Brain-Derived Neurotrophic Factor Activation of Trkb Protects Neuroblastoma Cells from Chemotherapy-Induced Apoptosis Via Phosphatidylinositol 3؅-Kinase Pathway

Home , Autophosphorylation

[CANCER RESEARCH 62, 6756–6763, November 15, 2002] Brain-derived Neurotrophic Factor Activation of TrkB Protects Neuroblastoma Cells from Chemotherapy-induced Apoptosis via Phosphatidylinositol 3؅-Kinase Pathway

Jerry Jaboin,1 Chong Jai Kim, David R. Kaplan, and Carol J. Thiele2 Pediatric Oncology Branch, National Cancer Institute, Bethesda, Maryland 20892 [J. J., C. J. T.]; Department of Pathology, Seoul National University College of Medicine, Seoul, Korea 100-799 [C. J. K.]; and Brain Tumor Research Centre, Montreal Neurological Institute, Montreal, Quebec, Canada, H3A 2B4 [D. R. K.]

ABSTRACT NTs are mediated by tyrosine kinase receptors (TrkA, TrkB, and TrkC) and the p75NTR, a member of the death receptor family, of Neuroblastoma (NB) tumors expressing high levels of brain-derived which the signaling may synergize or antagonize tropomysin-related neurotrophic factor (BDNF) and TrkB are associated with poor 5-year kinase signaling depending on cell context (10). NT activation of Trks survival outcomes. Our previous studies indicated that BDNF blocked the cytotoxic effects of vinblastine on NB cells. Here we evaluated the ability plays a role in the survival, axonal growth, and differentiation of of BDNF to decrease the chemosensitivity of NB cells to a number of neurons (11, 12). BDNF, in particular, has also demonstrated to have common chemotherapeutic agents. Two SH-SY5Y NB cell lines (TB3 and protective effects in several neuronal model systems (13–16). BDNF TB8) expressing TrkB under the control of a tetracycline (Tet)-repressible promotes survival in a variety of neuronal cell types including striatal promoter element were generated, and used to assess apoptosis resulting (17) and sensory neurons (18), and protects motor neurons from from treatment with cisplatin, doxorubicin, etoposide, and vinblastine. axotomy-induced cell death (19), hippocampal neurons from nitrous BDNF treatment of high TrkB-expressing TB8 (Tet؊) and TB3 (Tet؊) oxide induced apoptosis (20), and cerebellar granule neurons from cells blocked drug-induced cell death in a dose-dependent manner. Only metabolic and excitotoxic insults (14, 16, 21, 22). We hypothesized high-dose BDNF (100 ng/ml) could block the effects of chemotherapy in that tumor cells of neuroectodermal origin like NB might use a low TrkB-expressing cells. The ability of BDNF to rescue the cells from NT/Trk signal transduction pathway to protect themselves from chemotherapeutic agent-induced cell death was inhibited by treatment with the Trk tyrosine kinase inhibitor K252a or the phosphatidylinositol chemotherapy-induced cell death. In support of this hypothesis, prior 3؅-kinase (PI3K) inhibitor LY294002, but not by the mitogen-activated injection of NGF into the developing rat sympathetic nervous system protein kinase kinase inhibitor PD98059 or the peritoneal lymphocyte ␥ protected neurons from ablation by Vbl (13). We have also shown that inhibitor U73122, indicating that both TrkB and PI3K activities are BDNF protects NB cells expressing a transfected TrkB or endogenous required for the survival-promoting effects of BDNF. BDNF also pro- TrkB induced by retinoid treatment from Vbl-induced cell death (7). tected TrkB-expressing NGP and KCNR NB cells from chemotherapeutic These results were extended to include other chemotherapeutic agents agent-induced cell death, and LY294002 inhibited this protection. These used in the treatment of NB (8). However it was not clear using these results suggest that TrkB and BDNF can contribute to the chemoresis- model systems whether the effects of BDNF were solely dependent on tance of poor prognosis tumors, and that suppression of PI3K activity TrkB or whether p75NTR or other RA-induced genes contributed to might improve the ability of these agents to induce the death of NB the chemoprotection afforded by BDNF. tumors. Although the majority of poor prognosis patients with NB respond to induction chemotherapy, Ͼ75% of patients relapse (3). Because our INTRODUCTION studies indicate that the BDNF signaling system contributes to che- NB3 is a pediatric solid tumor derived from neural crest precursor moresistance (7), we should be able to improve NB cell chemosen- cells (1). It can be classified into good and poor prognostic categories sitivity if we can specifically target the pathway by which BDNF based on age, DNA ploidy, N-myc amplification, tumor location, and alters the response of NB cells to chemotherapy. To investigate the TrkA expression (1). Favorable-prognosis NB tumors express rela- signaling system by which BDNF protects NB cells from chemother- tively high levels of TrkA, whereas cell lines and tumors from patients apy, we developed a series of Tet-regulated TrkB-expressing SH- with poor prognosis are associated with expression of BDNF and its SY5Y cell NB cell lines (23). We found that BDNF stimulation of receptor, TrkB (1–3). Several studies indicate that BDNF increases TrkB rescued NB cells from apoptosis induced by Vbl, Doxo, Etop, NB cell survival (2, 4, 5), neurite extension (2, 6), and cell invasion and Cis. The BDNF-induced rescue of the cells was blocked by both (4), and protects cells from chemotherapy (7, 8). This indicates that pharmacologically inhibiting TrkB tyrosine kinase activity or by aside from being a marker of poor prognosis, BDNF and TrkB play a inhibiting PI3K activity. These results suggest that drugs that inhibit role in the biology of NB tumors. PI3K or Trk activity may enhance the effectiveness of chemothera- NTs are a family of growth factors that are important in the peutic agents in the treatment of NB tumors. development of the sympathetic nervous system (9). The activities of MATERIALS AND METHODS Received 6/6/02; accepted 9/20/02. The costs of publication of this article were defrayed in part by the payment of page Construction of Vectors and Establishment of Stable Transfectants. A charges. This article must therefore be hereby marked advertisement in accordance with 3.1-kb fragment of rat TrkB spanning a full coding region was subcloned into 18 U.S.C. Section 1734 solely to indicate this fact. the pBSTR1 vector (23) and transfected into the human NB cell line SH-SY5Y 1 J. J. is a MD/Ph.D. student from Howard University College of Medicine doing his using LipofectAMINE (Life Technologies, Inc.). Stable transfectants (TB3, thesis work in the Cell and Molecular Biology Section/Pediatric Oncology Branch/Center ␮ for Cancer Research/National Cancer Institute. TB8, and AS-TB8) were obtained after selection with puromycin (0.5 g/ml) 2 To whom requests for reprints should be addressed, at Cell and Molecular Biology and characterized in detail by Kim et al. (23). Section, Pediatric Oncology Branch, Center for Cancer Research, National Cancer Insti- Cell Culture. Cells were cultured in RPMI 1640 containing 10% fetal tute, 10 Center Dr. MSC-1928, Bethesda, MD 20892. Phone: (301) 496-1543; Fax: bovine serum, 2 mM glutamine, and antibiotics as described previously (23). (301) 402-0575; E-mail: [email protected]. 3 The abbreviations used are: NB, neuroblastoma; BDNF, brain-derived neurotrophic To maintain selection pressure the TB3 and TB8 cells were cultured in factor; NT, neurotrophin; RA, retinoic acid; PI3K, phosphatidylinositol 3Ј-kinase; TB3, puromycin (0.5 ␮g/ml), and in the presence of Tet (1 ␮g/ml) for repression of SH-SY5Y-TB3; TB8, SH-SY5Y-TB8; AS-TB8, SK-N-AS-TB8; Cis, cisplatin; Doxo, the transfected gene. doxorubicin; Etop, etoposide; Vbl, vinblastine; TdT, terminal deoxynucleotidyltrans- Reagents. Recombinant human BDNF (100 ng/␮l; PeproTech, Inc.) and ferase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Tet, tetracy- ␮ cline; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein 2.5S mouse NGF (100 ng/ l; Upstate Biotechnology) stocks were prepared as kinase. directed and stored at Ϫ20°C. The chemotherapeutic agents, Cis, Doxo, Etop, 6756

Tris-glycine bis-acrylamide gels, probed with mouse monoclonal antiphosphotyrosine antibody, and subsequently reprobed with anti-Cas antibody to determine total Cas protein levels. Akt and ERK phosphorylation changes were analyzed by Western blotting analysis of 40 ␮g of protein loaded onto 12% Tris-glycine bis-acrylamide gels, and probed with rabbit monoclonal antibodies to phospho-Akt (Cell Signaling) and phospho-p44/p42 (Cell Signaling). Total Akt and ERK protein levels were determined by Western blotting analysis of 40 ␮g of protein with rabbit monoclonal antibodies to Akt and p44/p42. Signals were detected using enhanced chemiluminescence reagents (Pierce). Apoptosis Assay. To assess a parameter of apoptosis, TB3 or TB8 cells were cultured as indicated in “Results,” washed, fixed in 4% paraformalde- hyde, and stored in ethanol. Cells were end-labeled with TdT-FITC using the ApoDirect kit (PharMingen) according to the manufacturer’s recommenda- Fig. 1. TrkB expression. Induction of TrkB mRNA expression in TB8 and TB3 cells. Cells were cultured for 3 days in the presence of various concentrations of Tet in RPMI tions and analyzed using a FACScan Flow Cytometer (Becton Dickinson). 1640 with 10% FBS. Total RNA was extracted, and 15 ␮g of total RNA was electro- Experiments were performed in triplicate, plates were combined, and three sets phoresed in a 1.2% formaldehyde gel and subsequently transferred to a nitrocellulose of 10,000 cells were collected for each condition. The data were acquired and membrane. analyzed by the CELLQuest software (Becton Dickinson). Cell Survival Analysis. Ten-thousand cells per well were plated into 96-well plates. After 24 h, the cells were treated with control medium or BDNF and Vbl were obtained from Sigma, and 1 mM stocks of each drug were (100 ng/ml). Twenty-four h later, indicated concentrations of chemotherapeu- prepared according to the manufacturer’s specification, and stored at Ϫ20°C. tic drugs were added for 2 h. After 2 h the culture supernatant was removed and The pharmacological inhibitors, LY294002 (1 mM), PD98059 (1 mM), U73122 fresh medium added for the remainder of the culture period. Similar results (1 mM), and K252a (1 mM) were obtained from Sigma, and reconstituted were obtained when the drug was cultured for the duration of 48 h without a according to manufacturer’s specification. medium change. Cell number was assessed 48 h later using the MTT assay. Northern Blot Analysis. RNA isolation and hybridization were performed Each value represents six replicates, and each experiment was repeated two to as described previously (24). Thirty ␮g total RNA for TrkB was electrophore- three times (26). sed in 1% agarose-6% formaldehyde gels. Gels were stained with 2 mg/ml ethidium bromide. Hybridization was performed with 32P-labeled insert DNA isolated from a plasmid containing TrkB (23). RESULTS Immunoblotting/Immunoprecipitation. For protein analysis, cells were washed in PBS, centrifuged, and cell pellets kept in Ϫ70°C until analysis, and Tet Suppresses BDNF-mediated Signal Transduction in SY5Y prepared as described previously (25). The cells were then lysed in NP40 lysis NB Cells Expressing TrkB under the Regulation of a Tet-respon- buffer containing 20 mM Tris (pH 7.4), 150 mM NaCl, 1% NP40, 10% glycerol, sive Promoter Element. The previous studies evaluating the effects 1mM phenylmethylsulfonyl fluoride, 10 ␮g/ml aprotinin, 1 ␮g/ml leupeptin, of TrkB on chemosensitivity were limited by cell model systems that and 500 ␮M sodium orthovanadate in ice-cold PBS. Cells were lysed on ice for used stable cell lines that constitutively overexpressed TrkB (7) or 30 min, and protein concentrations were determined using the Bradford assay used retinoids to induce endogenous TrkB expression (7, 8). To assess kit (Bio-Rad). For immunoprecipitation of phosphorylated TrkA and TrkB TrkB function in the absence of effects of receptor overexpression or receptors, 500 ␮g of protein from each lysate was immunoprecipitated with of retinoids, we generated SY5Y NB cell lines in which the levels of polyclonal rabbit anti-pan-Trk antibody (C-14; Santa Cruz) and protein A TrkB are controlled by a Tet-regulatable promotor element (23). In agarose (Life Technologies, Inc.). Immunoprecipitates were electrophoresed in two cell lines, TrkB-SY5Y (TB3) and TrkB-SY5Y (TB8), the levels 8% Tris-glycine bis-acrylamide gels (Invitrogen), probed with mouse monoclonal antiphosphotyrosine antibody (PY-99; Santa Cruz), and subsequently of TrkB mRNA were markedly repressed when the cells were cultured ␮ reprobed with anti-pan-Trk antibody to assess total Trk levels. For immuno- in Tet (1 g/ml; Fig. 1). TrkB mRNA expression was completely precipitation of phosphorylated Cas protein, 500 ␮g of protein of each lysate suppressed by Tet (1 ␮g/ml), and TrkB levels increased 4.6-fold in the was immunoprecipitated with rabbit polyclonal anti-Cas antibody (C-20; Santa absence of Tet (Fig. 1). Cruz) and protein A agarose. Immunoprecipitates were electrophoresed in 8% The time course of BDNF-induced TrkB autophosphorylation, and

Fig. 2. Activation of the TrkB signal transduction pathways. TB8 cells were cultured for 3 days in the presence (1 ␮g/ml) or absence of Tet in complete medium. The cells were then stimulated with BDNF (100 ng/ml) and harvested at various times thereafter. A, TrkB path. Protein lysates (500 ␮g) were immunoprecipitated with rabbit polyclonal anti-pan trk antibody (C-14) and subjected to Western analysis (as described in “Materials and Methods”) for evaluation of TrkB autophosphorylation and total TrkB levels.B,PI3K path. Forty ␮g of protein lysate was analyzed for P-Ser-Akt and total Akt by Western blot analysis.C,MAPK path. Forty ␮g of protein lysate was analyzed for P-Thr/ Tyr-ERK and total ERK by Western blot analysis. D, PLC-␥ path. Protein lysates (500 ␮g) were immunoprecipitated with rabbit polyclonal anti-Cas antibody (C-20) and subjected to Western blot analysis for evaluation of Cas autophosphorylation and total Cas levels.

6757

Fig. 3. BDNF rescues TrkB-expressing NB cells from chemotherapy-induced cytotoxicity. Morphology of TB3 cells expressing high levels of TrkB in the absence or presence of BDNF and chemotherapeutic agents. TB3 cells expressing high levels of TrkB were cultured in the absence (top panels) or presence of BDNF (100 ng/ml; bottom panels) followed by a 48-h incubation with the indicated chemotherapeutic agents: Cis (3 ␮g/ml), Etop (3 ␮g/ml), and Vbl (6 ng/ml). Photos were taken at ϫ100 magnification using a Nikon Diaphot inverted microscope with Hoffman modulation.

the phosphorylation of the signaling proteins AKT, ERK1/2, and Cas were investigated in TB8 cells in the presence or absence of Tet. TB8 cells expressed 2-fold higher TrkB levels in the absence of Tet as compared with the presence of Tet (Fig. 2A). Three major signaling pathways are activated after BDNF activation of the TrkB tyrosine kinase: the PI3k pathway, the Ras/MAPK pathway, and the PLC-␥ pathway (27). We used Akt, ERK1/2, and Cas phosphorylation as indicators of the activation of the PI3k, Ras/MAPK, or PLC-␥ pathway, respectively. In the presence of Tet (Tetϩ), Trk, Akt, and ERK1/2 phosphorylation were maximal at 5 min of BDNF treatment of cells, and declined to basal levels by 30 min for Trk and Akt, and 60 min for ERK1/2 (Fig. 2, A–C). In the absence of Tet (TetϪ), Akt and ERK1/2 phosphorylation were also maximal at 5 min but persisted for at least 180 min (Fig. 2, A–C). Cas phosphorylation was sustained in both the presence and absence of Tet addition, although phosphorylation was quantitatively more in the absence of Tet (Fig. 2D). BDNF Rescues TrkB-expressing Cells from Chemotherapy- induced Apoptosis. Scala et al. (7) showed previously that BDNF addition to NB cells induced to express TrkB by RA treatment could rescue the cells from Vbl-induced cytotoxicity. We assessed apoptosis in the Tet-responsive cells lines treated either with either Vbl or with three other chemotherapeutic agents commonly used to treat patients with NB, Cis, Doxo, or Etop. All four of the agents appeared to induce the death of TB3 cells in the absence of BDNF treatment (Fig. 3). Both Cis and Etop treatment resulted in detachment and loss of cell adherence (Fig. 3, top panel). Vbl treatment caused a detachment and loss of adhesion of some cells, whereas others seemed to have a more differentiated morphology (Fig. 3, top panel). The IC50 of these agents that was required to induce the apoptosis of TB8 cells was 3 ␮g/ml for Cis, 3 ␮g/ml for Etop, 0.05 ␮g/ml for Doxo, and 3 ng/ml for Vbl (data not shown). The IC50 for the drugs were similar in TB3 cells, and Fig. 4. BDNF rescues TrkB-expressing cells from chemotherapy-induced apoptosis. there was no statistical difference if cells were cultured in the absence TB3 (A) and TB8 (B) cells expressing high levels of TrkB were cultured for 24 h in the presence or absence of BDNF followed by a 2-h treatment with the indicated chemother- or presence of Tet (data not shown). apeutic agents and 48 h in control medium. Apoptosis was assessed by Tdt-mediated nick To determine whether BDNF would rescue the TrkB-expressing end labeling as described in “Materials and Methods.” Bars, ϮSD. 6758

Fig. 5. BDNF activation of TrkB is responsible for rescue. A–D, MTT analysis of TB8 cells in indicated conditions. TB8 cells expressing low levels of TrkB (A1–D1) or high levels of TrkB (A2–D2) were cultured with various concentrations of BDNF (100 ng/ml) or NGF (100 ng/ml) for 24 h followed by Cis (A1 and A2), Doxo (B1 and B2), ETOP (C1 and C2), or Vbl (D1 and D2) for 2 h, then washed and incu- value is significantly ,ء .bated with medium for 48 h different from chemotherapy-treated cells. ⌬, value is not significantly different from control cells. Per- centages were calculated with respect to the untreated cell conditions; bars, ϮSD.

cells (TetϪ) from chemotherapy-induced cell death, TB3 cells were labeling in cells not treated with the agents (Fig. 4A). TdT labeling cultured with BDNF (100 ng/ml) for 24 h, and Cis (3 ␮g/ml), Etop (3 was markedly decreased (16, 19, and 23%) if TB3 cells were pre- ␮g/ml), Doxo (30 ng/ml), or Vbl (6 ng/ml) were added for 48 h. The treated for 24 h with BDNF (100 ng/ml). Similarly, in TB8 cells, Cis, agents were removed, and apoptosis was assessed after an additional Etop, and Vbl induced 56, 55, and 35% apoptosis, and BDNF pre- 48 h. BDNF did not affect the cell growth or morphology of the TB3 treatment dramatically reduced apoptosis (Fig. 4B). In the presence of cells as compared with control medium (Fig. 3, bottom left panel). the chemotherapeutic agents, TdT labeling was detected in cells However, pretreatment with BDNF prevented the chemotherapy- throughout the cell cycle, indicating that the agents induced apoptosis induced reduction in cell number. These data indicate that BDNF can regardless of the phase of the cell cycle (data not shown). In addition, protect NB cells from cell death caused by a variety of chemothera- under these conditions the agents did not induce a cell cycle arrest peutic agents. (data not shown). These data indicate that the BDNF can protect To confirm that the reduction in cell number after treatment with TrkB-expressing NB cells from chemotherapy-induced apoptosis. the chemotherapeutic agents was because of an increase in apoptosis Dose Response of BDNF Required for Protection from Chemo- and not a cell cycle arrest, we performed TdT labeling to assess therapy-induced Cell Death. We first determined whether the apoptosis and flow cytometry to examine cell growth. Treatment of BDNF-induced protection from cell death was dependent on the level Ϫ TB3 cells (Tet ) with Cis, Etop, and Vbl at the IC50 dose for 2 h of expression of the TrkB receptor. TB8 cells were cultured in the followed by a 48-h incubation in control medium resulted in 58, 59, presence (Tetϩ) or absence (TetϪ) of Tet for 24 h, treated with and 61% increases in TdT labeling as compared with 10% TdT various concentrations of BDNF (10, 33, and 100 ng/ml) for 24 h, 6759

Fig. 6. Pharmacological inhibition of TrkB signal paths. A, inhibition of Trk signal transduction path. Western analysis of TrkB autophosphorylation in lysates from TB8 cells expressing high levels of TrkB pretreated for 30 min with various concentrations of K252a, and then stimulated for 5 min with BDNF (100 ng/ml). Immunoprecipitation (IP) conducted as described in “Materials and Methods.” B, inhibition of PI3k signal transduction path. TB8 cells expressing high levels of TrkB were pretreated for 30 min with indicted concentrations of LY294002 (LY) and stimulated for 5 min with BDNF (100 ng/ml). Lysates were immunoblotted for P-Ser-Akt and total Akt. C, inhibition of MAPK signal transduction path. TB8 cells expressing high levels of TrkB were pretreated for 30 min with indicated concentrations of PD98059 (PD) and then stimulated with BDNF (100 ng/ml) for 5 min. Lysates were assessed for P-Thr/Tyr-ERK and total ERK levels by Western analysis. D, inhibition of PLC-␥ signal transduction path. TB8 cells were pretreated for 30 min with indicated concentrations of U73122 (U) and then stimulated with BDNF (100 ng/ml) for 5 min. Lysates were analyzed for P-Tyr-Cas and total Cas levels by immunoprecipitation followed by immunoblotting as described in “Materials and Methods.” followed by a 2-h treatment with the chemotherapeutic agents. Cell BDNF rescue from chemotherapeutic agent-induced cell death. number was assessed after an additional 48 h in control medium. The inhibitors were added to the TB8 (TetϪ) cells 30 min before the BDNF at 10 or 33 ng/ml could not rescue the Tetϩ cells from cell addition of BDNF. After 24 h the cells were treated with Etop (3 death resulting from treatment with the agents (Fig. 5, A1–D1). ␮g/ml) for 2 h, the Etop was removed, and cell number was assessed However, there was a complete and statistically significant rescue after 48 h. Treatment of BDNF-stimulated cells with the inhibitors did from cell death when low TrkB-expressing cells were pretreated with not significantly affect cell survival under these conditions (Fig. 7). BDNF at 100 ng/ml (Fig. 5, A1–D1). In each case the values for the The Trk tyrosine kinase inhibitor K252a was able to completely block BDNF rescue of cells were significantly different from the values in the ability of BDNF to rescue cells from Etop-induced cell death (Fig. and for Cis, Doxo, and Vbl, the 7A). Similarly, the PI3K inhibitor LY294002 completely abrogated ,(ء ,the drug-treated cells (Fig. 5 values were not statistically different from the untreated or control the ability of BDNF to protect cells from Etop-induced cell death (Fig. cells (Fig. 5, ⌬). Thus, in TB8 cells expressing low levels of TrkB, 7B). In contrast, neither the mitogen-activated protein kinase kinase only BDNF (100 ng/ml) could protect from chemotherapy-induced inhibitor PD98059 nor the PLC-␥ inhibitor U73122 altered the ability cell death. NGF (100 ng/ml), which binds both the p75NTR and low of BDNF to protect cells from Etop-induced cell death (Fig. 7, C and levels of TrkA expressed in SY5Y cells, could not protect the cells D). These data indicate that the activities of TrkB and of PI3K are from chemotherapy-induced cell death. required for BDNF to protect the NB cells from Etop-induced cell In TB8 (TetϪ) cells, which express higher levels of TrkB as death. LY294002 also blocked the ability of BDNF to rescue cells compared with TB8 (Tetϩ) cells, a lower dose of BDNF (33 ng/ml) from Doxo-, Vbl-, and Cis-induced cell death (data not shown). protected the cells from chemotherapeutic agent induced cell death Similar results were found in the AS-TB8 cell lines, another TrkB- (Fig. 5, A2–D2). In the high TrkB (TetϪ)-expressing TB8 cells there transfected NB cell line derived from SK-N-AS (Fig. 7E, inset). was a dose-dependent increase in the ability of BDNF to protect cells Increasing concentrations of BDNF protect AS-TB8 cells from Etop- from Cis-induced cell death. For Cis and Vbl, as little as 10 ng/ml induced cell death (Fig. 7E), and the protection can be blocked by BDNF could significantly protect the cells from chemotherapeutic using LY294002 (Fig. 7F). agent-induced cell death (Fig. 5, A2 and D2). Again, NGF could not To determine whether inhibition of PI3K activity would sensitize protect the cells from death induced by the agents. These data indicate other NB cell lines to chemotherapeutic agent-induced cell death, we that the protection from cell death depends on both the levels of TrkB treated KCNR and NGP NB cell lines, which constitutively express and the concentration of BDNF. BDNF (25), with physiological concentrations of all-trans-RA (10 BDNF Protects TrkB-expressing NB Cells from Chemothera- nM) to induce TrkB expression. This concentration of RA does not peutic Agent-induced Cell Death via the PI3k Pathway. To con- significantly alter cell growth, although cells do extend neurites (4). firm that TrkB and not the p75NTR mediate the BDNF-induced LY294002 (10 ␮M) was able to cause a 29% and 36% increase in cell ϭ protection from cell death, and to identify the signal transduction death in RA and Etop-treated cells (KCNR IC50 4.1 ng/ml; NGP ϭ pathway(s) required for protection, we assessed the ability of BDNF IC50 2.3 ng/ml) when compared with cells not treated with ϭ ϭ to protect TB8 cells from cell death in the presence of pharmacolog- LY294002 (KCNR IC50 5.8 ng/ml, Fig. 8A; NGP IC50 3.6 ng/ml, ical inhibitors of Trk, PI3k, ERK1/2, and PLC-␥. First we determined Fig. 8, A and B). LY294002 also shifted the dose response of Etop- the concentrations of inhibitors required to suppress their respective induced cell death of BDNF and RA-treated NB cells. The IC50 was ϭ pathways. K252a, a relatively selective Trk tyrosine kinase inhibitor, increased nearly 2-fold in KCNR (IC50 8.14 ng/ml; Fig. 8C) and ϭ prevented BDNF activation of TrkB autophosphorylation in TB8 2.5-fold in NGP (IC50 8.5 ng/ml; Fig. 8D). Yet in each case, (TetϪ) cells at 500 nM (Fig. 6A). LY294002, a selective inhibitor of treatment with LY294002 blocked the effects of BDNF (KCNR ␮ ϭ ϭ PI3k, blocked the phosphorylation of the Akt at 10 M (Fig. 6B). IC50 3.9 ng/ml, Fig. 8C; NGP IC50 3.1 ng/ml, Fig. 8D). PD98059, an inhibitor of mitogen-activated protein kinase kinase, inhibited the phosphorylation of ERK1/2 at 50 ␮M (Fig. 6C). U73122, DISCUSSION a selective inhibitor of PLC-␥, completely blocked the phosphorylation of Cas activation at 10–20 ␮M (Fig. 6D). Each of the inhibitors Because NTs protect normal neurons from a variety of chemical was added to assess the contribution of the above proteins in the and physical insults (28), we hypothesized that the expression of NTs 6760

Fig. 7. Pharmacologic blockade of the PI3k pathway abrogates the cytoprotective effect of BDNF/TrkB activation. A–D, TB8 cells expressing high levels of TrkB were cultured as described in “Materials and Methods,” except that 30 min before addition of BDNF (100 ng/ml), cells were treated with indicated pharmacologic inhibitors. A, TB8 cells were pretreated with K252a (500 nM), a relatively selective Trk tyrosine kinase inhibitor. B, TB8 cells were pretreated with LY294002 (LY;10␮M), an inhibitor of the PI3K signaling pathway. C, TB8 cells were pretreated with PD98059 (PD;50␮M), an inhibitor of the MAPK signaling pathway. D, cells were pretreated with U73122 (U;20␮M), an inhibitor of the PLC-␥ pathway. E, AS-TB8 cells were cultured with various concentrations of BDNF or NGF (100 ng/ml) for 24 h followed by Etop for 2 h, then washed and incubated with medium for 48 h. Inset, AS-TB8 cells were treated with or without BDNF (100 ng/ml) for 15 min in culture and harvested. Protein lysates (500 ␮g) were immunoprecipitated with rabbit polyclonal anti-pan trk antibody (C-14) and subjected to Western analysis (as described “Materials and Methods”) for evaluation of TrkB autophosphorylation and TrkB levels. F, AS-TB8 cells were pretreated with LY294002 (LY;20␮M), an inhibitor of the PLC-␥ pathway. Data were generated using MTT analysis with percentages calculated with respect to the untreated cell conditions; bars, ϮSD. by tumors of neuroectodermal origin may antagonize the effects of therapy-induced apoptosis is TrkB-dependent is based on three find- chemotherapy and contribute to chemoresistance (7). A number of ings: (a) the inability of NGF to protect cells (Fig. 5); (b) the findings studies have shown that NTs such as BDNF promote the survival (2, that K252a could block the ability of BDNF to protect cells from 4, 5) and chemoprotection of NB cells (7, 8), although the mecha- chemotherapy-induced cell death (Fig. 7); and (c) that greater protec- nisms remain unclear. In this report, we demonstrate that BDNF via tion was afforded at high TrkB levels with low concentrations of activation of PI3K can block the apoptosis of NB cells treated with BDNF (Fig. 5). In our model, NGF, which binds TrkA and p75 but not Etop, Doxo, and Cis, chemotherapeutic agents commonly used in the TrkB, is unable to protect cells from chemotherapy-induced apoptosis. treatment of NB. This is in contrast to the finding that the ability of NGF to protect Our finding that the mechanism by which BDNF blocks chemo- SY5Y cells from neocarzinostatin-induced apoptosis is mediated by a 6761

Fig. 8. PI3K activation is responsible for mediating TrkB/BDNF protection in other NB cell lines. KCNR (A) and NGP (B) cells were treated with physiological levels of RA (10 nM) for 48 h, followed by treatment with LY294002 (20 ␮M in KCNR and 10 ␮M in NGP) and indicated concentrations of Etop for 48 h. KCNR (C) and NGP (D) cells were treated as in A, except for the addition of BDNF (100 ng/ml) before treatment with LY294002 and Etop. Data were generated using MTT analysis with percentages calculated with respect to the untreated cell conditions.

p75-dependent mechanism (29). p75 has both pro- and antiapoptotic cated in the ability of BDNF to rescue cerebellar granule neurons from signaling capacities, which are dependent on Trk activation status oxidative stress (35) and retinal ganglion cells from axotomy-induced (10). Thus, it may be that even the low levels of transfected TrkB in cell death (36). In fact, the survival of NB cells in serum-free medium our model system would mask the potential effect of p75 to block requires activation of the PI3K path (37). Despite the differing mech- apoptosis. We cannot rule out that the different cytotoxic agent, anisms of action of our chemotherapeutic agents, we have shown that neocarzinostatin, may account for the differences that we detect. BDNF is able to mediate rescue. This suggests that the ability of However, this seems unlikely, because we have shown that BDNF via BDNF activation of TrkB to rescue NB cells from chemotherapy may TrkB is able to protect NB cells from a variety of cytotoxic drugs of be the result of modulation of a key survival signaling molecule(s), which the mechanisms of action include changes in microtubule which is able to block the apoptosis-inducing signals of each of these polymerization, inhibition of topoisomerase activity, and intercalation drugs. or alkylation of DNA. Abrogation of Trk tyrosine kinase activity may be an important tool In our NB model system, in which Tet represses the levels of TrkB, in improving chemotherapeutic efficacy, as seen by the effect of we showed that the ability of low doses of BDNF (10–33 ng/ml) to K252a on BDNF activation of TrkB. Although K252a has not been protect NB cells from apoptosis was dependent on the level of TrkB used clinically, CEP-751, a K252a derivative with selectivity for Trk expression (Fig. 5). Moreover, at a constant TrkB level, BDNF tyrosine kinase, has been shown to inhibit growth in TrkB-expressing protection was dose-dependent (Fig. 5). This is of interest because we NB xenografts in nude mice (38). Inhibition of Trk tyrosine kinase have shown that in drug-resistant NB cells there is a direct correlation would abrogate all of the downstream signaling paths (MAPK, between increasing drug resistance and BDNF mRNA, whereas TrkB PLC-␥, and PI3K). We find that inhibition of the PI3K path is the only levels remained constant (7). This indicates that TrkB-expressing NB one required to block the ability of BDNF/TrkB to make cells resistant cells that constitutively express high levels of BDNF or metastasize to to chemotherapy. Utilization of an inhibitor with more restricted sites rich in BDNF or other NTs that activate TrkB may be more specificity to a downstream signaling path might be expected to have resistant to chemotherapy. In rodents, BDNF is expressed at high less or different toxicities. In the future we will analyze whether levels in adult brain, lung, and muscle. NB patients with metastases to differential TrkB expression affects NB tumor growth in an in vivo these sites have a shortened event-free survival (30). Furthermore, NB model (39) and assess the effects of the PI3K inhibitor LY294002. BDNF is highly expressed in active osteoblasts in bone, and in the LY294002 has been used in an in vivo ovarian cancer model to proliferating and mature zones of the epiphyseal growth plate (31) demonstrate growth inhibition with minimal toxicities (40). Our future raising the possibility that BDNF may contribute to the chemorefrac- studies will be aimed at determining the downstream targets of PI3K tiveness of bony lesions in NB (32). that are necessary to mediate BDNF/TrkB chemoprotection because a Our finding that K252a, a relatively selective Trk tyrosine kinase more precise delineation of the specific signaling path will lead to inhibitor, is able to block the ability of BDNF to rescue cells from more targeted therapy for the treatment of NB. chemotherapy-induced cell death indicates that activation of the Trk tyrosine kinase is necessary to mediate the protective effects of REFERENCES BDNF. Using a pharmacologic approach we have also shown that inhibition of the PI3K path is able to prevent BDNF rescue. Depend- 1. Brodeur, G. M., Pritchard, J., Berthold, F., Carlsen, N. L., Castel, V., Castelberry, R. P., De Bernardi, B., Evans, A. E., Favrot, M., Hedborg, F., and et al. Revisions of ing on the model system and insult used, different downstream sig- the international criteria for neuroblastoma diagnosis, staging, and response to treat- naling intermediaries of the Trk tyrosine kinase have been shown to ment. J. Clin. Oncol., 11: 1466–1477, 1993. mediate the protective effects of BDNF. For example, BDNF has 2. Nakagawara, A., Arima-Nakagawara, M., Scavarda, N. J., Azar, C. G., Cantor, A. B., and Brodeur, G. M. Association between high levels of expression of the TRK gene rescued neurons from death in culture with low KCl (33) through the and favorable outcome in human neuroblastoma. N. Engl. J. Med., 328: 847–854, PI3K pathway. In neonatal brains, the MAPK pathway was respon- 1993. 3. Brodeur, G. M., Maris, J. M., Yamashiro, D. J., Hogarty, M. D., and White, P. S. sible for mediating BDNF rescue from hypoxic-ischemic injury (34). Biology and genetics of human neuroblastomas. J. Pediatr. Hematol. Oncol., 19: Furthermore, both the MAPK and PI3K pathways have been impli- 93–101, 1997. 6762

4. Matsumoto, K., Wada, R. K., Yamashiro, J. M., Kaplan, D. R., and Thiele, C. J. 23. Kim, C. J., Matsuo, T., Lee, K. H., and Thiele, C. J. Up-regulation of insulin-like Expression of brain-derived neurotrophic factor and p145TrkB affects survival, growth factor-II expression is a feature of TrkA but not TrkB activation in SH-SY5Y differentiation, and invasiveness of human neuroblastoma cells. Cancer Res., 55: neuroblastoma cells. Am. J. Pathol., 155: 1661–1670, 1999. 1798–806, 1995. 24. Thiele, C., Cohen, P., and Israel, M. Regulation of c-myb expression in human 5. Feng, X., Jiang, H., Baik, J. C., Edgar, C., and Eide, F. F. BDNF dependence in neuroblastoma cells during retinoic acid induced differentiation. Mol. Cell. Biol., 8: neuroblastoma. J. Neurosci. Res., 64: 355–363, 2001. 1677–1683, 1988. 6. Lucarelli, E., Kaplan, D., and Thiele, C. J. Activation of trk-A but not trk-B signal 25. Kaplan, D. R., Matsumoto, K., Lucarelli, E., and Thiele, C. J. Induction of TrkB by transduction pathway inhibits growth of neuroblastoma cells. Eur. J. Cancer, 33: retinoic acid mediates biologic responsiveness to BDNF and differentiation of human 2068–2070, 1997. neuroblastoma cells. Eukaryotic Signal Transduction Group. Neuron, 11: 321–331, 7. Scala, S., Wosikowski, K., Giannakakou, P., Valle, P., Biedler, J. L., Spengler, B. A., 1993. Lucarelli, E., Bates, S. E., and Thiele, C. J. Brain-derived neurotrophic factor protects 26. Mosmann, T. Rapid colorimetric assay for cellular growth and survival: application neuroblastoma cells from vinblastine toxicity. Cancer Res., 56: 3737–3742, 1996. to proliferation and cytotoxicity assays. J. Immunol. Methods, 65: 55–63, 1983. 8. Middlemas, D. S., Kihl, B. K., Zhou, J., and Zhu, X. Brain-derived neurotrophic 27. Kaplan, D. R. Studying signal transduction in neuronal cells: the Trk/NGF system. factor promotes survival and chemoprotection of human neuroblastoma cells. J. Biol. Prog. Brain Res., 117: 35–46, 1998. Chem., 274: 16451–60, 1999. 28. Lindvall, O., Kokaia, Z., Bengzon, J., Elmer, E., and Kokaia, M. Neurotrophins and 9. Smeyne, R. J., Klein, R., Schnapp, A., Long, L. K., Bryant, S., Lewin, A., Lira, S. A., brain insults. Trends Neurosci., 17: 490–496, 1994. and Barbacid, M. Severe sensory and sympathetic neuropathies in mice carrying a 29. Cortazzo, M. H., Kassis, E. S., Sproul, K. A., and Schor, N. F. Nerve growth factor disrupted Trk/NGF receptor gene. Nature (Lond.), 368: 246–249, 1994. (NGF)-mediated protection of neural crest cells from antimitotic agent-induced ap- 10. Miller, F. D., and Kaplan, D. R. Neurotrophin signalling pathways regulating neu- optosis: the role of the low-affinity NGF receptor. J. Neurosci., 16: 3895–3899, 1996. ronal apoptosis. Cell Mol. Life Sci., 58: 1045–1053, 2001. 30. DuBois, S. G., Kalika, Y., Lukens, J. N., Brodeur, G. M., Seeger, R. C., Atkinson, 11. Persson, H., and Ibanez, C. F. Role and expression of neurotrophins and the trk family J. B., Haase, G. M., Black, C. T., Perez, C., Shimada, H., Gerbing, R., Stram, D. O., of tyrosine kinase receptors in neural growth and rescue after injury. Curr. Opin. and Matthay, K. K. Metastatic sites in stage IV and IVS neuroblastoma correlate with Neurol. Neurosurg., 6: 11–18, 1993. age, tumor biology, and survival. J. Pediatr. Hematol. Oncol., 21: 181–189, 1999. 12. Heumann, R. Neurotrophin signalling. Curr. Opin. Neurobiol., 4: 668–679, 1994. 31. Yamashiro, T., Fukunaga, T., Yamashita, K., Kobashi, N., and Takano-Yamamoto, T. 13. Johnson, E. M., Jr. Destruction of the sympathetic nervous system in neonatal rats and Gene and protein expression of brain-derived neurotrophic factor and TrkB in bone hamsters by vinblastine: prevention by concomitant administration of nerve growth and cartilage. Bone, 28: 404–409, 2001. factor. Brain Res., 141: 105–118, 1978. 32. Bhogate, B. M., Samuel, A. M., and Ramanathan, P. Bone scans in neuroblastoma. 14. Shimohama, S., Tamura, Y., Akaike, A., Tsukahara, T., Ohara, O., Watanabe, S., and Indian J. Cancer, 30: 5–9, 1993. Kimura, J. Brain-derived neurotrophic factor pretreatment exerts a partially protective 33. Bhave, S. V., Ghoda, L., and Hoffman, P. L. Brain-derived neurotrophic factor effect against glutamate-induced neurotoxicity in cultured rat cortical neurons. Neu- mediates the anti-apoptotic effect of NMDA in cerebellar granule neurons: signal rosci. Lett., 164: 55–58, 1993. transduction cascades and site of ethanol action. J. Neurosci., 19: 3277–3286, 1999. 15. Wozniak, W. Brain-derived neurotrophic factor (BDNF): role in neuronal develop- 34. Han, B. H., and Holtzman, D. M. BDNF protects the neonatal brain from hypoxic- ment and survival. Folia Morphol (Warsz)., 52: 173–181, 1993. ischemic injury in vivo via the ERK pathway. J. Neurosci., 20: 5775–5781, 2000. 16. Lindholm, D., Dechant, G., Heisenberg, C. P., and Thoenen, H. Brain-derived 35. Skaper, S. D., Floreani, M., Negro, A., Facci, L., and Giusti, P. Neurotrophins rescue neurotrophic factor is a survival factor for cultured rat cerebellar granule neurons and cerebellar granule neurons from oxidative stress-mediated apoptotic death: selective protects them against glutamate-induced neurotoxicity. Eur J. Neurosci., 5: 1455– involvement of phosphatidylinositol 3-kinase and the mitogen-activated protein ki- 1464, 1993. nase pathway. J. Neurochem., 70: 1859–1868, 1998. 17. Ventimiglia, R., Mather, P. E., Jones, B. E., and Lindsay, R. M. The neurotrophins 36. Klocker, N., Kermer, P., Weishaupt, J. H., Labes, M., Ankerhold, R., and Bahr, M. BDNF, NT-3 and NT-4/5 promote survival and morphological and biochemical Brain-derived neurotrophic factor-mediated neuroprotection of adult rat retinal gan- differentiation of striatal neurons in vitro. Eur J. Neurosci., 7: 213–222, 1995. glion cells in vivo does not exclusively depend on phosphatidyl-inositol-3Ј-kinase/ 18. Acheson, A., Conover, J. C., Fandl, J. P., DeChiara, T. M., Russell, M., Thadani, A., protein kinase B signaling. J. Neurosci., 20: 6962–6967, 2000. Squinto, S. P., Yancopoulos, G. D., and Lindsay, R. M. A BDNF autocrine loop in 37. Encinas, M., Iglesias, M., Llecha, N., and Comella, J. X. Extracellular-regulated adult sensory neurons prevents cell death. Nature (Lond.), 374: 450–453, 1995. kinases and phosphatidylinositol 3-kinase are involved in brain-derived neurotrophic 19. Wang, W., Salvaterra, P. M., Loera, S., and Chiu, A. Y. Brain-derived neurotrophic factor-mediated survival and neuritogenesis of the neuroblastoma cell line SH- factor spares choline acetyltransferase mRNA following axotomy of motor neurons in SY5Y. J. Neurochem., 73: 1409–1421, 1999. vivo. J. Neurosci. Res., 47: 134–143, 1997. 38. Evans, A. E., Kisselbach, K. D., Liu, X., Eggert, A., Ikegaki, N., Camoratto, A. M., 20. Kume, T., Kouchiyama, H., Kaneko, S., Maeda, T., Akaike, A., Shimohama, S., Dionne, C., and Brodeur, G. M. Effect of CEP-751 (KT-6587) on neuroblastoma Kihara, T., Kimura, J., Wada, K., and Koizumi, S. BDNF prevents NO mediated xenografts expressing TrkB. Med. Pediatr. Oncol., 36: 181–184, 2001. glutamate cytotoxicity in cultured cortical neurons. Brain Res., 756: 200–204, 1997. 39. Khanna, C., Jaboin, J., Drakos, E., Tsokos, M., and Thiele, C. J. Biologically relevant 21. Cheng, B., and Mattson, M. P. NT-3 and BDNF protect CNS neurons against orthotopic neuroblastoma xenograft models: primary adrenal tumor growth and metabolic/excitotoxic insults. Brain Res., 640: 56–67, 1994. spontaneous distant metastasis. In Vivo, 16: 77–85, 2002. 22. Courtney, M. J., Akerman, K. E., and Coffey, E. T. Neurotrophins protect cultured 40. Hu, L., Hofmann, J., Lu, Y., Mills, G. B., and Jaffe, R. B. Inhibition of phosphati- cerebellar granule neurons against the early phase of cell death by a two-component dylinositol 3Ј-kinase increases efficacy of Paclitaxel in in vitro and in vivo ovarian mechanism. J. Neurosci., 17: 4201–4211, 1997. cancer models. Cancer Res., 62: 1087–1092, 2002.

6763

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2002 American Association for Cancer Research. Brain-derived Neurotrophic Factor Activation of TrkB Protects Neuroblastoma Cells from Chemotherapy-induced Apoptosis via Phosphatidylinositol 3′-Kinase Pathway

Jerry Jaboin, Chong Jai Kim, David R. Kaplan, et al.

Cancer Res 2002;62:6756-6763.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/62/22/6756

Cited articles This article cites 39 articles, 11 of which you can access for free at: http://cancerres.aacrjournals.org/content/62/22/6756.full#ref-list-1

Citing articles This article has been cited by 23 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/62/22/6756.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/62/22/6756. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.