(2008) 27, 2910–2922 & 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc ORIGINAL ARTICLE Inhibitors of mammalian target of rapamycin downregulate MYCN protein expression and inhibit growth in vitro and in vivo

JI Johnsen1,6, L Segerstro¨ m1,6, A Orrego2, L Elfman1, M Henriksson3,BKa˚ gedal4, S Eksborg1, B Sveinbjo¨ rnsson1,5 and P Kogner1

1Department of Woman and Child Health, Karolinska Institutet, Childhood Research Unit, Stockholm, Sweden; 2Department of Oncology and Pathology, Karolinska Institutet, Stockholm, Sweden; 3Department of Microbiology, Tumor and Cellbiology, Karolinska Institutet, Stockholm, Sweden; 4Division of Clinical Chemistry, Faculty of Health Sciences, Linko¨ping University, Sweden and 5Department of Cell Biology and Histology, University of Tromso¨, Tromso¨, Norway

Mammalian target of rapamycin (mTOR) has been shown the most common and deadly solid tumor of childhood to play an important function in cell proliferation, (Brodeur, 2003). Amplification of the MYCN oncogene metabolism and tumorigenesis, and proteins that regulate is associated with rapid tumor progression and fre- signaling through mTOR are frequently altered in human quently detected in advanced-stage neuroblastoma, but . In this study we investigated the phosphorylation is also a major negative prognostic factor in localized status of key proteins in the PI3K/AKT/mTOR pathway tumors (Schwab et al., 2003). Advanced-stage tumors and the effects of the mTOR inhibitors rapamycin and and those with MYCN amplification show typically CCI-779 on neuroblastoma tumorigenesis. Significant emergence of treatment resistance, and alternative expression of activated AKT and mTOR were detected treatment strategies for these patients are therefore in all primary neuroblastoma tissue samples investigated, urgently needed. but not in non-malignant adrenal medullas. mTOR Proteins regulating signaling through the phosphatidy- inhibitors showed antiproliferative effects on neuroblasto- linositol 3-kinase (PI3K)/AKT/mTOR pathway are fre- ma cells in vitro. Neuroblastoma cell lines expressing high quently altered in various cancers (Hennessy et al., 2005; levels of MYCN were significantly more sensitive to Cully et al., 2006). Activation of PI3K, AKT or mTOR is mTOR inhibitors compared to cell lines expressing low associated with resistance to apoptosis, increased cell MYCN levels. Established neuroblastoma tumors treated proliferation and deregulated cellular energy metabolism with mTOR inhibitors in vivo showed increased apoptosis, (Hennessy et al., 2005; Cully et al., 2006). mTOR, a serine/ decreased proliferation and inhibition of angiogenesis. threonine kinase, is a central regulator of cell growth and Importantly, mTOR inhibitors induced downregulation of proliferation by controlling protein translation, cytoskele- vascular endothelial growth factor A (VEGF-A) secretion, ton organization and energy metabolism (Hay, 2005; cyclin D1 and MYCN protein expression in vitro and Hennessy et al., 2005; Cully et al., 2006). mTOR regulates in vivo. Our data suggest that mTOR inhibitors have translation by phosphorylation of ribosomal p70S6 kinase therapeutic efficacy on aggressive MYCN amplified 1 (S6K1), thus allowing translation of ribosomal proteins, . and cap-dependent translation through phosphorylation Oncogene (2008) 27, 2910–2922; doi:10.1038/sj.onc.1210938; of eukaryotic translation initiation factor 4E-binding published online 19 November 2007 protein 1 (4E-BP1) (Hay, 2005; Hennessy et al., 2005; Cully et al., 2006) Keywords: neuroblastoma; mTOR; rapamycin; CCI-779; In this study we investigated the activation status of MYCN; preclinical key protein involved in the PI3K/AKT/mTOR signaling pathway in neuroblastoma primary tumors and the effects of mTOR inhibition on neuroblastoma cell growth in vitro and in vivo.

Introduction Results Neuroblastoma, an embryonic tumor derived from primitive cells of the sympathetic nervous system, is Expression of phosphorylated AKT and mTOR in primary neuroblastoma, ganglioneuroma and Correspondence: L Segerstro¨ m, Childhood Cancer Research Unit, non-malignant adrenal tissue samples Q6:05, Department of Women & Child Health, Karolinska Institutet, We investigated 30 primary neuroblastoma tissue S-171-76, Stockholm, Sweden. E-mail: [email protected] samples from different biological subsets and all clinical 6These authors contributed equally to this work. Received 4 April 2007; revised 21 September 2007; accepted 17 October stages for the expression of activated AKT and mTOR 2007; published online 19 November 2007 using antibodies directed against phosphorylated AKT mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2911 (pAKT ser473) and mTOR (pmTOR ser2448). All 30 resulting in a potent inhibition of mTOR signaling samples analysed showed specific expression of pAKT (Hennessy et al., 2005). The effect of rapamycin or CCI- and pmTOR in the cytoplasm (Table 1, Figure 1). 779 on proliferation was investigated in seven neuro- Four ganglioneuromas were investigated and showed blastoma cell lines, of which three have MYCN gene pAKT and pmTOR immunopositivity in the tumor- amplification and express MYCN protein, three are not derived ganglion cells but not in the surrounding benign MYCN amplified but express MYCN protein and one stroma (Table 1, Figure 1). Non-malignant adrenals has neither MYCN gene amplification nor expresses high from children showed weak staining for pAKT and levels of MYCN protein (Table 2; Beppu et al., 2004). All pmTOR in the cortex whereas the medulla, where the neuroblastoma cell lines demonstrated a concentration- majority of primary neuroblastoma arises, was negative dependent decrease in cell viability after 48 h of exposure (Table 1, Figure 1). Antibodies directed against full- (Figures 2a and b). Concentrations associated with 50% length AKT and mTOR were used as controls to growth inhibition (biologic EC50) ranged from 10.6 to eliminate differences in the overall expression levels of 15.2 mM (median 13.7 mM) for rapamycin and 9.5 to the proteins (Figure 1). 19.1 mM (median 12.0 mM) for CCI-779 (Table 2). Neuro- blastoma cell lines containing MYCN amplifications or expressing high levels of MYCN protein (Beppu et al., Inhibition of mTOR decrease neuroblastoma cell 2004) were significantly more sensitive to treatment with proliferation and clonogenic capacity rapamycin or CCI-779 compared to cell lines expressing Rapamycin and its ester-analog CCI-779 both form a low MYCN protein levels (rapamycin; IMR-32 vs SK-N- complex with FK506-binding protein 12 and mTOR, SH Po0.01, SK-N-DZ vs SH-SY5Y Po0.001, SH-SY5Y

Table 1 Clinical features of tumors and control samples investigated for pAKT and pmTOR Sample Diaa Age Sex Stage MYCN 1p DNA High-riskc Outcome Survival PAKTd PmTORd (months) INSSb amplifications deletion Ploidy (months) (ser473) (ser2448)

1NBe 21 M 1 No No 4n No NEDf 159+ + + 2 NB 123 F 1 No No 3n No NED 131+ + + 3 NB 7 F 1 Yes Yes 2n No DODg 8++ 4 NB 13 M 1 No No No NED 41+ + + 5 NB 18 F 1 No No No NED 153+ + + 6 NB 31 M 2B No No 3n No NED 165+ + + 7 NB 33 F 2A No No 3n No NED 179+ + + 8 NB 8 F 2 No No 3n No NED 68+ + + 9 NB 110 M 2 No No 2n No NED 58+ + + 10 NB 5 F 2 No No 3n No AWDh 58+ + + 11 NB 103 F 2B No No 2n No NED 173+ + + 12 NB 6 M 3 No NDi 3n No NED 174+ + + 13 NB 12 F 3 No No 5n No NED 157+ + + 14 NB 0 M 3 No No 3n No DOCj 0++ 15 NB 79 M 3 Yes Yes 3n Yes NED 96+ + + 16 NB 43 F 3 No No 3n No NED 65+ + + 17 NB 43 F 3 No No 3n No NED 65+ + + 18 NB 35 F 4 No Yes Yes NED 57+ + + 19 NB 136 M 4 Yes Yes 2n Yes DOD 10 + + 20 NB 39 F 4 Yes Yes 2n Yes DOD 9 + + 21 NB 35 F 4 No Yes Yes NED 57+ + + 22 NB 28 M 4 Yes Yes 3n Yes NED 109+ + + k 23a NB 8 M 4M No No 3n No NED 59+ + + 23b NB 8 M 4M No No 4n/5n No NED 59+ + + 24 NB 41 F 4 No Yes 4n Yes NED 93+ + + 25 NB 22 M 4 Yes Yes Yes DOD 44 + + 26 NB 50 F 4 Yes Yes Yes DOD 18 + + 27 NB 0 M 4S No No 3n No NED 152+ + + 28 NB 10 M 4S No No 3n No NED 75+ + + 29 NB 0 M 4S No ND 4n No NED 174+ + + 30 NB 10 M 4S No No 3n No NED 75+ + + 31 GNl 145 M NED 108+ + + 32 GN 30 F AWD 92+ + + 33 GN 59 F NED 75+ + + 34 GN 137 M NED 122+ + + 35 ADRm 19 F ÀÀ 36 ADR 25 F ÀÀ 37 ADR 12 F ÀÀ aDiagnosis. bINSS, International Neuroblastoma Staging System. cPatient fulfilling clinico-biological criteria to obtain high-risk therapy. dPhosphorylated protein expression as assed by immunnohistochemistry according to Materials and methods. eNeuroblastoma. fNo evidence of disease. gDead of disease. hAlive with disease. iNot determined. jDead of surgical complications. kMultifocal primary. lGanglioneuroma. mNon- malignant adrenal gland.

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2912 Non-malignant Neuroblastoma Ganglioneuroma adrenal

pAKT (ser473)

AKT

pmTOR (ser2448)

mTOR

Figure 1 Immunohistochemistry of phosphorylated AKT(ser473), phosphorylated mTOR(ser2448), AKT and mTOR in neuroblastoma primary tumors, ganglioneuromas and non-malignant adrenals from children, showing specific staining of pAKT(ser473, top row) and pmTOR(ser2448, third row) in the cytoplasm of a neuroblastoma (sample no. 8, Table 1) and in differentiated ganglion cells of a benign ganglioneuroma (sample no 31, Table 1). Non-malignant adrenal showed weak staining of pAKT and pmTOR in the cortex but not in the medulla (sample no 36, Table 1). Staining of AKT (second row) and mTOR (bottom row) showed no differences in protein expression.

vs SK-N-SH Po0.001, CCI-779; IMR-32 vs SK-N-FI significant dose-dependent inhibition of colony forma- Po0.01, SK-N-DZ vs SK-N-FI Po0.001 and SK-N-DZ tion (Po0.001; Figures 2e and f). Examination of low- vs SK-N-AS Po0.01). No significant differences in dose, long-term drug exposure showed that both mTOR sensitivity was observed for the MYCN amplified, multi- inhibitors significantly inhibited colony formation resistant, mutated SK-N-BE(2) cell line compared to (Po0.001; Figures 2e and f). the remaining panel of cell lines. To investigate the significance of neuroblastoma MYCN expression for mTOR inhibition, we examined mTOR inhibitors induce cell cycle arrest and apoptosis of the effect of rapamycin or CCI-779 on the proliferation neuroblastoma cells of human neuroblastoma SH-EP cells containing a To study the mechanisms of mTOR inhibitors on tetracycline-regulated MYCN transgene (Tet21N cells; neuroblastoma cell growth, we evaluated the effect of Lutz et al., 1996). As shown in Figures 2c and d, Tet21N rapamycin or CCI-779 on cell cycle progression and cells expressing MYCN (MYCN ON) was significantly apoptosis. more sensitive to treatment with mTOR inhibitors as A pronounced accumulation of SH-SY5Y and SK-N-AS compared to Tet21N cells not expressing MYCN cells with hypodiploid DNA content (sub-G1) was observed (MYCN OFF, Po0.001). after CCI-779 treatment and SH-SY5Y cells treated Clonogenic assay on three of the neuroblastoma with rapamycin, whereas only a minor accumulation of cell lines further supported the antitumor effects of hypodiploid cells was observed in SK-N-AS cells treated rapamycin and CCI-779, of which both induced a with rapamycin (Table 3). Moreover, rapamycin induced a

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2913 Table 2 Rapamycin and CCI-779 treatment of neuroblastoma cell lines in vitro Rapamycin CCI-779

MYCN amplifications EC50a (mM) s.d. VC %b EC50a (mM) s.d. VC %b

SK-N-AS No 12.81 0.49 3.84 12.79 0.21 1.66 IMR-32 Yes 11.18 0.41 3.65 10.59 0.71 6.72 SK-N-FI No 14.58 0.45 3.12 19.05 0.75 3.95 SK-N-DZ Yes 15.23 0.48 3.14 9.47 0.31 3.23 SH-SY5Y No 10.56 0.63 5.95 10.49 0.60 5.72 SK-N-BE(2) Yes 13.66 0.31 2.25 12.07 0.52 4.31 SK-N-SH No 15.24 0.21 1.38 11.96 0.81 6.80 Median 13.66 0.45 3.14 11.96 0.60 4.31

Cell growth was measured after 48 h of treatment with rapamycin or CCI-779 using the colorimetric MTT assay. aEffective concentration that b inhibits 50% of cell proliferation (EC50). Variable coefficient. transient G1 arrest of SK-N-AS cells, whereas CCI-779 in vivo, we examined the effects on mTOR target treatment induced an accumulation of S-phase cells proteins. Western blotting to detect phosphorylation (Table 3). of the mTOR downstream target proteins, S6K1 and To determine whether the reduction in neuroblastoma 4E-BP1, revealed that both proteins displayed reduced cell viability was due to apoptosis, western blotting to phosphorylation after treatment with either mTOR detect proteins in the apoptotic cascade was performed inhibitor compared to untreated cells (Figure 5a). on rapamycin or CCI-779-treated cells. Activation of Analysis of SH-SY5Y xenografts by immunohisto- caspase-3 and subsequent cleavage of poly (ADP)-ribose chemistry demonstrated a reduced expression of polymerase (PARP), its downstream substrate, was pmTOR(ser2448) and pS6K1(thr389) in tumors isolated evident in both cell lines treated with either mTOR from mice treated with either of the mTOR inhibitors inhibitor (Figure 3). (Figure 5b). Rapamycin has been reported to activate AKT (Sun mTOR inhibitors have profound effects on the growth of et al., 2005). Therefore, we investigated the effect of established neuroblastoma xenografts mTOR inhibitors on the phosphorylation of AKT in neuro- To investigate the therapeutic effects of mTOR inhibi- blastoma cells. Western blotting using a pAKT(ser473) tors on neuroblastoma growth in vivo, nude mice specific antibody revealed a small increase of AKT carrying SH-SY5Y xenografts were treated with either phosphorylation and a reduction of pmTOR (ser2448) in rapamycin or CCI-779. Tumor growth was significantly all neuroblastoma cell lines investigated when treated with inhibited after treatment for 3 days with rapamycin rapamycin or CCI-779 (Figure 5c). Also, immunohisto- (P ¼ 0.016) or 4 days with CCI-779 (P ¼ 0.038) com- chemistry of xenografts isolated from mice treated with pared with untreated controls (Figure 4a). Tumor either rapamycin or CCI-779 demonstrated increased volumes were reduced by 56 and 71% after rapamycin phosphorylation of AKT. or CCI-779 treatment, respectively, compared to con- trols at the end of treatment (Figure 4a). Quantification showed that both mTOR inhibitors Combination of mTOR inhibitors with the PI3K-inhibitor significantly reduced expression of the nuclear prolifera- LY294002 enhances neuroblastoma growth inhibition tion marker Ki-67 (Po0.001), increased activation of The activation of AKT by mTOR inhibitors in caspase-3 (Po0.001) and decreased microvessel density neuroblastoma cells may attenuate the growth inhibi- (Po0.001) compared to untreated controls (Figure 4b). tory effect of these drugs. To investigate if this potential negative feedback mechanism had any effect on neuroblastoma cells, we treated SK-N-AS and SK-N- Inhibition of mTOR reduce neuroblastoma VEGF-A BE(2) cells with a combination of rapamycin or CCI-779 secretion and the PI3K inhibitor, LY294002. Both cell lines were Since inhibition of mTOR resulted in decreased vessel treated with increasing concentrations of rapamycin, density, we evaluated the effect of rapamycin or CCI- CCI-779 and LY294002. Single drug activity of 779 on the secretion of vascular endothelial growth LY294002 in SH-SY5Y and SK-N-BE(2) cells were factor A (VEGF-A) in four neuroblastoma cell lines. determined in initial experiments (data not shown). Concentrations of VEGF-A were significantly lower in Fixed concentration ratios of the drugs were used with cell-free culture media from rapamycin or CCI-779- serial dilutions for combination and single-drug treat- treated cells than in media from corresponding un- ment. As summarized in Table 4, showing the combina- treated controls (P ¼ 0.0022; all groups, Figure 4c). tion index (CI) at EC70, LY294002 induced a synergistic or additive cytotoxic effect in combination with Effect of rapamycin and CCI-779 on mTOR target rapamycin or CCI-779 in both the MYCN amplified proteins neuroblastoma cell line SK-N-BE(2) as well as in the Having established that both rapamycin and CCI-779 non-MYCN amplified cell line SK-N-AS. These results inhibited growth of neuroblastoma cells in vitro and indicate that the combination of rapamycin or CCI-779

Oncogene 2914 Oncogene

Cell growth Cell growth (% of control) (% of control) r( or fsxreplicates six of exposure. lngnccpct fnuolsoacls HS5,S--SadS--E2 el eetetdwt ,1 r15 or 10 5, with treated were cells SK-N-BE(2) and treated SK-N-AS were SH-SY5Y, cells. cells neuroblastoma Tet21N of capacity construct. human clonogenic are expression Tet21N MYCN cells. Tet21N tetracycline-regulated on a inhibitors mTOR with of transfected Effect ( results. with stably similar with cells times SH-EP three repeated neuroblastoma was experiment the and shown ( of concentrations 2 Figure 100 CCI-779 Rapamycin 100 25 50 75 25 50 75 0 0 f C-7,rsetvl o 8h rwt 0.5 with or h, 48 for respectively CCI-779, ) c

aayi r( or rapamycin ) Number of clones Number of clones 100 100 0 25 50 75 8 25 50 75 0 fet fmO niioso h rwho erbatm cells neuroblastoma of growth the on inhibitors mTOR of Effects 0 Rapamycin concentration (

Control Control 10 5 ±

5 Rapamycin (

5 Ii hw.Teeprmn a eetdtotmswt iia eut.Efc fmO niioso the on inhibitors mTOR of Effect results. similar with times two repeated was experiment The shown. is CI 95% 5 a µ µ aayi r( or rapamycin ) M / 48h M / 48h

d 10 SH-SY5Y 10 10 C-7 nteasneo rsneo oyyln.Cl rwhwsmaue yMTasy.Ma survival Mean assays. MTT by measured was growth Cell doxycycline. of presence or absence the in CCI-779 ) µ µ M / 48h M / 48h TRihbtr nnuolsoatherapy neuroblastoma in inhibitors mTOR

15 15 15

µ µ µ 15 0.5 M / 48h 0.5 M / 48h M) µ µ M / 7 days M / 7 days b 20 C-7 o 8hadcl rwhwsmaue sn T.Ma uvvlo i elcsis replicas six of survival Mean MTT. using measured was growth cell and h 48 for CCI-779 ) 20 µ M) 25 Tet21N MYCNON Rapamycin Tet21N MYCNOFF Rapamycin m M Number of clones Number of clones Johnsen JI SK-N-AS SK-N-SH SK-N-BE(2) SH-SY5Y SK-N-DZ SK-N-F1 IMR-32 25 50 75 aayi rCI79fr7dy oeaieln-emefcso o-oedrug low-dose of effects long-term examine to days 7 for CCI-779 or rapamycin 25 50 75 0 0

Control Control tal et 5 5 µ µ M / 48h M / 48h 10 10 Cell growth Cell growth

µ µ SK-N-AS M / 48h M / 48h (% of control) (% of control) 100 15 100 25 50 75 15 25 50 75 0 µ µ 0 0.5 M / 48h 0.5 M / 48h nvitro in µ µ M / 7 days M / 7 days 8 erbatm el eetetdwt increasing with treated were cells Neuroblastoma . CCI-779 concentration ( 10 CCI-779 (

Number of clones Number of clones 100 100 25 50 75 25 50 75 0 0 15 µ

Control Control M)

5 5 µ µ

M / 48h M / 48h 20 SK-N-BE(2) 10 µ 10 M) µ µ M / 48h M / 48h 25 m 15

M 15 Tet21N MYCNON CCI-779 µ µ Tet21N MYCNOFF CCI-779

( 0.5 M / 48h 0.5 M / 48h e

rapamycin ) µ µ M / 7 days M / 7 days SK-N-SH SK-N-BE(2) SH-SY5Y SK-N-DZ SK-N-F1 IMR-32 SK-N-AS mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2915 Table 3 Cell cycle analysis of rapamycin or CCI-779 treated Moreover, it has been shown that MYCN is destabilized neuroblastoma cells by GSK-3b through phosphorylation, whereas C-MYC

Time (h) Treatment Sub-G1 (%) G1 (%) S(%) G 2 (%) can drive oncogenic cell proliferation through transcrip- tional upregulation of cyclin D1 expression (Kenney SH-SY5Y et al., 2004). To study the potential of mTOR inhibitors 24 Control 1.7 56.2 28.3 13.8 Rapamycin 14.7 44.6 21.9 18.8 to modulate the activity of these proteins we treated CCI-779 3.3 35.6 39.9 21.2 neuroblastoma cells with rapamycin or CCI-779 and observed increased phosphorylation of GSK-3b 48 Control 1.3 47.3 42.2 9.2 and reduced expression of MYCN and cyclin D1 protein Rapamycin 34.1 38.6 21.2 6.1 levels (Figure 5d). To further study the observed in vitro CCI-779 20.3 39.5 24.9 15.2 effect of mTOR inhibitors on MYCN and cyclin D1 72 Control 1.2 57.5 31.8 9.5 protein expression, untreated neuroblastoma xenografts Rapamycin 36.2 36.5 19.3 8.0 were compared to xenografts treated with rapamycin or CCI-779 29.6 31.9 25.6 12.9 CCI-779. Western blotting revealed a reduction of both 96 Control 1.5 70.6 19.6 8.3 MYCN and cyclin D1 proteins in xenografts treated Rapamycin 39.2 32.6 19.9 8.3 with the mTOR inhibitors compared to untreated CCI-779 27.8 31.3 27 13.9 (Figure 5e). Similar, immunohistochemistry of xeno- grafts revealed a reduction of MYCN expression in SK-N-AS tumors treated with mTOR inhibitors (Figure 5b). We 24 Control 2.5 47.9 40.2 9.4 Rapamycin 3.8 64.7 22.8 8.7 also investigated the effect of rapamycin or CCI-779 on CCI-779 8 25.2 55.2 11.6 MYCN and cyclin D1 transcription using real-time PCR. TaqMan analysis revealed no significant differ- 48 Control 2.9 48.7 36.8 11.6 ences in MYCN or cyclin D1 mRNA expression in Rapamycin 3.8 47.7 37.3 11.2 CCI-779 10.3 24.5 54.4 10.7 neuroblastoma cell lines or SH-SY5Y xenograft tumors treated with mTOR inhibitors compared to untreated 72 Control 2.3 53.1 35.6 9 control cells or tumors (data not shown). These results Rapamycin 5.1 47.7 37.4 9.8 indicate that mTOR inhibitors regulate the expression of CCI-779 14.2 29.7 41.6 14.5 MYCN and cycling D1 at a post-transcriptional level. 96 Control 2.8 64.9 26.9 5.4 Rapamycin 4.6 54.3 33.2 7.9 CCI-779 29.1 27.6 29.7 13.6 Discussion

Recent studies indicate that numerous components of SH-SY5Y SK-N-AS the PI3K/AKT/mTOR signaling pathway frequently are Ctrl R CCI Ctrl R CCI targets for amplification, translocations and -19 kD in cancer, with resultant activation of the pathway Cleaved (Hennessy et al., 2005). We found that all primary Caspase 3 -17 kD neuroblastoma tumors investigated expressed activated -116 kD pAKT and pmTOR. We also detected both pAKT and PARP -89 kD pmTOR in differentiated ganglion cells of ganglio- neuroma, but not in the surrounding stroma or in non- β -actin malignant adrenal medullas from children. Activation of Figure 3 Effect of mTOR inhibitors on apoptotic proteins in AKT has recently been suggested to be a factor neuroblastoma cells. SH-SY5Y and SK-N-AS cells were treated predicting poor outcome in neuroblastoma (Opel with 12 mM rapamycin or CCI-779 for 24 h and protein extracts et al., 2007). Neuroblastoma exhibit increased expres- were subjected to western blotting using antibodies detecting cleaved caspase-3 and poly (ADP)-ribose polymerase (PARP). The sion of several growth factor receptors that transmit experiment was repeated twice with similar results and b-actin was their signals through the PI3K/AKT/mTOR pathway used to ensure equal protein loading. (Vanhaesebroeck et al., 2001; Kozma and Thomas, 2002; Cully et al., 2006), such as insulin-like growth factor I receptor (IGF-IR; Singleton et al., 1996; Wang with LY294002 augmented the growth-inhibitory effect et al., 2001; Kim et al., 2004), epidermal growth factor of neuroblastoma cells. receptor (EGFR; Ho et al., 2005), tyrosine receptor kinase B (TrkB; Macdonald et al., 2001; Ho et al., 2002; Jaboin et al., 2002), platelet-derived growth factor mTOR inhibition downregulate expression of MYCN and receptor B (PDGFRB; Eggert et al., 2000; Beppu cyclin D1 protein in vitro and in vivo et al., 2004) and c-KIT (Cohen et al., 1994; Vitali Other studies suggest cyclin D1 as a key target of et al., 2003; Beppu et al., 2004; Uccini et al., 2005). mTOR, and that activation of GSK-3b is critical for Although a number of agents that target the PI3K or regulating cyclin D1 expression (Nelsen et al., 2003; AKT pathway have been developed, no drugs have yet Aguirre et al., 2004; Gera et al., 2004; Dong et al., 2005). progressed to clinical cancer trials, whereas agents that

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2916 Ki-67 Casp-3 BS-1

Control

Rapamycin

CCI-779

Figure 4 Effect of mTOR inhibitors on neuroblastoma xenograft growth in vivo.NMRInu/nu mice engrafted with 30 Â 106 SH-SY5Y cells subcutaneously was randomized to receive rapamycin (5 mg kgÀ1; n ¼ 9), CCI-779 (20 mg kgÀ1; n ¼ 8) intraperitoneally daily for 12 days or no treatment (n ¼ 9), starting at the appearance of palpable tumors of B0.20 ml (mean 0.26 ml). (a) Comparison of tumor volumes from mice treated with rapamycin, CCI-779 or untreated controls (mean±s.d.). (b) Immunohistochemical staining of untreated xenografts (top row), xenografts treated with rapamycin (5 mg kgÀ1, second row) or CCI-779 (20 mg kgÀ1, third row). Proliferation was detected by Ki-67 (left panel; Â 400 magnification) and apoptosis by cleaved caspase-3 (middle panel; Â 400 magnification), whereas BS-1 lectin staining for murine endothelial cells was used for visualizing microvessel density (right panel; Â 200 magnification). Quantification (bottom row) of proliferation, apoptosis and microvessel density in xenografts treated with rapamycin or CCI-779 is shown in the diagrams (c). Effect of mTOR inhibitors on vascular endothelial growth factor A (VEGF-A) secretion by neuroblastoma cells. SH-SY5Y, SK-N-AS, SK-N-BE(2) and IMR-32 cells were incubated with 6 mM rapamycin or CCI- 779 for 72 h, cell free growth media collected and the concentration of VEGF-A was measured using VEGF-A DuoSet ELISA. Media from untreated cells were used as controls.

inhibit mTOR, like rapamycin and its analogs CCI-779 f) showed a concentration-dependent inhibition of SK- and RAD001, have been approved for clinical use and N-AS cell proliferation as well as an inhibition of trials (Hennessy et al., 2005). Inhibitors of mTOR have clonogenic capacity. The reason for this discrepancy been shown to function both as inhibitors of cell might be that some of the neuroblastoma cells are proliferation, leading to cell cycle arrest, and as subjected to autophagy, a process that is closely cytotoxic agents, inducing apoptosis, depending on the regulated by mTOR (Easton and Houghton, 2006). tumor cell type (Cully et al., 2006). We show that We observed significant effects with rapamycin or treatment of neuroblastoma cells with rapamycin or CCI-779 treatment in vitro on all seven neuroblastoma CCI-779 induced apoptosis (Table 3), which was cell lines, regardless of MYCN amplification or expres- preceded by activation of caspase-3 and PARP sion levels. Interestingly, MYCN-amplified cell lines (Figure 3). No activation of caspase-9 was observed were more sensitive to rapamycin or CCI-779 treatment for either treatment, indicating that neuroblastoma cells compared to non-amplified cells, whereas cells with the underwent apoptosis through the extrinsic pathway lowest MYCN-expression showed the least sensitivity to (data not shown). Rapamycin treatment induced a G1 rapamycin treatment (Table 2). To investigate if the arrest whereas CCI-779 induced cell cycle arrest in the S difference in sensitivity to mTOR inhibitors was due to and G2 phase (Table 3). The observed S/G2 arrest may expression of MYCN, we investigated the effect of be due to repression of MYCN since MYCN has been mTOR inhibitors in SH-EP cells containing a stably shown to activate expression of S-phase specific cyclins, transfected tetracycline-inducible MYCN construct with cyclin D1 being one of the major mediators (Yaari (Tet21N cells). Tet21N cells expressing MYCN was et al., 2005). We also noticed that only a minor fraction significantly more sensitive to treatment with mTOR of SK-N-AS cells treated with rapamycin had sub-G1 inhibitors compared to Tet21N cells expressing no DNA content. However, we detected activation of MYCN (Figures 2c and d). Taken together, these data caspase-3 and PARP cleavage in these cells (Table 3, suggest that MYCN is a target for mTOR inhibitors. Figure 3). Moreover, data from the proliferation assay Almost one-fourth of all neuroblastoma patients have (Figures 2a and b) and clonogenic assay (Figures 2e and amplification of MYCN and MYCN is a powerful

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2917 predictor of aggressive biological behavior and poor of a mouse model having a MYCN transgene targeted clinical outcome. Dysregulation of MYCN expression is to the developing neural crest by the tyrosine hydro- critically involved in the pathogenesis of the disease xylase promoter. Depending on gene dose, these mice (Brodeur, 2003). The importance of MYCN amplifica- develop tumors with biological and genetic features that tion and overexpression in malignant progression of are highly similar to the aggressive MYCN amplified neuroblastoma is further supported by the development neuroblastomas seen in humans (Weiss et al., 1997).

SH-SY5Y SK-N-AS Ctrl R CCI Ctrl R CCI p-S6K1 (thr389) S6K1 p4E-BP1 (ser65) 4E-BP1 β-actin

SH-SY5Y SK-N-AS SK-N-BE(2) IMR-32 Ctrl R CCI Ctrl R CCI Ctrl R CCI Ctrl R CCI pmTOR (Ser 2448)

mTOR

pAkt (Ser 473)

Akt

β-actin

SH-SY5Y SK-N-AS SK-N-BE(2) IMR-32 Ctrl R CCI Ctrl R CCI Ctrl R CCI Ctrl R CCI pGSK-3β (ser9)

GSK-3β

MYCN

Cyclin D1

β-actin

Ctrl R CCI

Cyclin D1

MYCN

β-actin

Figure 5 Effects of rapamycin or CCI-779 on mTOR downstream targets and the expression of cyclin D1 and MYCN in neuroblastoma cells and xenografts. (a) Effect of mTOR inhibition on S6K1 and 4E-BP1 phosphorylation. Cells were treated with 12 mM rapamycin or CCI-779 for 24 h, and protein extracts were subjected to western blotting to detect pS6K1(thr389), S6K1, p4E- BP1(ser65) and 4E-BP1. (b) Immunohistochemistry (following page) of xenograft tumors from untreated controls (left panel) or animals treated with rapamycin (5 mg kgÀ1, middle panel) or CCI-779 (20 mg kgÀ1, right panel). Sections were stained with antibodies detecting pAKT (ser473) (top row), AKT (second row), pmTOR (ser2448) (third row), mTOR (fourth row), pS6K1 (thr389) (fifth row), S6K1 (sixth row) and MYCN (bottom row). All  400 magnification. (c) Western blot of AKT and mTOR phosphorylation in neuroblastoma cells treated with mTOR inhibitors. Cells were treated as above and protein extracts were subjected to western blotting to detect pAKT(ser473) and mTOR(ser2448). (d) Effect of rapamycin or CCI-779 on GSK-3b phosphorylation, cyclin D1 and MYCN protein expression. Cells were treated as above, and protein extracts were subjected to western blotting to detect pGSK-3b(ser9), GSK-3b, cyclin D1 and MYCN. (e) Effect of mTOR inhibitors on cyclin D1 and MYCN protein expression in SH-SY5Y xenografts. Protein extracts from frozen xenografts tissue were subjected to western blotting with antibodies detecting cyclin D1 and MYCN. Antibodies detecting unphosphorylated AKT, mTOR, GSK-3b, S6K1 and 4E-BP1 were used as controls to exclude possible differences in total protein expression. b-actin was used to ensure equal protein loading.

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2918 Control Rapamycin CCI-779

pAKT (ser473)

AKT

pmTOR (ser2448)

mTOR

pS6K1 (thr389)

S6K1

MYCN

Figure 5 Continued.

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2919 Table 4 Combination of mTOR inhibitors and the PI3K-inhibitor LY294002 on neuroblastoma cells Combinationsa SK-N-AS SK-N-BE(2)

b b LY294002 (mM) Rapamycin (mM) CCI-779 (mM) n Mean CI at EC70 Effect n Mean CI at EC70 Effect (95% CI) (95% CI)

3.45 3.45 — 5 0.70 (0.61–0.80) Synergistic 5 0.37 (0.32–0.42) Synergistic 6.9 6.9 — 5 0.81 (0.66–0.95) Synergistic 5 0.45 (0.33–0.58) Synergistic 13.8 13.8 — 5 1.01 (0.68–1.33) Additive 5 0.46 (0.41–0.52) Synergistic 3.45 — 3.45 5 0.53 (0.50–0.56) Synergistic 5 0.49 (0.46–0.53) Synergistic 6.9 — 6.9 5 0.76 (0.67–0.86) Synergistic 5 0.65 (0.60–0.71) Synergistic 13.8 — 13.8 5 0.92 (0.76–1.08) Additive 5 0.48 (0.43–0.54) Synergistic aNeuroblastoma cells were treated with a combination of equimolar concentrations of rapamycin, CCI-779 and LY294002 for 48 h and cell survival was measured by the MTT assay. bSynergistic effects are defined as a CI mean statistically significantly lower than 1 and additive effects as a CI mean not significantly higher or lower than 1 (one sample t-test, Po0.05).

Given the important role of MYCN expression in high- Moreover, VEGF-A secreted by neuroblastoma cells risk neuroblastoma and the limited expression in other contributes to the growth of endothelial cells in vitro and postnatal tissues (Grimmer and Weiss, 2006), the to angiogenesis in vivo (Eggert et al., 2002) whereas MYCN protein appears to be an attractive candidate blockade of VEGF-A function is associated with for targeted therapy. We here show that both rapamycin suppression of neuroblastoma growth (Davidoff et al., and CCI-779 have profound effects on MYCN protein 2001; Kim et al., 2002; Segerstrom et al., 2006). VEGF- expression, suggesting that mTOR inhibitors may be A expression has also been correlated with higher levels effective in the treatment of aggressive neuroblastoma of AKT and mTOR phosphorylation in various human expressing high levels of MYCN. This is supported by tumors (Klos et al., 2006). Our data showing that the recent findings that small-molecule inhibitors of rapamycin or CCI-779 treatment reduce the production PI3K, LY294002 or wortmannin, represent an effective of VEGF-A (Figure 4c) suggest that inhibition of preclinical therapy for neuroblastoma through destabi- mTOR may suppress neuroblastoma growth by inhibi- lization of MYCN (Chesler et al., 2006). This further tion of angiogenesis, contributing to the observed indicate that dysregulation of the PI3K/AKT/mTOR reduction in proliferation and increase of apoptosis. pathway are important in the malignant progression of AKT and mTOR are linked to each other via positive neuroblastoma. and negative regulatory feedback loops, which might Treatment of neuroblastoma cells with mTOR in- have been evolved as a protective mechanism to inhibit hibitors also resulted in phosphorylation of GSK-3b and uncontrolled cell survival and proliferation (Hay, 2005). reduced expression of cyclin D1. GSK-3b has been Rapamycin has been shown to activate AKT (Sun et al., shown to be critical for the regulation of cyclin D1 2005). We observed increased phosphorylation of AKT expression and to destabilize MYCN through phos- when neuroblastoma cells or xenograft tumors were phorylation (Nelsen et al., 2003; Aguirre et al., 2004; treated with rapamycin or CCI-779 (Figures 5b and c). Gera et al., 2004; Kenney et al., 2004; Dong et al., 2005). This may potentially be a dilemma when designing Moreover, C-MYC has been shown to transcriptionally anticancer therapies using these compounds. This is upregulate cyclin D1 expression (Yu et al., 2005). Hence, supported by our findings that a combination of mTOR inhibition of mTOR signaling affects the expression of inhibitors with LY294002 inhibited neuroblastoma cell several proteins shown to be important in the tumor- proliferation in a synergistic or additive manner in vitro igenesis of neuroblastoma. (Table 4). Therefore, treatment with agents that We found that rapamycin or CCI-779 inhibited simultaneously target receptors transmitting signals growth of human neuroblastoma xenograft tumors in through the PI3K/AKT/mTOR pathway or AKT itself, nude mice (Figure 4a). Proliferation, apoptosis and in combination with inhibitors of mTOR, may prove to microvessel density was quantified (Figure 4b) to further be more efficacious. understand the basis for this effect. Rapamycin and Taken together, our results suggest that the PI3K/ CCI-779-treated xenografts showed significant reduc- AKT/mTOR signaling pathway is constitutively activated tion of proliferation and increased apoptosis in vivo in neuroblastoma and that mTOR inhibitors targeting key compared to untreated controls (Figure 4b). Treated proteins in this pathway may represent an approach for tumors also had a significant reduction in microvessel the treatment of children with neuroblastoma. density compared to untreated (Figure 4b), a parameter that is previously shown to correlate to poor prognosis (Meitar et al., 1996). One of the most potent inducers of angiogenesis is VEGF-A, which induces endothelial cell Materials and methods proliferation and migration (Ferrara and Alitalo, 1999). Tissue samples and patient characteristics The majority of neuroblastoma cell lines and tumors Thirty neuroblastoma tumor tissue samples from different clinical express VEGF-A and the levels of expression have been (age, stage and risk group) or biological (MYCN amplification, 1p correlated with both disease progression and poor deletion and DNA ploidy) subsets were examined (Table 1). Four prognosis (Meitar et al., 1996; Eggert et al., 2000). ganglioneuromas and three non-malignant adrenal glands from

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2920 children, ages 12–25 months, were also included (Table 1). Ethical Immunohistochemistry approval was obtained by the Karolinska University Hospital Sections were incubated with primary antibody phospho-AKT Research Ethics Committee (Approval no. 03–708). (ser473), phospho-mTOR (ser2448), phospho-S6K1 (thr389), AKT, mTOR and S6K1 (Cell Signaling), Ki-67 (NeoMarkers, Chemicals Fremont, CA, USA), MYCN (Oncogene Research Products, Rapamycin (Sirolimus, LC Laboratories, Woodburn, MA, Darmstadt, Germany) or cleaved caspase-3 (R&D Systems, USA) or CCI-779, an ester-analog of rapamycin (Temsiroli- Abingdon, UK). As a secondary antibody, the HRP-Super- mus, a kind gift from Wyeth Pew River, NY, USA), was Picture Polymer detection kit and matched isotype controls dissolved in 99.5% ethanol while the PI3K-inhibitor were used (Zymed Laboratories Inc., San Francisco, CA, LY294002 (Cell Signaling, Beverly, MA, USA) was dissolved USA). Bandeirea simplifolica (BS-1; Sigma-Aldrich) lectin was in dimethyl sulfoxide. All inhibitors were further diluted in used to highlight murine endothelial cells as described OptiMEM (Gibco BRL, Sudbyberg, Sweden) to the desired previously (Segerstrom et al., 2006; Ponthan et al., 2007). in vitro concentration. For in vivo use of rapamycin and CCI- Proliferation (Ki-67) and apoptosis (cleaved caspase-3) were 779, the stock was diluted in 5% PEG 400 (Sigma-Aldrich, quantified at  400 magnification, whereas microvessel density Solna, Sweden), 5% Tween 20 (BioRad, Sundbyberg, Sweden) (BS-1) was quantified at  200. Fifteen randomly chosen fields and 0.9% sterile saline. Each dose was freshly prepared prior per slide and four slides per group were quantified for each to injection. staining. Ki-67 is presented as the proportion of positively staining cells over the total number of cells, whereas Neuroblastoma cell lines microvessel density and caspase-3 positive cells are presented Human neuroblastoma cell lines were grown in Eagle Minimal as an average number per field. Essential Medium (SH-SY5Y, Sigma-Aldrich) or RPMI 1640 (SK-N-BE(2), SK-N-AS, SK-N-FI, SK-N-SH, SK-N-DZ, Western blotting IMR-32 and Tet21N; Gibco) medium supplemented with Cells were treated with 16 mM rapamycin or CCI-779 and 10% fetal bovine serum (Gibco) and 2 mM L-glutamine proteins extracted on ice either in 1  SDS sample buffer (Sigma-Aldrich) at 37 1C humidified 5% CO2 atmosphere. (BioRad) or in RIPA buffer (Cell Signaling) containing Tet21N are neuroblastoma SH-EP cells containing a tetra- protease inhibitors (Roche Diagnostic, Mannheim, Germany). cycline-regulated NMYC transgene (Tet-Off) and were main- Frozen xenograft tissues were homogenized on ice in RIPA tained as described previously (Lutz et al., 1996). The cells buffer and cleared by repeated centrifugations. Protein were continuously grown as MYCN ON and MYCN were concentrations were measured using Bradford reagent switched OFF ( þ Tet) 24 ho prior to experiments. (BioRad). Equal quantities were separated by SDS–PAGE, transferred to nylon membranes (Millipore Inc., Sundbyberg, Cytotoxic and clonogenic assay Sweden) and probed with antibodies against phospho- Effects of rapamycin, CCI-779 or LY294002 on neuroblasto- AKT(ser473), phospho-mTOR(ser2448), phospho-S6K1(thr389), ma cell growth were determined using a colorimetric 3-(4,5- phospho-4E-BP1(ser65), phospho-GSK-3b(ser9),AKT,mTOR, dimethylthiazol-2-yl)-2,5-diphenyl-tetrazoleum (MTT) assay S6K1, 4E-BP1, GSK-3b, caspase-9, cleaved caspase-3, (Sigma-Aldrich). Five or six parallels of each treatment were PARP, cyclin D1 (Cell Signaling), MYCN (Oncogene Research performed in each experiment. The concentration that Products) and b-actin (Sigma-Aldrich). Anti-mouse IgG or inhibited 50% of cell proliferation (EC50) was calculated. anti-rabbit IgG, conjugated with horseradishperoxidase (Cell To determine clonogenic capacity, neuroblastoma cells were Signaling) were used as secondary antibodies, and Pierce Super seeded, allowed to attach for 24 h and treated with rapamycin Signal (Pierce, Rockford, IL, USA) for chemiluminescent or CCI-779 for 48 h (5, 10 or 15 mM) or 7 days (0.5 mM). Clones detection. were then grown in drug-free media for 10 days, fixed in formaldehyde, stained with Giemsa (Gibco) and colonies (>50 cells) were counted manually. Analysis of cyclin D1, MYCN and b2-microglobulin transcripts RNA was isolated and cDNA synthesized as earlier described (Kagedal et al., 2004). For MYCN mRNA quantification Fluorescence-activated cell sorting the forward primer was ACCCTGAGCGATTCAGATGAT, SH-SY5Y and SK-N-AS cells were treated with 16 mM the reverse primer GTGGTGACAGCCTTGGTGTT and the rapamycin or CCI-779 for 24, 48, 72 or 96 h. Cells were probe TGGAGAAGCGGCGTTCCTCCTC. For cyclin D1 harvested, stained with 40,6-diamidino-2-phenylindole (DAPI) mRNA quantification the forward primer was AACAAACAG and subjected to cell cycle analysis using single parameter ATCATCCGCAAAC, the reverse primer ACCATGGA DNA flow cytometry as described previously (Ponthan et al., GGGCGGATT and the probe TCTGTGCCACAGATGT 2003). GAA. For b2-microglobulin mRNA quantification the for- ward primer was GAGTATGCCTGCCGTGTG, the reverse In vivo xenografts and administration of mTOR inhibitors primer AATCCAAATGCGGCATCT and the probe Nude animal housing and engrafting was done as described CCTCCATGATGCTGCTTACATGTCTC. The probes were previously (Segerstrom et al., 2006). Tumors were measured labeled with FAM at the 50 ends and TAMRA at the 30 ends daily with a digital caliper and tumor volume calculated as (Applied Biosystems, Foster City, CA, USA). Double stranded 2 length  width  0.44. At a tumor volume of X0.2 ml (mean DNA calibrators were synthesized for each transcript accord- 0.26 ml) the animals were randomized to receive either ing to principles described earlier (Kagedal et al., 2004). rapamycin (5 mg kgÀ1) or CCI-779 (20 mg kgÀ1) intraperitone- ally daily for 12 days or no treatment (controls). Tumor materials was fixed in 4% paraformaldehyde and frozen in Vascular endothelial growth factor A measurement liquid nitrogen and stored at 4 or À80 1C, respectively. Animal Neuroblastoma cells were treated with 6 mM rapamycin or experiments were approved by the regional ethics committee CCI-779 for 72 h and cell-free culture supernatant harvested. (N234-05) in accordance with national regulations (SFS Soluble VEGF-A was measured using the VEGF DuoSet 1988:534, SFS 1988:539 and SFS 1988:541). ELISA (R&D Systems) to manufacturers specifications.

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2921 Statistical analysis comparison test were used for analysis of statistical differences EC50 values were evaluated from a plot of survival versus the between two and several independent populations, respec- logarithm of the molar drug concentration, using a standard tively. All statistical tests were two-sided. dose–response curve defined by four parameters, that is, the Testing for synergistic or additive effects of combina- baseline response (Bottom), the maximum response (Top), the tion therapy the data was done as described (Ponthan et al., slope (Hill slope) and the drug concentration. 2007).

^ Response ¼ Bottom þðTop À BottomÞ=ð1 þ 10ððlog EC50ÀxÞÃHill slopeÞÞ ð1Þ Acknowledgements where x is the logarithm of the drug concentration. Non-linear regression analysis was performed by the PCNONLIN We thank M Schwab (German Cancer Research Centre, program (version 2.0). Data was initially fitted to Eq. (1) with DKFZ, Heidelberg, Germany) for providing us with the the Hill slope fixed to À1, but also fitted along with the other Tet21N cell line. This work was supported by grants from the parameters. The choice of the final model was based on the Swedish Childhood Cancer Foundation, The Swedish Cancer F-ratio test. Mann–Whitney U-test and the Kruskal–Wallis Foundation, The Swedish Research Council and Ma¨ rta and test (non-parametric ANOVA) followed by Dunn’s multiple Gunnar V Philipson Foundation.

References

Aguirre D, Boya P, Bellet D, Faivre S, Troalen F, Benard J et al. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. (2005). Exploiting (2004). Bcl-2 and CCND1/CDK4 expression levels predict the the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug cellular effects of mTOR inhibitors in human ovarian carcinoma. Discov 4: 988–1004. Apoptosis 9: 797–805. Ho R, Eggert A, Hishiki T, Minturn JE, Ikegaki N, Foster P et al. Beppu K, Jaboine J, Merchant MS, Mackall CL, Thiele CJ. (2004). (2002). Resistance to mediated by TrkB in neuro- Effect of imatinib mesylate on neuroblastoma tumorigenesis and blastomas. Cancer Res 62: 6462–6466. vascular endothelial growth factor expression. J Natl Cancer Inst 96: Ho R, Minturn JE, Hishiki T, Zhao H, Wang Q, Cnaan A et al. (2005). 46–55. Proliferation of human neuroblastomas mediated by the epidermal Brodeur GM. (2003). Neuroblastoma: biological insights into a clinical growth factor receptor. Cancer Res 65: 9868–9875. enigma. Nat Rev Cancer 3: 203–216. Jaboin J, Kim CJ, Kaplan DR, Thiele CJ. (2002). Brain-derived Chesler L, Schlieve C, Goldenberg DD, Kenney A, Kim G, Mcmillan A neurotrophic factor activation of TrkB protects neuroblastoma et al. (2006). Inhibition of phosphatidylinositol 3-kinase destabilizes cells from chemotherapy-induced apoptosis via phosphatidylinositol Mycn protein and blocks malignant progression in neuroblastoma. 30-kinase pathway. Cancer Res 62: 6756–6763. Cancer Res 66: 8139–8146. Kagedal B, Kullman A, Lenner L, Trager C, Kogner P, Farneback M. Cohen PS, Chan JP, Lipkunskaya M, Biedler JL, Seeger RC. (1994). (2004). Pterin-dependent tyrosine hydroxylase mRNA is not Expression of stem cell factor and c-kit in human neuroblastoma. expressed in human melanocytes or melanoma cells. Pigment Cell The Children’s Cancer Group. Blood 84: 3465–3472. Res 17: 346–351. Cully M, You H, Levine a J, Mak TW. (2006). Beyond PTEN Kenney AM, Widlund HR, Rowitch DH. (2004). Hedgehog and PI-3 mutations: the PI3K pathway as an integrator of multiple inputs kinase signaling converge on Nmyc1 to promote cell cycle during tumorigenesis. Nat Rev Cancer 6: 184–192. progression in cerebellar neuronal precursors. Development 131: Davidoff AM, Leary MA, Ng CY, Vanin EF. (2001). Gene therapy- 217–228. mediated expression by tumor cells of the angiogenesis inhibitor flk-1 Kim B, Van Golen CM, Feldman EL. (2004). Insulin-like growth results in inhibition of neuroblastoma growth in vivo. J Pediatr Surg factor-I signaling in human neuroblastoma cells. Oncogene 23: 36: 30–36. 130–141. Dong J, Peng J, Zhang H, Mondesire WH, Jian W, Mills GB et al. Kim ES, Serur A, Huang J, Manley CA, Mccrudden KW, Frischer JS (2005). Role of glycogen synthase kinase 3beta in rapamycin- et al. (2002). Potent VEGF blockade causes regression of coopted mediated cell cycle regulation and chemosensitivity. Cancer Res 65: vessels in a model of neuroblastoma. Proc Natl Acad Sci USA 99: 1961–1972. 11399–11404. Easton JB, Houghton PJ. (2006). mTOR and cancer therapy. Klos KS, Wyszomierski SL, Sun M, Tan M, Zhou X, Li P et al. (2006). Oncogene 25: 6436–6446. ErbB2 increases vascular endothelial growth factor protein synthesis Eggert A, Grotzer MA, Ikegaki N, Liu XG, Evans a E, Brodeur GM. via activation of mammalian target of rapamycin/p70S6 K leading (2002). Expression of the neurotrophin receptor TrkA down-regulates to increased angiogenesis and spontaneous of human expression and function of angiogenic stimulators in SH-SY5Y breast cancer cells. Cancer Res 66: 2028–2037. neuroblastoma cells. Cancer Res 62: 1802–1808. Kozma SC, Thomas G. (2002). Regulation of cell size in growth, Eggert A, Ikegaki N, Kwiatkowski J, Zhao H, Brodeur GM, development and human disease: PI3K, PKB and S6K. Bioessays Himelstein BP. (2000). High-level expression of angiogenic factors 24: 65–71. is associated with advanced tumor stage in human neuroblastomas. Lutz W, Stohr M, Schurmann J, Wenzel A, Lohr A, Schwab M. Clin Cancer Res 6: 1900–1908. (1996). Conditional expression of N-myc in human neuroblastoma Ferrara N, Alitalo K. (1999). Clinical applications of angiogenic cells increases expression of alpha-prothymosin and ornithine growth factors and their inhibitors. Nat Med 5: 1359–1364. decarboxylase and accelerates progression into S-phase early after Gera JF, Mellinghoff IK, Shi Y, Rettig MB, Tran C, Hsu JH et al. mitogenic stimulation of quiescent cells. Oncogene 13: 803–812. (2004). AKT activity determines sensitivity to mammalian target of Macdonald TJ, Brown KM, Lafleur B, Peterson K, Lawlor C, Chen Y rapamycin (mTOR) inhibitors by regulating cyclin D1 and c-myc et al. (2001). Expression profiling of medulloblastoma: PDGFRA expression. J Biol Chem 279: 2737–2746. and the RAS/MAPK pathway as therapeutic targets for metastatic Grimmer MR, Weiss WA. (2006). Childhood tumors of the nervous disease. Nat Genet 29: 143–152. system as disorders of normal development. Curr Opin Pediatr 18: Meitar D, Crawford SE, Rademaker AW, Cohn SL. (1996). Tumor 634–638. angiogenesis correlates with metastatic disease, N-myc amplifica- Hay N. (2005). The Akt-mTOR tango and its relevance to cancer. tion, and poor outcome in human neuroblastoma. J Clin Oncol 14: Cancer Cell 8: 179–183. 405–414.

Oncogene mTOR inhibitors in neuroblastoma therapy JI Johnsen et al 2922 Nelsen CJ, Rickheim DG, Tucker MM, Hansen LK, Albrecht JH. mediated mammalian target of rapamycin inhibition. Cancer Res 65: (2003). Evidence that cyclin D1 mediates both growth and 7052–7058. proliferation downstream of TOR in hepatocytes. J Biol Chem Uccini S, Mannarino O, Mcdowell HP, Pauser U, Vitali R, Natali PG 278: 3656–3663. et al. (2005). Clinical and molecular evidence for c-kit receptor as Opel D, Poremba C, Simon T, Debatin KM, Fulda S. (2007). a therapeutic target in neuroblastic tumors. Clin Cancer Res 11: Activation of Akt predicts poor outcome in neuroblastoma. Cancer 380–389. Res 67: 735–745. Vanhaesebroeck B, Leevers SJ, Ahmadi K, Timms J, Katso R, Driscoll PC Ponthan F, Johnsen JI, Klevenvall L, Castro J, Kogner P. (2003). The et al. (2001). Synthesis and function of 3-phosphorylated inositol lipids. synthetic retinoid RO 13-6307 induces neuroblastoma differentia- Annu Rev Biochem 70: 535–602. tion in vitro and inhibits neuroblastoma tumour growth in vivo. Vitali R, Cesi V, Nicotra MR, Mcdowell HP, Donfrancesco A, Int J Cancer 104: 418–424. Mannarino O et al. (2003). c-Kit is preferentially expressed in Ponthan F, Wickstrom M, Gleissman H, Fuskevag OM, Segerstrom L, MYCN-amplified neuroblastoma and its effect on cell proliferation Sveinbjornsson B et al. (2007). Celecoxib prevents neuroblastoma is inhibited in vitro by STI-571. Int J Cancer 106: 147–152. tumor development and potentiates the effect of chemotherapeutic Wang JY, Del Valle L, Gordon J, Rubini M, Romano G, Croul S et al. drugs in vitro and in vivo. Clin Cancer Res 13: 1036–1044. (2001). Activation of the IGF-IR system contributes to malignant Schwab M, Westermann F, Hero B, Berthold F. (2003). Neuroblas- growth of human and mouse medulloblastomas. Oncogene 20: toma: biology and molecular and chromosomal pathology. Lancet 3857–3868. Oncol 4: 472–480. Weiss WA, Aldape K, Mohapatra G, Feuerstein BG, Bishop JM. Segerstrom L, Fuchs D, Backman U, Holmquist K, Christofferson R, (1997). Targeted expression of MYCN causes neuroblastoma in Azarbayjani F. (2006). The anti-VEGF antibody bevacizumab transgenic mice. EMBO J 16: 2985–2995. potently reduces the growth rate of high-risk neuroblastoma Yaari S, Jacob-Hirsch J, Amariglio N, Haklai R, Rechavi G, Kloog Y. xenografts. Pediatr Res 60: 576–581. (2005). Disruption of cooperation between Ras and MycN in human Singleton JR, Randolph a E, Feldman EL. (1996). Insulin-like growth neuroblastoma cells promotes growth arrest. Clin Cancer Res 11: factor I receptor prevents apoptosis and enhances neuroblastoma 4321–4330. tumorigenesis. Cancer Res 56: 4522–4529. Yu Q, Ciemerych MA, Sicinski P. (2005). Ras and Myc can drive Sun SY, Rosenberg LM, Wang X, Zhou Z, Yue P, Fu H et al. (2005). oncogenic cell proliferation through individual D-cyclins. Oncogene Activation of Akt and eIF4E survival pathways by rapamycin- 24: 7114–7119.

Oncogene