbamate (PDTC), a NFκB inhibitor, orl-1-tosylamido-2-phenylethyl chloromethyl ketone (TPCK), a serine protease inhibitor that blocks IκBα degradation. Both agents induced cell death, suggesting that constitutive NFκB activity is required for survival. The transient expression of a super-repressor IκBα mutant killed S-type cells. The inhibition of NFκB produced an apoptotic response characterized by the collapse of the mitochondrial transmembrane electrochemical gradient, caspase-9 activation, and apoptotic DNA changes. Constitutive NFκB DNA binding activity specifically involving p65 and p50 was demonstrated in S- but not N-type cells by electromobility supershift and gene reporter assays. This study demonstrates a role for NFκB in the survival of S-type NB tumor cells and suggests that NFκB activity and function differ according to NB tumor cell phenotype.

Nuclear factor-B (NFB) 1 is a transcription factor comprised of heterodimers and homodimers of a family of related proteins including NFB1/p50, NFB2/p52, RelA/p65, RelB, and c-Rel (1,2). In the absence of appropriate signals, NFB is sequestered in the cytoplasm by association with one of several inhibitory IB species (1). Its release from IB and translocation to the nucleus results from activation of IB kinase (IKK) (3). NFB transcriptionally regulates an array of genetic responses, many of which determine the balance between cell survival and death. Its activity is particularly important for cells to resist apoptosis triggered by surface death receptors (4). As an illustration, mice lacking p65 die at embryonic day 12 as a result of massive liver apoptosis triggered by TNF-␣ (5). However, the nature of actions of NFB on the balance between life and death is complex because NFB also mediates cell death. For example, wild-type p53 induces activation of NFB in response to genotoxic stress, and inhibition of NFB activity under these circumstances abrogates p53-mediated apoptosis (6).
NB is a malignant tumor that develops in children, arising from neoplastic cells of neural crest origin (7). Unfortunately, the cure rate of children with high risk NB remains at Ͻ20%, providing a compelling reason to better understand the molecular mechanisms that can be targeted to treat this disease (8,9). The study of this disease is further complicated by the fact that NB tumors are composed of a mixture of histologically distinct malignant cell types identified as neuroblastic and stromal cells (10). In culture, tumor cell explants reflect this heterogeneity in that cells demonstrate one of the three phenotypes: neuronal (N)-, stromal (S)-, or intermediate (I)-type. N-type cells contain enzymes involved in neurotransmitter synthesis, grow as poorly adherent aggregates of small round cells with neuritic processes, and express cell surface receptors characteristic of embryonic neuronal precursors. S-type cells grow flattened and adherent to substrate and are devoid of the synthetic machinery for neurotransmitter synthesis (11). I-type cells share certain characteristics with S-and N-type cells and can be induced to further differentiate along either lineage (12).
Compared with S-type cells, N-type cells are more tumorigenic in immunodeficient mice and express higher levels of the tumor marker MYCN and the anti-apoptotic protein Bcl-2 (13). In human NB, disease prognosis worsens as the proportion of neuroblastic cells in a tumor increases. These characteristics have made N-type cells the primary focus of studies of NB chemotherapy responsiveness. In previous work, we reported that N-type cells respond to doxorubicin, an agent used to treat this disease, by inducing NFB activation and that NFB is necessary for the cytotoxic effects of this drug (14).
However, the heterogeneity of NB tumors presents the possibility that malignant behavior (including development, persistence, and treatment response) is influenced by the nonneuroblastic components that correspond to S-and I-type cells in culture. Despite their low tumorigenicity in xenograft mod-els, S-type cells in particular have characteristics that suggest an important role in determining disease outcome. Most importantly, recent data show that S-and N-type cells are derived from neoplastic clones that are genetically identical (15). Secondly, stromal tumor cells appear more able to persist after treatment with cytotoxic drugs. Following cytotoxic therapy, residual tumor deposits are primarily composed of mature atypical ganglion cells and stromal components and lack neuroblastic elements that predominate in specimens prior to treatment (16). Chemotherapy-induced differentiation does not correlate with more favorable outcomes, further suggesting that stromal cells have malignant capability. These observations may be partly explained by the increased expression of membrane-bound complement inhibitors such as CD59 on S-type cells (17). By resisting complement-mediated cell lysis, S-type cells evade an arm of immune surveillance as well as a mechanism that eliminates tumor cells following cytotoxic therapy (18). Because S-type cells are capable of trans-differentiating into N-type cells in vitro (19), a model of NB tumorigenesis has been proposed in which S-type cells provide a reserve pool of tumor cells that survive even when N-type cells are successfully targeted (15).
Bcl-2 and MYCN are highly expressed in N-type cells and appear to primarily account for their survival advantage and tumorigenic phenotype (20). S-type cells express at most low levels of Bcl-2 and are rarely MYCN-amplified. Consequently, their survival must depend on other molecular mechanisms that could be important targets for therapy. This report describes experiments performed to test the hypothesis that NFB supports S-type cell survival. Results show that NFB is constitutively active in two S-type NB cell lines and that this activity is necessary for their survival. Pharmacological or molecular inhibition of NFB induced apoptosis that was associated with the collapse of the mitochondrial transmembrane gradient and caspase-9 activation. NFB species comprised of p50/p50 homodimers and p65/p50 heterodimers were present in the nuclei of S-type cells but not in N-type cells. Consistent with these results, constitutive NFB-dependent gene transcription was detected in S-type cells. These findings add to the understanding of the mechanistic importance of NFB in NB and define a significant difference between S-and N-type cells with respect to NFB. Future therapeutic strategies should consider the complex behavior of NFB as an arbiter of NB cell survival.
Cell Culture and Transfections-The S-type cell lines, SH-EP1 and SK-N-AS, were cultured under 5% CO 2 at 37°C in minimal essential medium supplemented with 10% fetal bovine serum, 100 g/ml penicillin, and 100 units/ml streptomycin. NB cells were transiently transfected with the indicated expression plasmids using LipofectAMINE Plus according to the manufacturer's instructions. Cells were seeded in tissue culture plates to achieve 50% confluence. 24 h later, cells were transfected using a mixture of DNA and LipofectAMINE Plus in Opti-MEM I. 8 h following transfection, cells were supplemented with serum at a final concentration of 10%. Cells were then treated as indicated.
MTT Cytotoxicity Assay-NB cells were plated in triplicates at 10 4 cells/well in 96-well culture plates in minimal essential medium. For consecutive days following treatment, cells were incubated with MTT dye (5 mg/ml) at 37°C for 4 h and lysed in buffer containing 20% (w/v) SDS, 50% (v/v) N,N-dimethylformamide (pH 4.5). Absorbance at 600 nm (A 600 ) was determined for each well using an automated microplate reader (Biotek Instruments, Winnoski, VT). After subtraction of the background absorbance, the absorbance of treated cells was divided by that of the untreated cells to obtain the percent viability. The Student's t test was used to test the significance of the difference in cell viability for these assays.
Determination of NFB-dependent Reporter Gene Activation-S-type NB cells were transfected with 1 g of the reporter plasmid pBVIx-Luc as described. To assure identical transfection efficiency in control and treated cells, cells were replated 12 h after transfection into 12-well plates, and after attachment they were treated as indicated. After cells were harvested, luciferase activity was determined according to the manufacturer's instructions using the luciferase reagent kit (Promega, Madison, WI) and a Monolight 2010 luminometer (Analytical Luminescence Laboratory, Ann Arbor, MI). An aliquot of the same samples was subjected to protein concentration determination (Bio-Rad). Luciferase activity was then normalized to protein concentration.
Measurement of NFB Activation by Electromobility Shift Assay (EMSA)-Nuclear extracts were prepared as described previously (22) and incubated in EMSA reaction buffer containing 500 mM NaCl, 2 g/ml poly(dI⅐dC), 100 mM Tris (pH 7.6), 10 mM DTT with a 32 P-labeled oligonucleotide encompassing the consensus NFB binding site (5Ј-AGTTGAGGGGACTTTCCCAGGC-3Ј). In the antibody supershift assays, nuclear extracts were incubated with 0.2 g of anti-p65 or anti-p50 polyclonal antibody for 15 min at room temperature prior to the addition of the 32 P-labeled oligonucleotides. After incubation, samples were resolved on 6% polyacrylamide gels, dried, and exposed to BioMax MS film (Eastman Kodak Co.).
Flow Cytometric Analysis of the Mitochondrial Transmembrane Gradient (⌬M)-Cells were plated in duplicate at a density of 2 ϫ 10 5 cells/well and treated as indicated. Cells were then incubated at 37°C for 20 min with JC-1 (1 M) and harvested by trypsinization. ⌬ M was determined for 10,000 cells by flow cytometry, and the data were analyzed using CellQuest software (BD Biosciences).

Chemical Inhibitors of NFB Induce Apoptosis of S-type
Cells-To explore the role of NFB in S-type cell survival, two S-type cells lines were treated with small molecules known to inhibit NFB activation. PDTC is an antioxidant and chelator of heavy metals that blocks NFB activity by suppressing the release of IB␣ from NFB (23). TPCK inhibits NFB by block-ing IB␣ degradation (24). Both compounds dose-dependently kill SH-EP1 and SK-N-AS cells after 24 h of treatment (Fig.  1A). When visualized by interference contrast microscopy after treatment with PDTC or TPCK, cells demonstrated an apoptotic appearance including plasma membrane blebbing, nuclear condensation, and detachment from the culture flask (Fig.  1B). These findings were in sharp contrast with our previous results that showed that PDTC blocks cell death of N-type cells, suggesting that NFB activity and function might differ dramatically according to NB cell phenotype (14).
The mechanism and time course of cell death was further characterized with respect to mitochondria dysfunction and caspase activation. JC-1, a fluorescent potentiometric indicator, selectively accumulates in mitochondria and was used to monitor the mitochondrial transmembrane gradient (⌬ M ) by flow cytometry. As shown in Fig. 2A, the treatment of SH-EP1 cells with PDTC or TPCK causes ⌬ M collapse beginning within 4 h of treatment. Caspase-9 activation, which depends on apoptosome formation that is in turn triggered by cytochrome c release from mitochondria (25), was also detected in treated cells. Both drugs the induced processing of procaspase-9 to the active 35-kDa form between 8 and 12 h after treatment (Fig. 2B). Thus, its activation is slightly delayed relative to ⌬ M collapse. Of note, caspase-9 is processed into two forms in the cells treated with TPCK, corresponding to molecular masses of 35 and 37 kDa. In PDTC-treated cells, only the 35-kDa form is detected. This difference may be related to serine protease inhibition by TPCK that can result in altered caspase processing (26). Together, these data are consistent with an apoptotic response induced by PDTC and TPCK that involves mitochondria-associated signals beginning within 4 h following treatment. Identical changes in ⌬ M and in caspase processing were detected in SK-N-AS cells after these treatments (data not shown).
NFB Inactivation Is Associated with the S-type Response to PDTC or TPCK-Because IB␣ levels inversely correlate with NFB activity, its levels were measured in SH-EP1 cells during treatment to directly correlate the death response with the inhibition of NFB. As expected, increased IB␣ levels in SH-EP1 cells were detected by 8 h of treatment with PDTC and 4 h of treatment with TPCK (Fig. 3A). Similar increases were also observed in SK-N-AS cells after treatment with these agents (data not shown). To confirm the increase in IB␣ reflects diminished NFB activity, NFB-dependent gene expression was measured using a NFB-dependent luciferase reporter construct. SH-EP1 and SK-N-AS cells were transiently transfected with pBVIx-Luc, which contains six tandemly placed NFB consensus binding sites within the promoter region upstream of sequence encoding firefly luciferase. Using this construct, constitutive NFB-dependent luciferase activity was detectable in both S-type cell lines in the absence of any treatment (Fig. 3B). When cells were treated with PDTC or TPCK, luciferase activity decreased in a time-dependent fashion beginning within 4 h of treatment. With PDTC, decreased transcriptional activity occurred prior to the detectable increase in IB␣. This difference may be attributed to the relative insensitivity of immunoblotting to detect early small increases in the protein. On the basis of measurements using the NFB-dependent luciferase assay, NFB transcriptional activity decreases within the time frame during which ⌬ M is disrupted and prior to pro-caspase-9 processing or morphological evidence of cell death. These results are consistent with the possibility that changes in NFB activity underlie the apoptotic response to PDTC and TPCK and suggest that constitutive NFB activity is necessary for SH-EP1 and SK-N-AS cell survival.
NFB Is Constitutively Active in S-type Cells-The reporter gene assays above, which showed basal activity of the NFB promoter construct in untreated cells, provided initial direct evidence that NFB is indeed activated in untreated S-type cells. To more firmly establish that this is indeed the case, we directly measured NFB sequence-specific DNA binding. Nuclear extracts prepared from untreated S-type cells (SH-EP1 and SK-N-AS) were incubated with a 32 P end-labeled double- stranded DNA oligonucleotide with the sequence 5Ј-AGTT-GAGGGGACTTTCCCAGGC-3Ј. As seen in Fig. 4A, extracts of both S-type lines shifted the labeled NFB oligonucleotide during electrophoresis, resulting in three distinct complexes. Additional experiments were performed to determine which shifted complexes correspond to NFB-specific binding. First, nuclear extracts were prepared from cells treated with TNF-␣, a potent inducer of NFB activity (27). Upon comparison with the EMSA results from untreated cells, the intensity of two bands were increased by TNF-␣ treatment, consistent with their identity as NFB⅐DNA complexes (Fig. 4A, asterisk). Second, confirmation was obtained by antibody supershift experi- ments in which antibodies specific for either p50 or p65 were added to the mixture of nuclear extract and labeled oligonucleotide. As seen in Fig. 4B, each of the two putative NFB⅐DNA complexes was supershifted by the p50-specific antibody, whereas the larger complex (upper band) was also supershifted by anti-p65. Hence, constitutively active NFB is detected in S-type cells and consists of both p50/p50 and p50/ p65 dimers.
N-type cells were expected to have little or no basal NFB DNA binding activity, because prior experiments showed the absence of constitutive NFB reporter gene activation in N-type cells (14). We tested this possibility and directly compared S-type and N-type cells with respect to constitutive NFB DNA binding activity. As expected, when the EMSA was performed with nuclear extracts prepared from untreated N-type cells (SH-SY5Y and IMR32), neither of the two complexes specific for NFB⅐DNA binding was detected (Fig. 4C). We confirmed that the cytotoxic action of PDTC in S-type cells is related to a reduction of NFB⅐DNA binding. Treating SH-EP1 cells with PDTC resulted in a dose-dependent reduction in DNA binding activity (Fig. 4D). Similar results were obtained when cells were treated instead with TPCK (data not shown). These results confirm the constitutive NFB activity in S-type cells and show that the inhibition of NFB is associated with the response of S-type cells to small molecule inhibitors of NFB signaling that induce apoptosis.
SR-IB␣ Induces Apoptosis of S-type Cells-Experiments were next performed to test whether the specific inhibition of NFB is sufficient to induce S-type cell death. The superrepressor mutant form of IB␣ (SR-IB␣) binds NFB but can not be phosphorylated on the basis of alanine substitution for serines 32 and 36 (28). SH-EP1 and SK-N-AS cells were cotransfected with pCMV4-SR-IB␣-FLAG along with the NFBresponsive luciferase reporter (pBVIx-Luc) and the expression plasmid pCDNA3-p35 (21). pCDNA3-p35 encodes a dominant negative inhibitor of caspase-9 and was included in these experiments to block the expected cytotoxic effect of inhibiting NFB so that changes in transcriptional activity could be detected. SR-IB␣ expression was first confirmed by immunoblotting (Fig. 5A). Additionally, cells transfected in this manner showed no evidence of increased cell death compared with cells transfected with SR-IB␣ vector-only plasmid (data not shown). In a dose-dependent manner, SR-IB␣ expression decreased NFB-dependent luciferase activity in both cell lines, confirming the activity of the super-repressor against NFB transcriptional activation in these cells (Fig. 5A).
To determine whether the inhibition of NFB caused cell death, SH-EP1 cells were transiently co-transfected to express SR-IB␣ and GFP (as a transfection marker) without the expression plasmid pCDNA3-p35. Control-transfected cells demonstrated normal morphology including intact nuclei (Fig. 5B,  upper panels). In contrast, 24 h following transfection, cells expressing SR-IB␣ demonstrated apoptotic morphology: rounding up, detachment, and nuclear condensation (Fig. 5B,  lower panels). This cytotoxic effect was further quantified by enumerating cell survival following transfection on the basis of expression of a co-transfected marker gene, ␤-galactosidase. SH-EP1 and SK-N-AS cells were transfected with efficiencies of 5.5 and 8.0%, respectively, using the ␤-galactosidase expression vector. ␤-Galactosidase-positive cells per high power field were counted 24 h after six independent co-transfections with pEF-1-Bos/␤-galactosidase and either pCMV4 control or pCMV4-SR-IB␣-FLAG. For SH-EP1 cells, ␤-galactosidase positivity decreased from 50 Ϯ 1.9 cells/high power field with control to 17 Ϯ 1.1 following pCMV4-SR-IB␣-FLAG transfection, revealing a 60% reduction in survival (p Ͻ 0.01). For SK-N-AS cells, 84 Ϯ 10 ␤-galactosidase-positive cells/high power field were counted for control compared with 45 Ϯ 5.2 with pCMV4-SR-IB␣-FLAG, a 40% reduction in survival (p Ͻ 0.01). Together, these results confirm that specific molecular inhibition of NFB changes NFB-dependent gene expression FIG. 4. NFB is constitutively active in S-type NB cells. A, nuclear extracts prepared from SH-EP1 and SK-N-AS cells cultured with or without TNF-␣ (100 ng/ml) for 20 min were used in EMSAs with a NFB consensus sequence probe. NFB-specific bands (indicated by asterisk) were identified on the basis of their increase in response to TNF-␣. B, EMSA performed with supershift analysis using anti-p65 or anti-p50 antibodies. The NFB probe and nuclear extracts from untreated SH-EP1 and SK-N-AS cells confirm NFB proteins are involved in binding activity. C, EMSA using nuclear extracts from SH-EP1 and N-type cell lines (SH-SY5Y, IMR32) indicates that NFB-specific binding is not detected in nuclear extracts from untreated N-type cells. D, EMSA using nuclear extracts from SH-EP1 cells treated with PDTC (50 -150 M, 4 h), TNF-␣ (100 ng/ml, 20 min), or control medium indicates PDTC dose-dependently reduces NFB-specific binding activity. and concomitantly reduces the survival of S-type cells in culture.

DISCUSSION
In the absence of appropriate stimuli, NFB is sequestered in the cytoplasm by physical association with IB, which prevents exposure of nuclear localization signals. As an immediate response to stimuli including pro-inflammatory cytokines, lipopolysaccharides, and phorbol esters, IB is phosphorylated, ubiquitinated, and degraded by proteasomes (29). The cytoplasmic to nuclear translocation of NFB and transcriptional activation of NFB-responsive genes follow (30). In turn, NFB induces the synthesis of IB␣, which results in negative feedback, limiting the magnitude and duration of the response (31).
Despite tight coupling of NFB to specific signals as outlined above, constitutive NFB activation can occur. Constitutive NFB activation appears to have an important role in tumorigenesis. For example, persistent nuclear NFB localization and NFB-dependent transcription is detected in breast (32), ovarian (33,34), colon (34), thyroid (35), prostate (36), and melanoma (37) tumors. In breast and prostate tumor cells, constitutive NFB activity is associated with reduced levels of IB␣ that appears related to increased degradation of IB proteins in these cells (38,39). Constitutive NFB activity is also associated with the tumorigenic properties of viruses. The viral homologue of c-rel (v-rel) is transduced in the reticuloen-dotheliosis virus and is responsible for the acute-transforming properties of this virus in mammalian cells (40). Constitutive activation of cellular gene-encoded NFB mediates transformation by human T-cell lymphotrophic virus, type I. The human T-cell lymphotrophic virus, type I-encoded Tax protein activates NFB by stably activating IKK, providing an ongoing signal for IB␣ degradation (41).
IB␣ degradation is primarily regulated through the phosphorylation of N-terminal serine residues 32 and 36 by the IKK complex. Phosphorylation at these residues leads to dissociation from NFB and targets the protein for ubiquitination and proteasomal degradation. Alternatively, the phosphorylation of C-terminal serine residues by casein kinase 2, phosphorylation at tyrosine 42 under conditions of anoxia, and treatment with UV radiation can result in NFB activation independent of IKK (42). Our results are most consistent with the two N-terminal serine residues determining constitutive NFB activity in S-type cells. This conclusion is based on the inhibitory activity of the mutant SR IB␣ that derives its dominantnegative action solely on the basis of alanine substitutions at serine 32 and 36. However, to completely exclude a contribution of the C-terminal region or tyrosine 42, additional mutants involving these residues would need to be studied.
Because the IKK complex is the only kinase recognized to directly phosphorylate serine 32 and 36 on IB␣ (3), constitutive IKK activity is suspected to underlie constitutive NFB activation in this model. The regulation of IKK is complex, and input converges from multiple signaling cascades including those coupled to TrkA (the high affinity receptor for nerve growth factor) and tumor necrosis receptor-1. Additionally, other signals that activate Ras, phosphatidylinositol 3-kinase, and MEKK1 also increase IKK activity. Because S-type cells neither express nerve growth factor receptors nor demonstrate nerve growth factor responsiveness (43,44), the activity of TrkA or the low affinity nerve growth factor receptor is unlikely to sustain NFB activation. The autocrine activity of TNF-␣ seems a more probable possibility. S-type cells are responsive to this cytokine (see Fig. 2), and cells derived from the neural crest (including sympathetic neurons) are known to synthesize both TNF-␣ and TNFR1 (45). NFB supports neoplastic growth by preventing cell death. In this study, we found that inhibition of constitutive NFB activity engages an apoptotic response that specifically involves the collapse of the mitochondrial transmembrane gradient and caspase-9 activation, both responses associated with the mitochondrial permeability transition. The anti-apoptotic function of NFB in these cells could be mediated by transcriptional up-regulation of any one of the array of NFB-responsive genes including TRAF1, TRAF2, c-IAP-1, c-IAP-2, A20, and Bcl-x L (46 -48). With the involvement of the mitochondria and caspase-dependent processes in the death response, it is possible that c-IAP-1, c-IAP-2 (inhibitors of caspase activation), and Bcl-x L (stabilizer of the mitochondrial permeability transition pore) are potentially important. Indeed, S-type cells have previously been shown to constitutively express Bcl-x L (49).
In preliminary experiments using an activation-specific antibody against p65, we have uncovered evidence of NFB activation in a NB tumor specimen (data not shown). From hereon, it will be important to firmly establish whether NFB is indeed activated specifically within the stromal components of NB tumor specimens and whether differences in activation are associated with stage and histopathologic grade. Efforts to directly measure transcription factor activation are best combined with experiments to quantify the expression of NFB target genes to more strongly support conclusions regarding NFB activation. Along these lines, a recent study has ele-

FIG. 5. SR-IB␣ inhibits NFB-dependent transcription and decreases survival of S-type cells.
A, SH-EP1 and SK-N-AS cells were co-transfected with pCMV4-SR-IB␣-FLAG (0 -500 ng/well) and the reporter pBIVx-Luc (50 ng/well). SR-IB␣-FLAG expression was confirmed by immunoblotting. Luciferase activity was normalized to protein concentration and expressed as arbitrary relative light units. Mean Ϯ S.D. from three experiments is shown and indicates that SR-IB␣ expectedly decreases NFB-dependent luciferase activity. B, SH-EP1 cells were co-transfected with the pEGFP-C1 (as a marker of transfection, 20 ng/well) and either pCMV4-SR-IB␣-FLAG or the control plasmid (100 ng/well). 24 h after transfection, cells were stained with 4,6-diamidino-2-pheylindole. Fluorescent microscopy (ϫ400) enumerated surviving transfected clones (cells with green fluorescence). Cells with SR-IB␣ had reduced survival and increased DNA condensation (fragmented nuclear staining), consistent with apoptosis. gantly used gene expression profiling to show that diffuse large B cell lymphoma consists of two molecularly distinct diseases that are distinguished on the basis of their constitutive expression of NFB-induced genes (50). Similar experimental strategies are planned to address the question of in vivo NFB activation in NB.
In conclusion, NFB is constitutively active in S-type NB cells. Moreover, because it is necessary for S-type cell survival in culture, it has the potential to contribute to cellular transformation, tumor progression and resistance to chemotherapy. Given its opposing, pro-apoptotic activity in N-type NB cells, therapeutic strategies designed to inhibit NFB must be approached cautiously.