- Email: [email protected]
AKT pathway in neuroblastoma and its regulation by thioredoxin 1
Human Pathology (2011) 42, 1727–1739
www.elsevier.com/locate/humpath
Original contribution
Activation of the phosphatidylinositol 3′-kinase/AKT pathway in neuroblastoma and its regulation by thioredoxin 1☆ Hervé Sartelet MD, PhD a,b,⁎, Anne-Laure Rougemont MD b , Monique Fabre MD c , Marine Castaing PhD d , Michel Duval MD, PhD e , Raouf Fetni PhD b , Stefan Michiels PhD e , Mona Beaunoyer MD f , Gilles Vassal MD, PhD a,g a
UPRES EA3535, University of Paris South, Institut Gustave Roussy, 94805 Villejuif, France Department of Pathology, Sainte Justine Hospital, University of Montreal, H3T1C5 Montreal, Quebec, Canada c Department of Pathology, University of Paris South 11, Hôpital Bicêtre, 94275 Le Kremlin-Bicêtre, France d Department of Biostatistics and Epidemiology, Institut Gustave Roussy, 94805 Villejuif, France e Department of Pediatric Oncology, Sainte Justine Hospital, Montreal, H3T1C5 Quebec, Canada f Department of Surgery, Sainte Justine Hospital, Montreal, Quebec, H3T1C5 Canada g Department of Pediatric Oncology, Institut Gustave Roussy, 94805 Villejuif, France b
Received 28 September 2010; revised 21 January 2011; accepted 28 January 2011
Keywords: AKT; Neuroblastoma; Tissue microarray; Targeted therapy; Thioredoxin
Summary Neuroblastoma is a malignant pediatric tumor with poor survival. The phosphatidylinositol 3′-kinase/AKT pathway is a crucial regulator of cellular processes including apoptosis. Thioredoxin 1, an inhibitor of tumor-suppressor phosphatase and tensin homolog, is overexpressed in many tumors. The objective of this study was to explore phosphatidylinositol 3′-kinase/AKT pathway activation and regulation by thioredoxin 1 to identify potential therapeutic targets. Immunohistochemical analysis was done on tissue microarrays from tumor samples of 101 patients, using antibodies against phosphatidylinositol 3′-kinase, AKT, activated AKT, phosphatase and tensin homolog, phosphorylated phosphatase and tensin homolog, thioredoxin 1, epidermal growth factor receptor, vascular endothelial growth factor and receptors (vascular endothelial growth factor 1 and vascular endothelial growth receptor 2), platelet-derived growth factor receptors, insulin-like growth factor 1 receptor, neurotrophic tyrosine kinase receptor type 2, phosphorylated 70-kd S6 protein kinase, 4E-binding protein 1, and phosphorylated mammalian target of rapamycin. Using 3 neuroblastoma cell lines, we investigated cell viability with AKT-specific inhibitors (LY294002, RAD001) and thioredoxin 1 alone or in combination. We found activated AKT and AKT expressed in 97% and 98%, respectively, of neuroblastomas, despite a high expression of phosphatase and tensin homolog correlated with thioredoxin 1. AKT expression was greater in metastatic than primary tumors. Insulin-like growth factor 1 receptor, tyrosine kinase receptor type 2, vascular endothelial growth receptor 1, and downstream phosphorylated 70-kd S6 protein kinase were correlated with activated AKT. LY294002 and RAD001 significantly reduced AKT activity and cell viability and induced a G1 cell cycle
☆ This research was funded by foundation Charles Bruneau (Montreal, Quebec, Canada) and Leucan (Montreal, Quebec, Canada). ⁎ Corresponding author. Department of Pathology, CHU Sainte Justine, Université de Montréal, 3175, Côte Sainte-Catherine Montréal (QC) H3T1C5 Canada. E-mail address: [email protected] (H. Sartelet).
0046-8177/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.01.019
1728
H. Sartelet et al. arrest. Thioredoxin 1 decreased cytotoxicity of AKT inhibitors and doxorubicin, up-regulated AKT activation, and induced cell growth. Thus, vascular endothelial growth receptor 1, tyrosine kinase receptor type 2, insulin-like growth factor 1 receptor, and thioredoxin 1 emerged as preferentially committed to phosphatidylinositol 3′-kinase/AKT pathway activation as observed in neuroblastoma. Thioredoxin 1 is a potential target for therapeutic intervention. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Neuroblastoma, a tumor of peripheral neural crest origin, is the most common extracranial solid tumor of childhood, with an incidence of 10.9 cases per million children [1]. Even with aggressive treatment, survival in children older than 1 year with advanced disease is poor, with only 34% surviving in the long term [2]. The activation of the serine/threonine protein kinase B pathway, also known as the AKT pathway, has emerged as a crucial regulator of cellular processes including apoptosis, proliferation, differentiation, and metabolism [3-5]. Phosphatase and tensin homolog (PTEN) is a tumor-suppressor protein that negatively regulates the phosphatidylinositol 3′kinase (PI3K)/AKT signaling pathway by dephosphorylating phosphatidylinositol 3 [6]. As previously reported in in vitro studies, the AKT pathway is present and activated in neuroblastoma cells [7-9]. Inhibition of the PI3K/AKT pathway in vitro decreased neuroblastoma tumor mass and oncogene N-myc protein expression while affecting neither the levels of N-myc messenger RNA nor N-myc amplification [10]. In neuroblastoma cell lines, the use of AKTspecific inhibitors [11-14] or of small interfering RNA targeting AKT [15] induced apoptotic cell death. Indeed, AKT activation in neuroblastoma samples was found to be associated with poor prognosis in vivo [16]. Because the AKT pathway has many therapeutic implications in neuroblastoma [17] and other cancers, several AKT inhibitors have been assessed at the preclinical level. The most frequently described is LY294002, which has demonstrated very selective total inhibition of PI3K activity [18]. The naturally occurring rotenoid deguelin is an AKT inhibitor that increases the sensitivity of cells to chemotherapeutic drugs, as shown in leukemia cells with an active PI3K/AKT signaling network [19]. At the clinical level, the mammalian target of rapamycin (mTOR) has emerged as an important therapeutic target as it induces phosphorylation of AKT. By inhibiting the AKT pathway, mTOR inhibitors are a promising therapeutic option in cancers and in pediatric malignancies in particular [20]. PTEN activity is regulated in 2 ways: (a) phosphorylation of PTEN decreases its phosphatase activity [21] and (b) intracytoplasmic binding of PTEN to proteins such as thioredoxin 1 (TRX-1) [22]. TRX-1, a small ubiquitous protein with multiple biologic functions, is overexpressed in many tumor cell lines, including neuroblastomas [23]. It is
present in various compartments of the cell [21], including the cytosol [24]. Essential for the first step of DNA synthesis, TRX-1 regulates the activity of proteins that control cell growth such as PTEN and AKT [22]; induction of TRX-1 promotes oncogenicity. Indeed, increased TRX-1 levels are seen in many human primary cancers such as colorectal [25]; and TRX-1–transfected cells are resistant to classical therapeutic drugs such as doxorubicin [26]. Newly developed TRX-1 inhibitors such as PMX464 (Pharminox 464) have been shown to decrease proliferation and survival of colorectal cancer cell lines [27]. In neuroblastoma, TRX-1 protects the cell against oxidative stress–induced apoptosis [23]. TRX-1 up-regulation has been described as a compensatory cell survival mechanism when the expression of antiapoptotic B-cell lymphoma 2 is blocked [28]. Finally, in neuroblastoma, TRX-1 stimulates neuroblast invasion by decreasing the expression of metalloproteinase inhibitors [29]. The aims of our study were (1) to quantify the activation of the AKT pathway in tissue samples from patients with neuroblastoma; (2) to explore the interrelationship between intrapathway proteins; and (3) to study the mechanisms of PTEN regulation, with a special focus on the importance of TRX-1. Mapping of protein signaling networks within tumor cells is important as these may prove useful in identifying the best therapeutic interventions (or combinations thereof) for targeting the AKT pathway.
2. Materials and methods 2.1. Patients and samples We obtained tumor samples from 101 patients with neuroblastic neoplasms treated and managed at the 2 centers in France: first in Hôpital Bicêtre (Le Kremlin-Bicêtre) and Institut Gustave Roussy (Villejuif) and second in Hôpital Américain (Reims). The samples were fixed in 10% neutralbuffered formalin. A tissue microarray was constructed using on average 4 tissue cores per sample with a 0.6-mm diameter. The cores were transferred into a recipient paraffin block using a tissue arrayer. Four tissue microarray blocks were constructed containing 101 primary tumors, 39 paired metastases (35 lymph nodes), and 56 paired control normal tissues. For Western blot analysis, we used 8 frozen samples obtained from patients with neuroblastoma treated and followed up at Sainte-Justine Hospital (Montreal, Canada).
Thioredoxin and AKT pathway in neuroblastoma Four were from infants younger than 1 year with stage 1 disease, and 4 were from children older than 1 year with stage 4 disease. Informed consent and assent were obtained from patients and/or parents.
2.2. Immunohistochemical study An immunohistochemical study was performed using 5-μm sections of the tissue microarray blocks. These sections were deparaffinized and incubated with the following antibodies for immunohistochemical staining (Fig. 1): AKT (mouse monoclonal, 1:50, B-1; Santa Cruz Biotechnology, Santa Cruz, CA), phosphorylated AKT (pAKT; at serine 473) (rabbit polyclonal, 1:100, S473-r; Santa Cruz Biotechnology), PI3K (phosphatidylinositol 3-kinase) (mouse monoclonal, 1:50, p85a[B9]; Santa Cruz Biotechnology), PTEN (phosphatase and tensin homolog) (mouse monoclonal, 1:100, A2B1; Santa Cruz Biotechnology), phosphorylated PTEN (pPTEN) (rabbit polyclonal, 1:200, Ser380; Santa Cruz Biotechnology), TRX-1 (rabbit polyclonal, 1:400, C63C6; Cell Signaling Technologies, Danvers, MA), epidermal growth factor receptor (EGFR) (mouse monoclonal, 1:20, E30; Dako, Glostrup, Denmark), human epidermal growth factor receptor 2 (HER2) (mouse monoclonal, 1:700; CB11, Novocastra Newcastle, UK), insulin-like growth factor 1 receptor (IGF1R) (rabbit polyclonal, 1:300, H-60; Santa Cruz Biotechnology), platelet-derived growth factor receptor α (rabbit polyclonal, 1:100, C-20; Santa Cruz Biotechnology) and platelet-derived growth factor receptor β (PDGFRβ) (rabbit polyclonal, 1:200, P-20; Santa Cruz Biotechnology), vascular endothelial growth factor (VEGF) (mouse monoclonal 1:500, C-1; Santa Cruz Biotechnology), VEGF receptor 1 (VEGFR1) (rabbit polyclonal, 1:200, C-17; Santa Cruz Biotechnology) and VEGF receptor 2 (VEGFR2) ( mouse monoclonal, 1:200, A-3; Santa Cruz Biotechnology), neurotrophic tyrosine kinase receptor type (TRKB) (rabbit polyclonal, 1:400, H-118; Santa Cruz Biotechnology), phosphorylated mTOR (p-mTOR [Ser 2448]) (rabbit monoclonal, 1:50, 49F9; Cell Signaling Technologies), eukaryotic translation initiation factor 4E-binding protein 1 (4EBP1) (rabbit polyclonal, 1:50, 53H11; Cell Signaling Technologies), and phosphorylated 70-kd S6 protein kinase (p-p70S6K [Ser 411]) ( mouse monoclonal, 1:100, A-6; Santa Cruz Biotechnology). As a negative control, the slides were incubated with normal rabbit IgG at the same concentration as the primary antibody. Samples were revealed with the LSAB II Kit (Dako, Glostrup, Denmark), according to manufacturer's instructions. Two investigators blinded for clinical data independently evaluated immunostaining under a light microscope at an original magnification of ×400. Immunostaining scores were established by a semiquantitative optical analysis of samples containing more than 10 neuroblasts, assessing the percentage of positive cells in each sample: 0, all cells negative; 1+, up to 25% positive tumor cells; 2+, 26% to 50% positive cells; 3+, 51% to 75% positive cells; and 4+, more than 75% positive cells. Interobserver agreement was calculated using the κ coefficient. Discordant
1729 cases were discussed by the 2 investigators, and a consensus was reached.
2.3. Cell lines We used 3 neuroblastoma cells lines: 2 non–N-mycamplified cell lines (SK-N-SH and SK-N-AS) purchased from American Type Culture Collection (Manassas, VA) and 1 N-myc–amplified cell line (NB 10) [30] from Saint Jude's Children's Research Hospital (Memphis, TN). Cells were cultured in Dulbecco Modified Eagle's Medium (GibcoBRL, Rockville, MD) supplemented with 10% fetal bovine serum (HyClone, Logan, UT) at 37°C in a humidified atmosphere consisting of 5% CO2 and 95% air. The culture medium was changed every 48 hours.
2.4. Western blot The 8 frozen patient tumor samples were used for Western blot analysis. A very small piece of tumor sample was crushed with a homogenizer, and the temperature was maintained at 4°C throughout. All samples were centrifuged at 10 000g for 10 minutes at 4°C. The supernatant fluid represented the total cell lysate. SK-N-SH cells were incubated with either LY294002 20 μmol/L (Calbiochem, Darmstadt, Germany), everolimus (RAD001) 10 μmol/L (Novartis, Basel, Switzerland), or human recombinant TRX-1 (hrTRX-1) 10 μmol/L (Calbiochem) for 3 hours at 37°C in a CO2 incubator. The medium was removed, and a cell lysis buffer was added for 15 minutes at 4°C. Fifteen micrograms of proteins from each sample were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. After transfer to a polyvinylidene fluoride membrane, the resultant was immunoblotted with antibodies against pAKT (1/500, B-1; Santa Cruz Biotechnology), TRX-1 (1/2000, C63C6; Cell Signaling Technologies), p-mTOR (1/500, rabbit, 49F9; Cell Signaling Technologies), or β-actin (positive control) and then incubated for 1 hour at room temperature. These were followed by incubation with donkey secondary antimouse or antirabbit antibody (horseradish peroxidase conjugated) (Jackson Immunoresearch Lab, West Grove, PA). Blots were visualized with enhanced chemiluminescence before exposing the membrane to photosensitive paper.
2.5. AKT kinase assay Active AKT was immunoprecipitated from 1 mg of clarified total cell lysate of SK-N-SH, SK-N-AS, or NB-10 cell lines, according to the manufacturer's protocol. Five micrograms of mouse monoclonal anti-AKT antibody, near the Pleckstrin homology-domain amino acid 107-122 (5G12; Santa Cruz Biotechnology) were used per 500 μg of cell lysate. After immunoprecipitation, equivalent amounts of
1730
H. Sartelet et al.
Fig. 1 Immunohistochemical study of AKT pathway in tissue microarray of neuroblastoma tumor cells. Immunohistochemical staining of 5-μm sections of microarray blocks, deparaffinized, treated with 1% H2O2, and submitted to antigen retrieval by microwave oven treatment for 15 minutes in 0.01 mmol/L citrate buffer at pH 6.0. Samples were then incubated with biotinylated immunoglobulin at room temperature for 30 minutes, followed by avidin-biotin-peroxidase complexes for 30 minutes. Three-amino-9-ethylcarbazole was used as the chromogen; and hematoxylin, as the nuclear counterstain. AKT pathway at original magnification ×100. Red arrow indicates significant correlation between 2 proteins (A); AKT pathway markers at high magnification: IGFR1, PDGFRβ, AKT, pAKT, p-p70S6K, p-mTOR are visualized at original magnification ×400; EGFR, VEGFR1, TRKB, TRX-1, PTEN, pPTEN, PI3K, and 4EBP1, at original magnification ×630. Abbreviations: AKT indicates serine/threonine protein kinase.
Thioredoxin and AKT pathway in neuroblastoma
1731
Fig. 1
eluate were used for the kinase assay with an enzyme-linked immunosorbent assay–based AKT activity assay using a biotinylated peptide substrate phosphorylated by AKT kinase (K-LISA AKT activity assay; Calbiochem). AKT
(continued).
activity was quantified by reading the absorbance at 450 nm, with a reference wavelength set at 540 nm. All mesurements were performed in triplicate, each with 3 determinations for each condition.
1732
H. Sartelet et al.
2.6. Cell proliferation assay
3. Results
Chemotherapy-induced cytotoxicity was determined by MTT (3-(4, 5-dimethylthiazol-2-y1)2, 5-diphenyltetrazolium) cell proliferation assay (Cell Titer 96; Fisher Scientific, Rockville, MD). For each of the 3 cell lines, we incubated cells for 24 hours with various concentrations of doxorubicin, a chemotherapeutic agent commonly used in the treatment of neuroblastoma (0.1-36 μmol/L); LY294002, an AKT-specific inhibitor (2.5-605 μmol/L; Calbiochem); deguelin, an AKT-specific inhibitor (1.5-252 μmol/L; Calbiochem); or everolimus (RAD001), a specific mTOR inhibitor (0.65-104 μmol/L; Novartis). Absorbency was measured at 570 nm. Assays were performed 3 times. The mean cell viability was compared with that of positive control cells receiving only medium.
3.1. Patient characteristics
2.7. Cell cycle analysis by flow cytometry Each of the cell lines, when almost 60% confluent, was incubated with dimethyl sulfoxide alone (control) or added to one of the following interventions: LY294002 20 μmol/L, everolimus (RAD001) 10 μmol/L, or hrTRX-1 10 μmol/L for 24 hours at 37°C in a CO2 incubator. Once attached the cells were trypsinized, 1 × 106 fixed cells were harvested by centrifugation and washed 3 times with phosphate-buffered saline. Cells were then resuspended in 0.5-mL fluorochrome solution containing 50 μg/mL propidium iodide, 0.1% sodium citrate, 0.1% Triton X-100, and 0.1 mg/mL ribonuclease A. After a 1-hour incubation at 4°C protected from light, the cells were analyzed on a Beckman Coulter (Brea, CA, USA) EPICS-XL flow cytometer. The mean value was determined from 3 independent experiments.
2.8. Statistical analysis We compared paired data using the Wilcoxon signed rank test. The univariate relationships between immunohistochemical expression in tumor tissues and clinical variables such as age, disease stage as per the International Neuroblastoma Staging System, and histologic type were investigated using a Wilcoxon test. Spearman correlation values (ρ) were used to compare the expression of proteins in the primary tumors. Event-free survival was computed from the time of surgery of the primary tumor (baseline) to the time of first event (local relapse, metastasis, or death) or last follow-up; overall survival was computed from the time of surgery to the time of death or last follow-up. Differences in survival between patients with a low versus high level of expression were assessed on a log-rank test. For the in vitro study, results were expressed as means ± SD of at least 3 independent experiments. The Student t test was used to determine statistical significance. Statistical analyses were performed with SAS software version 8.2 (SAS, Cary, NC). A P value b.05 indicated statistical significance.
Clinicopathologic characteristics of the 101 patients and associated tumors are detailed in Table 1. The median follow-up was 60 months (range, 1-174 months), with a median age at diagnosis of 30 months (range, newborn to 151 months).
3.2. Activation of the AKT pathway in neuroblastoma Interobserver agreement in immunostaining scores between the 2 pathologists was κ = 0.61 (95% confidence interval, 0.58-0.64; P b .0001), before consensus was reached. AKT and pAKT were expressed in 98% and 97% of tumors, respectively, with a median semiquantitative score of 2 Table 1
Patient characteristics at diagnosis
Patient characteristics Sex Male, n (%) Female, n (%) Age (mo) Median (range) b12 mo, n (%) 1-5 y, n (%) N5 y, n (%) Follow-up (mo) Median (range) Survival Alive at time of last follow-up, n (%) INSS stage 1, n (%) 2, n (%) 3, n (%) 4, n (%) 4S, n (%) N-myc oncogene analysis b10 copies, n (%) N10 copies, n (%) Unknown, n (%) Shimada histopathologic classification Favorable, n (%) Unfavorable, n (%) Children's Oncology Group Risk Classification Low, n (%) Intermediate, n (%) High, n (%) Unknown, n (%) Sample type Primary tumor, n (%) Paired metastasis, n (%) Paired control normal tissue, n (%)
(N = 101) 52 (51.4) 49 (48.6) 30 (0-151) 30 (29.7) 53 (52.4) 18 (17.8) 60 (1-174) 64 (63.3) 16 (15.8) 7 (6.9) 22 (21.8) 48 (47.5) 8 (7.9) 67 (77.9) 19 (22.1) 15 49 (48.5) 52 (51.5) 19 (20.2) 21 (22.3) 54 (57.4) 7 101 39 (38.6) 56 (55.4)
Abbreviation: INSS indicates International Neuroblastoma Staging System.
Thioredoxin and AKT pathway in neuroblastoma (Table 2, Fig. 1A and B). These data were confirmed by Western blot analysis with pAKT expression found in 7 (87.5%) of the 8 tumors studied (Fig. 2A). PI3K was expressed in 61% of tumors. Among the membranous tyrosine kinase receptors studied, only IGF1R, TRKB, and PDGFRβ showed an intense and very frequent expression (Table 2, Fig. 1A and B). Among the VEGF receptors, the ligands VEGF (58% of tumors) and VEGFR1 (63%) were moderately expressed (median score of 1 for both), as opposed to VEGFR2 (27%; median score 0). The EGF (epidermal growth factor) family of receptors was rarely expressed. Among the downstream proteins, p-p70S6K (98% of tumors; median score 3) and 4EBP1 (95% of tumors; median score 3) were highly and frequently expressed, as opposed to p-mTOR (Table 2, Fig. 1A and B). IGF1R was more highly expressed in metastases than in primary tumors (P = .01). There was a strong correlation between PI3K and pAKT expression (ρ = 0.54; P b .0001) and between AKT and pAKT expression (ρ = 0.46; P b .0001). Among the tyrosine kinase receptors capable of upstream activation of AKT, there was a significant positive correlation between pAKT and VEGFR1 (ρ = 0.25; P = .01), IGF1R (ρ = 0.24; P = .02), and TRKB (ρ = 0.27; P = .008). Moreover, VEGF was highly correlated with pAKT (ρ = 0.33; P = .0008). The correlation between VEGF and its receptors was significant for VEGFR1 (ρ = 0.27; P = .007) and VEGFR2 (ρ = 0.46; P b .0001). Among downstream proteins, the correlation was high between pAKT and p-p70S6K (ρ = 0.43; P b .0001). Three proteins were expressed in the nucleus: pPTEN, pAKT, and Table 2
Protein expression in primary neuroblastoma
Protein
Median score (range)
% of positive tumors
Staining location
EGFR HER2 VEGF VEGFR1 VEGFR2 IGF1R TRKB PDGFRα PDGFRβ AKT pAKT PI3K PTEN pPTEN TRX-1 p-p70S6K p-mTOR 4EBP1
0 0 1 1 0 2 2 1 2 2 2 1 2.5 0 1 3 1 3
19 2 58 63 27 96 94 76 92 98 97 61 92 63 76 98 51 95
M M C M and C M and C M and C M M M C C and N C C C and N C and N C C C
(0-3) (0-0) (0-4) (0-4) (0-4) (0-3.5) (0-4) (0-3) (0-4) (0-4) (0-4) (0-3) (0-4) (0-3) (0-4) (0-4) (0-3) (0-4)
Immunostaining scores were established by semiquantitative optical analysis of samples containing more than 10 neuroblasts, assessing the percentage of positive cells in each sample: 0, none, all cells negative; 1+, up to 25% of cells were positive; 2+, 26% to 50%; 3+, 51% to 75%; 4+, more than 75%. Abbreviations: M indicates membrane; C, cytoplasm; N, nucleus; PDGFRα, platelet-derived growth factor receptor.
1733
Fig. 2 Activation of the AKT pathway by TRX-1. A, Western blot analysis of frozen neuroblastoma samples. Four samples from stage 1 tumors of infants younger than 1 year (tumors 1-4) and 4 from stage 4 tumors of patients older than 1 year (tumors 5-8) were immunoblotted with antibodies against pAKT, TRX-1, or β-actin (positive control) for 1 hour at room temperature. B, Western blot analysis of a neuroblastoma cell line. Almost 60% confluent SK-N-SH cells were incubated with hrTRX-1 10 μmol/L, RAD001 10 μmol/L, or LY294002 20 μmol/L for 3 hours at 37°C in a CO2 incubator and immunoblotted with antibodies against pAKT, p-mTOR, and β-actin (positive control). C, Cell line AKT kinase assay. SK-N-SH, SK-N-AS, and NB10 cells from cultures, almost 60% confluent, were incubated with LY294002 20 μmol/L, RAD001 20 μmol/L, or hrTRX-1 10 μmol/L for 3 hours at 37°C in a CO2 incubator. To assess TRX-1 activity, cells were incubated with LY294002 or RAD001 in combination with TRX-1. Means of 3 separate experiments are shown as well as SDs (error bars). ⁎P values b.05.
TRX-1 (5%, 40%, 75% of cells, respectively; Fig. 1B). This nuclear expression was not correlated with patient outcome, cell proliferation, or mitosis-karyorrhexis index.
3.3. Relationship between protein expression and clinical covariates On immunohistochemical studies, no relationship was found between the expression of pAKT or p-mTOR in the
1734 tumors and clinical variables. On Western blot analysis, however, the expression of pAKT was demonstrably lower in stage 1 than 4 (Fig. 2A), with 2 low expressions and 1 moderate in stage 1 versus 2 high and 3 moderate expressions in stage 4. Similarly, AKT was more highly expressed in metastatic versus nonmetastatic stage tumors (P = .001). Survival analysis found no correlation between patient outcome and tumor AKT expression (P = .1), but event-free survival tended to be lower in patients with a high expression of AKT (P = .04). TRKB expression was higher in older patients (P = .004) and in metastatic stages (P b .001). Survival analysis found that patients whose tumor expressed more TRKB (above a median score of 2) had a significantly worse survival and lower event-free survival rate than those whose tumors showed a low expression of TRKB (P = .004 and P b .001, respectively). The N-myc
H. Sartelet et al. amplification status showed no correlation with any of the proteins studied.
3.4. Regulation of PTEN activity in neuroblastoma PTEN was only expressed in the cytoplasm of tumors (92%; median score, 2.5). Expression was significantly correlated with that of pAKT (ρ = 0.33; P = .0009). pPTEN, an inactivated form of PTEN, had a very low and inconsistent expression (63% of tumors; median score, 0) (Table 2, Fig. 1A and B), whrereas TRX-1 expression was moderate and frequent (76% of tumors; median score, 1.3). The highly significant correlations between TRX-1 and PTEN (ρ = 0.26; P b .0001) and between TRX-1 and pAKT (ρ = 0.17; P b .0001) were confirmed by Western
Fig. 3 The cytotoxic effect of chemotherapeutic agents modified by TRX-1. A, The effect of specific inhibitors alone on cell viability. The effect of chemotherapy alone in neuroblasts from 3 cell lines (SK-N-SH, SK-N-AS, and NB10) was assessed using a standard MTT assay. Varying concentrations of doxorubicin, a chemotherapeutic agent commonly used in the treatment of neuroblastoma; AKT inhibitors LY294002 and deguelin; and mTOR inhibitor everolimus (RAD001) were added to the culture medium. Data are expressed as a percentage of control culture conditions. Means of 3 separate experiments are shown as well as SDs (error bars); ⁎P b .01 for the 3 cell lines. B, The combined effect of TRX-1 and cytotoxic agents. SK-N-SH, SK-N-AS, and NB10 cells were incubated for 24 hours with cytotoxic agents in combination with hrTRX-1 at progressively increased concentrations (0-16 μmol/L) or with hrTRX-1 alone. Doxorubicin, LY294002, and RAD001 were used at the half maximal inhibitory concentrations as determined from the experiments in A: 4, 20, and 10 μmol/L, respectively. Cell viability was measured by the MTT assay. Data are expressed as a percentage of control culture conditions. Means of 3 separate experiments are shown as well as SDs (error bars). ⁎P b .05 for the 3 cell lines.
Thioredoxin and AKT pathway in neuroblastoma
1735
Fig. 3
blot, with correlation between the expression of pAKT and TRX-1 (or lack thereof) in 7 of the 8 tumors studied (Fig. 2A).
3.5. Inhibition of the AKT pathway in neuroblastoma cell lines Of the 3 AKT inhibitors tested, 2 (LY294002 and RAD001) significantly reduced the activation of AKT; this was correlated with decreased kinase activity of AKT in cell
(continued).
lines (P b .05), the activation of mTOR being also decreased (Fig. 2B and C). Treatment with LY294002 and RAD001 also induced a significant decrease of viable cells in all 3 cell lines studied, as did doxorubicine, a chemotherapeutic agent often prescribed in the treatment of neuroblastoma (P b .01; Fig. 3A). The half maximal inhibitory concentrations for LY294002 and RAD001 for 24 hours of incubation were determined as 20 and 10 μmol/L, respectively (Fig. 3A). In SK-N-SH, SK-N-AS, and NB10 cell lines, the percentage of cells in S phase (synthesis) was significantly reduced when treated with LY294002 (8.62% ± 0.3% and 6.44% ± 0.3%,
1736 respectively) and RAD001 (5.04% ± 1.1% and 9.56% ± 0.4%, respectively), as compared with control medium (12.11% ± 0.2% and 13.4% ± 1.2%) (P b .01; Fig. 4). These observations suggest that the AKT inhibitors induced a G1 cell cycle arrest. Treatment with up to 252 μmol/L of deguelin showed no significant change in cell viability in any of the cell lines studied (Fig. 3A).
H. Sartelet et al. significant (Fig. 3B). When cell cycle analysis was performed in SK-N-SH, SK-N-AS, and NB-10, a significantly higher number of cells were found to be in the mitotic (G2/M) phase when treated with TRX-1 as compared with those without TRX-1 (9.1% ± 0.4%, 11.2% ± 0.2%, and 10.7% ± 0.8% versus 3.5% ± 0.5%, 4.2% ± 0.2%, and 5.1% ± 0.3%; P b .01) (Fig. 4). These results indicated a cell growth-induction effect by TRX-1.
3.6. Activation of the AKT pathway by TRX-1 TRX-1 significantly up-regulated AKT activation in neuroblasts, as demonstrated by an in vitro kinase assay performed on total cellular extracts after exposure to 10 μmol/L hrTRX-1 (P b .05; Fig. 2C). These data were confirmed through Western blot, by the increased level of the activated form of AKT (pAKT) after exposure (Fig. 2B). The hrTRX-1 dampened the down-regulation of AKT activity by LY294002 and RAD001 (Fig. 2C). TRX-1 decreased the cytotoxicity of both AKT inhibitors as well as that of doxorubicin (Fig. 3B). When used alone, TRX-1 induced a mild increase in cell viability, which was not statistically
4. Discussion Although substantial progress has been made in the treatment for children with low- and intermediate-risk neuroblastoma, the cure rate for high-risk patients remains poor. To identify novel therapeutic targets, it is important to uncover pathways critical to neuroblastoma tumorigenesis. The AKT pathway is of particular interest because it is associated with several tyrosine kinase receptors currently targeted by numerous anticancer drugs [17].
Fig. 4 Cell cycle analysis. SK-N-SH, SK-N-AS, and NB-10 cells were incubated with dimethyl sulfoxide (control) or with LY294002 20 μmol/L or everolimus (RAD001) 10 μmol/L or hrTRX-1 10 μmol/L. Percentages of cells in gap (G0/G1), synthesis (S), and mitosis (G2/M) phases are presented in histograms. The experiment was repeated 3 times. Data are shown as mean ± SD.
Thioredoxin and AKT pathway in neuroblastoma Our study confirmed that the AKT pathway was activated in neuroblastoma but failed to demonstrate a correlation between this activation and prognostic factors, in contrast to a previous study [16]. This difference may be explained in part by the different methodologies used, such as the number of core biopsies per tumor, doublecontrol analysis by independent pathologists, quantification of positive cells, and statistical design addressing the issue of clinical correlations. Nonetheless, in our study, the level of AKT protein expression was correlated with a poorer outcome, where event-free survival was significantly lower in patients displaying a high level of AKT. A significant correlation was observed between PI3K, an AKT activator, and pAKT, the activated form of AKT. Moreover, downstream proteins (p-p70S6K and 4EBP1) were present in more than 92% of primary tumors and metastases, a high expression confirming AKT pathway activation. Our data suggested that, among the tyrosine kinase receptors, TRKB, PDGFRβ, and IGF1R might represent targets of interest for specific therapeutic intervention. We also found that VEGF and VEGFR1 had moderate but frequent expression, the significant correlation between the molecule and its receptor strongly suggesting paracrine and autocrine activation [31]. With respect to the EGF receptor family, our results indicated that HER2 and EGFR expressions were very rare in neuroblastoma and showed no correlation with clinical findings, in concordance with a previous study [32] but contrary to others [33]. Of the 3 AKT inhibitors tested, only LY294002 and RAD001 (not deguelin) significantly decreased neuroblast survival and induced a G1 cell cycle arrest. RAD001 is a specific mTOR inhibitor; it likely blocks AKT activation by inhibiting the formation of mTOR complex 2; mTOR complex 2 is known to phosphorylate and activate AKT [34]. In neuroblastoma and acute myeloid leukemia, RAD001 also decreased cell survival. TRX-1, which activates the AKT pathway, partially reversed the action of RAD001, LY294002, and doxorubicin. Several studies have demonstrated that chemosensitivity to doxorubicin was regulated by the AKT pathway [35]. PTEN is a tumor-suppressor protein that negatively regulates the PI3K/AKT signaling pathway by dephosphorylating phosphatidylinositol 3′-kinase [6]. Although found in many malignancies [36], mutations in the PTEN gene are rare in neuroblastoma and may be responsible for malignant progression in only a limited percentage of cases [37]. In many cancers, the presence of molecular alterations of PTEN is often not significantly correlated with PTEN expression, as evidenced from immunohistochemical assays [38]. In our study, the monoclonal antibody assay for PTEN demonstrated only cytoplasmic staining and never nuclear expression. Despite an expression of PTEN in 92% of paired primary neuroblastomas, it is worth noting that pAKT and pp70S6K were still expressed in 97% and 98% of tumors, respectively, demonstrating continued activation of the AKT
1737 pathway. Further to a previous report of a positive correlation between the expression of PTEN and that of pAKT [38], we investigated pPTEN, which is the inactivated form of PTEN (phosphorylation inactivates PTEN by decreasing its phosphatase activity [21]), and TRX-1, a protein that inhibits dephosphorylation of phosphatidylinositol 3′-kinase by PTEN [21]. We observed an inconsistent presence of pPTEN and at low levels thereof. This finding in itself, therefore, cannot explain the high levels of PTEN expression in neuroblastoma without inactivation of the AKT pathway. Thioredoxin (TRX) is a key molecule for redox regulation. TRX-transgenic mice are more resistant to infection, inflammation, and ischemic diseases and survive longer than control mice [39]. TRX is an important regulator of the cell cycle in the G1 phase via cyclin D1 transcription and the ERK/AP-1 (extracellular signal-regulated kinase/ activator protein-1) signaling pathways [40]. However, TRX-1 was found to bind to the catalytic site of PTEN and to its C2 lipid membrane-binding domain [22]. Indeed, previous studies had suggested that the increased levels of TRX observed in human tumors could lead to a functional inhibition of PTEN tumor-suppressor activity [22]. In cancer cells, TRX-1 overexpression has been associated with a biologically aggressive cancer phenotype [25] and resistance to chemotherapeutic agents such as doxorubicin and cysplatin, drugs currently used in the treatment of neuroblastoma [26]. TRX was highly expressed in several neuroblastoma cell lines as well [23,28,29]. Our study was the first to show that TRX-1 was expressed in a large series of neuroblastomas from patients and that its expression was correlated with both PTEN and pAKT expressions. Hence, AKT activation despite a high level of PTEN was associated with the expression of TRX-1 in neuroblastoma. In in vitro [41,42] and in vivo studies [43], human recombinant TRX-1 enhanced cellular resistance to chemotherapy [41] and prolonged survival of cancer cells [42]. In this study, we demonstrated that hrTRX-1 induced AKT activation in neuroblastoma cell lines. Moreover, it partially inhibited the action of several chemotherapeutic agents, including AKT inhibitors; increased cell viability; and induced cell growth. Together, these data strongly suggest that specific inhibitors of TRX-1 alone or in combination with classical chemotherapy could be beneficial in the treatment of neuroblastoma. PX-12 (1-methylpropyl 2-imidazolyl disulfide) is a specific TRX-1 inhibitor that shows both excellent in vitro and promising in vivo antitumor activity [44]. A detailed knowledge of tumor cell genomics is essential in establishing the risk classifications in neuroblastomas [45]. We know that DNA copy number gain represents only one of the many mechanisms that can cause protein overexpression. The evaluation of the intensity of protein expression in tumor samples by immunohistochemistry, including tissue microarray and Western blot analysis, as was done in our study, represents a very efficient means of detecting new therapeutic targets in neuroblastoma and of providing patient follow-up.
1738 Our study confirmed that the AKT pathway was activated in primary and metastatic neuroblastomas and demonstrated a correlation between this AKT activation and the presence of TRKB, IGF1R, VEGFR1, and, in particular, TRX-1. TRX-1 seems to be a key player. These data suggest the feasibility of therapeutic intervention on the AKT pathway in neuroblastoma through a combination of targeted therapies.
Acknowledgment The authors thank Danielle Buch, our medical editor, for editing and revising the manuscript.
References [1] Spix C, Pastore G, Sankila R, et al. Neuroblastoma incidence and survival in European children (1978-1997): report from the Automated Childhood Cancer Information System project. Eur J Cancer 2006;42: 2081-91. [2] Escobar MA, Grosfeld JL, Powell RL, et al. Long-term outcomes in patients with stage IV neuroblastoma. J Pediatr Surg 2006;41:377-81. [3] Brazil DP, Park J, Hemmings BA. PKB binding proteins. Getting in on the Akt. Cell 2002;111:293-303. [4] Kandel ES, Hay N. The regulation and activities of the multifunctional serine/threonine kinase Akt/PKB. Exp Cell Res 1999;253:210-29. [5] Nicholson KM, Anderson NG. The protein kinase B/Akt signalling pathway in human malignancy. Cell Signalling 2002;14:381-95. [6] Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate. J Biol Chem 1998;273:13375-8. [7] Fang CX, Yang X, Sreejayan N, et al. Acetaldehyde promotes rapamycin-dependent activation of p70(S6K) and glucose uptake despite inhibition of Akt and mTOR in dopaminergic SH-SY5Y human neuroblastoma cells. Exp Neurol 2007;203:196-204. [8] Fleming JM, Desury G, Polanco TA, et al. IGF-I and epidermal growth factor receptors recruit distinct upstream signalling molecules to enhance AKT activation in mammary epithelial cells. Endocrinology 2006;147:6027-35. [9] Kyrylenko S, Roschier M, Korhonen P, et al. Regulation of PTEN expression in neuronal apoptosis. Brain Res Mol Brain Res 1999;73: 198-202. [10] Chesler L, Schlieve C, Goldenberg DD, et al. Inhibition of phosphatidylinositol 3-kinase destabilizes Mycn protein and blocks malignant progression in neuroblastoma. Cancer Res 2006;66: 8139-46. [11] Efron PA, Chen MK, Iyengar M, et al. Differential response of neuroblastoma cells to TRAIL is independent of PI3K/AKT. J Pediatr Surg 2006;41:1072-80. [12] Kim S, Kang J, Qiao J, et al. Phosphatidylinositol 3-kinase inhibition down-regulates survivin and facilitates TRAIL-mediated apoptosis in neuroblastomas. J Pediatr Surg 2004;39:516-21. [13] Ruffels J, Griffin M, Dickenson JM. Activation of ERK1/2, JNK and PKB by hydrogen peroxide in human SH-SY5Y neuroblastoma cells: role of ERK1/2 in H2O2-induced cell death. Eur J Pharmacol 2004;483:163-73. [14] Shankar SL, Krupski M, Parashar B, et al. UCN-01 alters phosphorylation of Akt and GSK3beta and induces apoptosis in six independent human neuroblastoma cell lines. J Neurochem 2004;90: 702-11. [15] Liu XH, Yu EZ, Li YY, et al. RNA interference targeting Akt promotes apoptosis in hypoxia-exposed human neuroblastoma cells. Brain Res 2006;1070:24-30.
H. Sartelet et al. [16] Opel D, Poremba C, Simon T, et al. Activation of Akt predicts poor outcome in neuroblastoma. Cancer Res 2007;67:735-74. [17] Sartelet H, Oligny LL, Vassal G. AKT pathway in neuroblastoma and its therapeutic implication. Expert Rev Anticancer Ther 2008;8:757-69. [18] Vlahos CJ, Matter WF, Hui KY, et al. A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002). J Biol Chem 1994;269:5241-8. [19] Bortul R, Tazzari PL, Billi AM, et al. Deguelin, A PI3K/AKT inhibitor, enhances chemosensitivity of leukaemia cells with an active PI3K/AKT pathway. Br J Haematol 2005;129:677-86. [20] Witzig TE, Kaufmann SH. Inhibition of the phosphatidylinositol 3kinase/mammalian target of rapamycin pathway in hematologic malignancies. Curr Treat Options Oncol 2006;7:285-94. [21] Gericke A, Munson M, Ross AH. Regulation of the PTEN phosphatase. Gene 2006;374:1-9. [22] Meuillet EJ, Mahadevan D, Berggren M, et al. Thioredoxin-1 binds to the C2 domain of PTEN inhibiting PTEN's lipid phosphatase activity and membrane binding: a mechanism for the functional loss of PTEN's tumor suppressor activity. Arch Biochem Biophys 2004;429: 123-33. [23] Andoh T, Chock PB, Chiueh CC. The roles of thioredoxin in protection against oxidative stress–induced apoptosis in SH-SY5Y cells. J Biol Chem 2002;277:9655-60. [24] Nakamura H, Nakamura K, Yodoi J. Redox regulation of cellular activation. Annu Rev Immunol 1997;15:351-69. [25] Noike T, Miwa S, Soeda J, et al. Increased expression of thioredoxin-1, vascular endothelial growth factor, and redox factor-1 is associated with poor prognosis in patients with liver metastasis from colorectal cancer. HUM PATHOL 2008;39:201-8. [26] Yokomizo A, Ono M, Nanri H, et al. Cellular levels of thioredoxin associated with drug sensitivity to cisplatin, mitomycin C, doxorubicin, and etoposide. Cancer Res 1995;55:4293-6. [27] Mukherjee A, Huber K, Evans H, et al. A cellular and molecular investigation of the action of PMX464, a putative thioredoxin inhibitor, in normal and colorectal cancer cell lines. Br J Pharmacol 2007;151:1167-75. [28] Li Y, Lu Z, Chen F, et al. Antisense bcl-2 transfection up-regulates anti-apoptotic and anti-oxidant thioredoxin in neuroblastoma cells. J Neurooncol 2005;72:17-23. [29] Farina AR, Tacconelli A, Cappabianca L, et al. Thioredoxin alters the matrix metalloproteinase/tissue inhibitors of metalloproteinase balance and stimulates human SK-N-SH neuroblastoma cell invasion. Eur J Biochem 2001;268:405-13. [30] Shapiro DN, Valentine MB, Rowe ST, et al. Detection of N-myc gene amplification by fluorescence in situ hybridization. Diagnostic utility for neuroblastoma. Am J Pathol 1993;142:1339-46. [31] Langer I, Vertongen P, Perret J, et al. Expression of vascular endothelial growth factor (VEGF) and VEGF receptors in human neuroblastomas. Med Pediatr Oncol 2000;34:386-93. [32] Langer I, Vertongen P, Perret J, et al. Expression of HER2/neu is uncommon in human neuroblastic tumors and is unrelated to tumor progression. Cancer Immunol Immunother 2003;52:116-20. [33] Zeng Z, Sarbassov dos D, Samudio IJ, et al. Rapamycin derivatives reduce mTORC2 signaling and inhibit AKT activation in AML. Blood 2007;109:3509-12. [34] Ho R, Minturn JE, Hishiki T, et al. Proliferation of human neuroblastomas mediated by the epidermal growth factor receptor. Cancer Res 2005;65:9868-75. [35] Gagnon V, Van Themsche C, Turner S, et al. Akt and XIAP regulate the sensitivity of human uterine cancer cells to cisplatin, doxorubicin and taxol. Apoptosis 2008;13:259-71. [36] Athanassiadou P, Athanassiades P, Grapsa D. The prognostic value of PTEN, p53, and beta-catenin in endometrial carcinoma: a prospective immunocytochemical study. Int J Gynecol Cancer 2007;17:697-704. [37] Moritake H, Horii Y, Kuroda H, et al. Analysis of PTEN/MMAC1 alteration in neuroblastoma. Cancer Genet Cytogenet 2001;125: 151-5.
Thioredoxin and AKT pathway in neuroblastoma [38] Pallares J, Bussaglia E, Martínez-Guitarte JL, et al. Immunohistochemical analysis of PTEN in endometrial carcinoma: a tissue microarray study with a comparison of four commercial antibodies in correlation with molecular abnormalities. Mod Pathol 2005;18:719-27. [39] Mitsui A, Hamuro J, Nakamura H, et al. Overexpression of human thioredoxin in transgenic mice controls oxidative stress and life span. Antioxid Redox Signal 2002;4:693-6. [40] Mochizuki M, Kwon YW, Yodoi J, et al. Thioredoxin regulates cell cycle via the ERK1/2-cyclin D1 pathway. Antioxid Redox Signal 2009;11:2957-71. [41] Ueda S, Nakamura T, Yamada A, et al. Recombinant human thioredoxin suppresses lipopolysaccharide-induced bronchoalveolar neutrophil infiltration in rat. Life Sci 2006;79:1170-7.
1739 [42] Sasada T, Nakamura H, Ueda S, et al. Secretion of thioredoxin enhances cellular resistance to cis-diamminedichloroplatinum (II). Antioxid Redox Signal 2000;2:695-705. [43] Nilsson J, Söderberg O, Nilsson K, et al. Thioredoxin prolongs survival of B-type chronic lymphocytic leukemia cells. Blood 2000; 95:1420-6. [44] Kirkpatrick DL, Kuperus M, Dowdeswell M, et al. Mechanisms of inhibition of the thioredoxin growth factor system by antitumor 2-imidazolyl disulfides. Biochem Pharmacol 1998;55:987-94. [45] Ambros PF, Ambros IM, Brodeur GM, et al. International consensus for neuroblastoma molecular diagnostics: report from the International Neuroblastoma Risk Group (INRG) Biology Committee. Br J Cancer 2009;100:1471-82.
www.elsevier.com/locate/humpath
Original contribution
Activation of the phosphatidylinositol 3′-kinase/AKT pathway in neuroblastoma and its regulation by thioredoxin 1☆ Hervé Sartelet MD, PhD a,b,⁎, Anne-Laure Rougemont MD b , Monique Fabre MD c , Marine Castaing PhD d , Michel Duval MD, PhD e , Raouf Fetni PhD b , Stefan Michiels PhD e , Mona Beaunoyer MD f , Gilles Vassal MD, PhD a,g a
UPRES EA3535, University of Paris South, Institut Gustave Roussy, 94805 Villejuif, France Department of Pathology, Sainte Justine Hospital, University of Montreal, H3T1C5 Montreal, Quebec, Canada c Department of Pathology, University of Paris South 11, Hôpital Bicêtre, 94275 Le Kremlin-Bicêtre, France d Department of Biostatistics and Epidemiology, Institut Gustave Roussy, 94805 Villejuif, France e Department of Pediatric Oncology, Sainte Justine Hospital, Montreal, H3T1C5 Quebec, Canada f Department of Surgery, Sainte Justine Hospital, Montreal, Quebec, H3T1C5 Canada g Department of Pediatric Oncology, Institut Gustave Roussy, 94805 Villejuif, France b
Received 28 September 2010; revised 21 January 2011; accepted 28 January 2011
Keywords: AKT; Neuroblastoma; Tissue microarray; Targeted therapy; Thioredoxin
Summary Neuroblastoma is a malignant pediatric tumor with poor survival. The phosphatidylinositol 3′-kinase/AKT pathway is a crucial regulator of cellular processes including apoptosis. Thioredoxin 1, an inhibitor of tumor-suppressor phosphatase and tensin homolog, is overexpressed in many tumors. The objective of this study was to explore phosphatidylinositol 3′-kinase/AKT pathway activation and regulation by thioredoxin 1 to identify potential therapeutic targets. Immunohistochemical analysis was done on tissue microarrays from tumor samples of 101 patients, using antibodies against phosphatidylinositol 3′-kinase, AKT, activated AKT, phosphatase and tensin homolog, phosphorylated phosphatase and tensin homolog, thioredoxin 1, epidermal growth factor receptor, vascular endothelial growth factor and receptors (vascular endothelial growth factor 1 and vascular endothelial growth receptor 2), platelet-derived growth factor receptors, insulin-like growth factor 1 receptor, neurotrophic tyrosine kinase receptor type 2, phosphorylated 70-kd S6 protein kinase, 4E-binding protein 1, and phosphorylated mammalian target of rapamycin. Using 3 neuroblastoma cell lines, we investigated cell viability with AKT-specific inhibitors (LY294002, RAD001) and thioredoxin 1 alone or in combination. We found activated AKT and AKT expressed in 97% and 98%, respectively, of neuroblastomas, despite a high expression of phosphatase and tensin homolog correlated with thioredoxin 1. AKT expression was greater in metastatic than primary tumors. Insulin-like growth factor 1 receptor, tyrosine kinase receptor type 2, vascular endothelial growth receptor 1, and downstream phosphorylated 70-kd S6 protein kinase were correlated with activated AKT. LY294002 and RAD001 significantly reduced AKT activity and cell viability and induced a G1 cell cycle
☆ This research was funded by foundation Charles Bruneau (Montreal, Quebec, Canada) and Leucan (Montreal, Quebec, Canada). ⁎ Corresponding author. Department of Pathology, CHU Sainte Justine, Université de Montréal, 3175, Côte Sainte-Catherine Montréal (QC) H3T1C5 Canada. E-mail address: [email protected] (H. Sartelet).
0046-8177/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.01.019
1728
H. Sartelet et al. arrest. Thioredoxin 1 decreased cytotoxicity of AKT inhibitors and doxorubicin, up-regulated AKT activation, and induced cell growth. Thus, vascular endothelial growth receptor 1, tyrosine kinase receptor type 2, insulin-like growth factor 1 receptor, and thioredoxin 1 emerged as preferentially committed to phosphatidylinositol 3′-kinase/AKT pathway activation as observed in neuroblastoma. Thioredoxin 1 is a potential target for therapeutic intervention. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Neuroblastoma, a tumor of peripheral neural crest origin, is the most common extracranial solid tumor of childhood, with an incidence of 10.9 cases per million children [1]. Even with aggressive treatment, survival in children older than 1 year with advanced disease is poor, with only 34% surviving in the long term [2]. The activation of the serine/threonine protein kinase B pathway, also known as the AKT pathway, has emerged as a crucial regulator of cellular processes including apoptosis, proliferation, differentiation, and metabolism [3-5]. Phosphatase and tensin homolog (PTEN) is a tumor-suppressor protein that negatively regulates the phosphatidylinositol 3′kinase (PI3K)/AKT signaling pathway by dephosphorylating phosphatidylinositol 3 [6]. As previously reported in in vitro studies, the AKT pathway is present and activated in neuroblastoma cells [7-9]. Inhibition of the PI3K/AKT pathway in vitro decreased neuroblastoma tumor mass and oncogene N-myc protein expression while affecting neither the levels of N-myc messenger RNA nor N-myc amplification [10]. In neuroblastoma cell lines, the use of AKTspecific inhibitors [11-14] or of small interfering RNA targeting AKT [15] induced apoptotic cell death. Indeed, AKT activation in neuroblastoma samples was found to be associated with poor prognosis in vivo [16]. Because the AKT pathway has many therapeutic implications in neuroblastoma [17] and other cancers, several AKT inhibitors have been assessed at the preclinical level. The most frequently described is LY294002, which has demonstrated very selective total inhibition of PI3K activity [18]. The naturally occurring rotenoid deguelin is an AKT inhibitor that increases the sensitivity of cells to chemotherapeutic drugs, as shown in leukemia cells with an active PI3K/AKT signaling network [19]. At the clinical level, the mammalian target of rapamycin (mTOR) has emerged as an important therapeutic target as it induces phosphorylation of AKT. By inhibiting the AKT pathway, mTOR inhibitors are a promising therapeutic option in cancers and in pediatric malignancies in particular [20]. PTEN activity is regulated in 2 ways: (a) phosphorylation of PTEN decreases its phosphatase activity [21] and (b) intracytoplasmic binding of PTEN to proteins such as thioredoxin 1 (TRX-1) [22]. TRX-1, a small ubiquitous protein with multiple biologic functions, is overexpressed in many tumor cell lines, including neuroblastomas [23]. It is
present in various compartments of the cell [21], including the cytosol [24]. Essential for the first step of DNA synthesis, TRX-1 regulates the activity of proteins that control cell growth such as PTEN and AKT [22]; induction of TRX-1 promotes oncogenicity. Indeed, increased TRX-1 levels are seen in many human primary cancers such as colorectal [25]; and TRX-1–transfected cells are resistant to classical therapeutic drugs such as doxorubicin [26]. Newly developed TRX-1 inhibitors such as PMX464 (Pharminox 464) have been shown to decrease proliferation and survival of colorectal cancer cell lines [27]. In neuroblastoma, TRX-1 protects the cell against oxidative stress–induced apoptosis [23]. TRX-1 up-regulation has been described as a compensatory cell survival mechanism when the expression of antiapoptotic B-cell lymphoma 2 is blocked [28]. Finally, in neuroblastoma, TRX-1 stimulates neuroblast invasion by decreasing the expression of metalloproteinase inhibitors [29]. The aims of our study were (1) to quantify the activation of the AKT pathway in tissue samples from patients with neuroblastoma; (2) to explore the interrelationship between intrapathway proteins; and (3) to study the mechanisms of PTEN regulation, with a special focus on the importance of TRX-1. Mapping of protein signaling networks within tumor cells is important as these may prove useful in identifying the best therapeutic interventions (or combinations thereof) for targeting the AKT pathway.
2. Materials and methods 2.1. Patients and samples We obtained tumor samples from 101 patients with neuroblastic neoplasms treated and managed at the 2 centers in France: first in Hôpital Bicêtre (Le Kremlin-Bicêtre) and Institut Gustave Roussy (Villejuif) and second in Hôpital Américain (Reims). The samples were fixed in 10% neutralbuffered formalin. A tissue microarray was constructed using on average 4 tissue cores per sample with a 0.6-mm diameter. The cores were transferred into a recipient paraffin block using a tissue arrayer. Four tissue microarray blocks were constructed containing 101 primary tumors, 39 paired metastases (35 lymph nodes), and 56 paired control normal tissues. For Western blot analysis, we used 8 frozen samples obtained from patients with neuroblastoma treated and followed up at Sainte-Justine Hospital (Montreal, Canada).
Thioredoxin and AKT pathway in neuroblastoma Four were from infants younger than 1 year with stage 1 disease, and 4 were from children older than 1 year with stage 4 disease. Informed consent and assent were obtained from patients and/or parents.
2.2. Immunohistochemical study An immunohistochemical study was performed using 5-μm sections of the tissue microarray blocks. These sections were deparaffinized and incubated with the following antibodies for immunohistochemical staining (Fig. 1): AKT (mouse monoclonal, 1:50, B-1; Santa Cruz Biotechnology, Santa Cruz, CA), phosphorylated AKT (pAKT; at serine 473) (rabbit polyclonal, 1:100, S473-r; Santa Cruz Biotechnology), PI3K (phosphatidylinositol 3-kinase) (mouse monoclonal, 1:50, p85a[B9]; Santa Cruz Biotechnology), PTEN (phosphatase and tensin homolog) (mouse monoclonal, 1:100, A2B1; Santa Cruz Biotechnology), phosphorylated PTEN (pPTEN) (rabbit polyclonal, 1:200, Ser380; Santa Cruz Biotechnology), TRX-1 (rabbit polyclonal, 1:400, C63C6; Cell Signaling Technologies, Danvers, MA), epidermal growth factor receptor (EGFR) (mouse monoclonal, 1:20, E30; Dako, Glostrup, Denmark), human epidermal growth factor receptor 2 (HER2) (mouse monoclonal, 1:700; CB11, Novocastra Newcastle, UK), insulin-like growth factor 1 receptor (IGF1R) (rabbit polyclonal, 1:300, H-60; Santa Cruz Biotechnology), platelet-derived growth factor receptor α (rabbit polyclonal, 1:100, C-20; Santa Cruz Biotechnology) and platelet-derived growth factor receptor β (PDGFRβ) (rabbit polyclonal, 1:200, P-20; Santa Cruz Biotechnology), vascular endothelial growth factor (VEGF) (mouse monoclonal 1:500, C-1; Santa Cruz Biotechnology), VEGF receptor 1 (VEGFR1) (rabbit polyclonal, 1:200, C-17; Santa Cruz Biotechnology) and VEGF receptor 2 (VEGFR2) ( mouse monoclonal, 1:200, A-3; Santa Cruz Biotechnology), neurotrophic tyrosine kinase receptor type (TRKB) (rabbit polyclonal, 1:400, H-118; Santa Cruz Biotechnology), phosphorylated mTOR (p-mTOR [Ser 2448]) (rabbit monoclonal, 1:50, 49F9; Cell Signaling Technologies), eukaryotic translation initiation factor 4E-binding protein 1 (4EBP1) (rabbit polyclonal, 1:50, 53H11; Cell Signaling Technologies), and phosphorylated 70-kd S6 protein kinase (p-p70S6K [Ser 411]) ( mouse monoclonal, 1:100, A-6; Santa Cruz Biotechnology). As a negative control, the slides were incubated with normal rabbit IgG at the same concentration as the primary antibody. Samples were revealed with the LSAB II Kit (Dako, Glostrup, Denmark), according to manufacturer's instructions. Two investigators blinded for clinical data independently evaluated immunostaining under a light microscope at an original magnification of ×400. Immunostaining scores were established by a semiquantitative optical analysis of samples containing more than 10 neuroblasts, assessing the percentage of positive cells in each sample: 0, all cells negative; 1+, up to 25% positive tumor cells; 2+, 26% to 50% positive cells; 3+, 51% to 75% positive cells; and 4+, more than 75% positive cells. Interobserver agreement was calculated using the κ coefficient. Discordant
1729 cases were discussed by the 2 investigators, and a consensus was reached.
2.3. Cell lines We used 3 neuroblastoma cells lines: 2 non–N-mycamplified cell lines (SK-N-SH and SK-N-AS) purchased from American Type Culture Collection (Manassas, VA) and 1 N-myc–amplified cell line (NB 10) [30] from Saint Jude's Children's Research Hospital (Memphis, TN). Cells were cultured in Dulbecco Modified Eagle's Medium (GibcoBRL, Rockville, MD) supplemented with 10% fetal bovine serum (HyClone, Logan, UT) at 37°C in a humidified atmosphere consisting of 5% CO2 and 95% air. The culture medium was changed every 48 hours.
2.4. Western blot The 8 frozen patient tumor samples were used for Western blot analysis. A very small piece of tumor sample was crushed with a homogenizer, and the temperature was maintained at 4°C throughout. All samples were centrifuged at 10 000g for 10 minutes at 4°C. The supernatant fluid represented the total cell lysate. SK-N-SH cells were incubated with either LY294002 20 μmol/L (Calbiochem, Darmstadt, Germany), everolimus (RAD001) 10 μmol/L (Novartis, Basel, Switzerland), or human recombinant TRX-1 (hrTRX-1) 10 μmol/L (Calbiochem) for 3 hours at 37°C in a CO2 incubator. The medium was removed, and a cell lysis buffer was added for 15 minutes at 4°C. Fifteen micrograms of proteins from each sample were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. After transfer to a polyvinylidene fluoride membrane, the resultant was immunoblotted with antibodies against pAKT (1/500, B-1; Santa Cruz Biotechnology), TRX-1 (1/2000, C63C6; Cell Signaling Technologies), p-mTOR (1/500, rabbit, 49F9; Cell Signaling Technologies), or β-actin (positive control) and then incubated for 1 hour at room temperature. These were followed by incubation with donkey secondary antimouse or antirabbit antibody (horseradish peroxidase conjugated) (Jackson Immunoresearch Lab, West Grove, PA). Blots were visualized with enhanced chemiluminescence before exposing the membrane to photosensitive paper.
2.5. AKT kinase assay Active AKT was immunoprecipitated from 1 mg of clarified total cell lysate of SK-N-SH, SK-N-AS, or NB-10 cell lines, according to the manufacturer's protocol. Five micrograms of mouse monoclonal anti-AKT antibody, near the Pleckstrin homology-domain amino acid 107-122 (5G12; Santa Cruz Biotechnology) were used per 500 μg of cell lysate. After immunoprecipitation, equivalent amounts of
1730
H. Sartelet et al.
Fig. 1 Immunohistochemical study of AKT pathway in tissue microarray of neuroblastoma tumor cells. Immunohistochemical staining of 5-μm sections of microarray blocks, deparaffinized, treated with 1% H2O2, and submitted to antigen retrieval by microwave oven treatment for 15 minutes in 0.01 mmol/L citrate buffer at pH 6.0. Samples were then incubated with biotinylated immunoglobulin at room temperature for 30 minutes, followed by avidin-biotin-peroxidase complexes for 30 minutes. Three-amino-9-ethylcarbazole was used as the chromogen; and hematoxylin, as the nuclear counterstain. AKT pathway at original magnification ×100. Red arrow indicates significant correlation between 2 proteins (A); AKT pathway markers at high magnification: IGFR1, PDGFRβ, AKT, pAKT, p-p70S6K, p-mTOR are visualized at original magnification ×400; EGFR, VEGFR1, TRKB, TRX-1, PTEN, pPTEN, PI3K, and 4EBP1, at original magnification ×630. Abbreviations: AKT indicates serine/threonine protein kinase.
Thioredoxin and AKT pathway in neuroblastoma
1731
Fig. 1
eluate were used for the kinase assay with an enzyme-linked immunosorbent assay–based AKT activity assay using a biotinylated peptide substrate phosphorylated by AKT kinase (K-LISA AKT activity assay; Calbiochem). AKT
(continued).
activity was quantified by reading the absorbance at 450 nm, with a reference wavelength set at 540 nm. All mesurements were performed in triplicate, each with 3 determinations for each condition.
1732
H. Sartelet et al.
2.6. Cell proliferation assay
3. Results
Chemotherapy-induced cytotoxicity was determined by MTT (3-(4, 5-dimethylthiazol-2-y1)2, 5-diphenyltetrazolium) cell proliferation assay (Cell Titer 96; Fisher Scientific, Rockville, MD). For each of the 3 cell lines, we incubated cells for 24 hours with various concentrations of doxorubicin, a chemotherapeutic agent commonly used in the treatment of neuroblastoma (0.1-36 μmol/L); LY294002, an AKT-specific inhibitor (2.5-605 μmol/L; Calbiochem); deguelin, an AKT-specific inhibitor (1.5-252 μmol/L; Calbiochem); or everolimus (RAD001), a specific mTOR inhibitor (0.65-104 μmol/L; Novartis). Absorbency was measured at 570 nm. Assays were performed 3 times. The mean cell viability was compared with that of positive control cells receiving only medium.
3.1. Patient characteristics
2.7. Cell cycle analysis by flow cytometry Each of the cell lines, when almost 60% confluent, was incubated with dimethyl sulfoxide alone (control) or added to one of the following interventions: LY294002 20 μmol/L, everolimus (RAD001) 10 μmol/L, or hrTRX-1 10 μmol/L for 24 hours at 37°C in a CO2 incubator. Once attached the cells were trypsinized, 1 × 106 fixed cells were harvested by centrifugation and washed 3 times with phosphate-buffered saline. Cells were then resuspended in 0.5-mL fluorochrome solution containing 50 μg/mL propidium iodide, 0.1% sodium citrate, 0.1% Triton X-100, and 0.1 mg/mL ribonuclease A. After a 1-hour incubation at 4°C protected from light, the cells were analyzed on a Beckman Coulter (Brea, CA, USA) EPICS-XL flow cytometer. The mean value was determined from 3 independent experiments.
2.8. Statistical analysis We compared paired data using the Wilcoxon signed rank test. The univariate relationships between immunohistochemical expression in tumor tissues and clinical variables such as age, disease stage as per the International Neuroblastoma Staging System, and histologic type were investigated using a Wilcoxon test. Spearman correlation values (ρ) were used to compare the expression of proteins in the primary tumors. Event-free survival was computed from the time of surgery of the primary tumor (baseline) to the time of first event (local relapse, metastasis, or death) or last follow-up; overall survival was computed from the time of surgery to the time of death or last follow-up. Differences in survival between patients with a low versus high level of expression were assessed on a log-rank test. For the in vitro study, results were expressed as means ± SD of at least 3 independent experiments. The Student t test was used to determine statistical significance. Statistical analyses were performed with SAS software version 8.2 (SAS, Cary, NC). A P value b.05 indicated statistical significance.
Clinicopathologic characteristics of the 101 patients and associated tumors are detailed in Table 1. The median follow-up was 60 months (range, 1-174 months), with a median age at diagnosis of 30 months (range, newborn to 151 months).
3.2. Activation of the AKT pathway in neuroblastoma Interobserver agreement in immunostaining scores between the 2 pathologists was κ = 0.61 (95% confidence interval, 0.58-0.64; P b .0001), before consensus was reached. AKT and pAKT were expressed in 98% and 97% of tumors, respectively, with a median semiquantitative score of 2 Table 1
Patient characteristics at diagnosis
Patient characteristics Sex Male, n (%) Female, n (%) Age (mo) Median (range) b12 mo, n (%) 1-5 y, n (%) N5 y, n (%) Follow-up (mo) Median (range) Survival Alive at time of last follow-up, n (%) INSS stage 1, n (%) 2, n (%) 3, n (%) 4, n (%) 4S, n (%) N-myc oncogene analysis b10 copies, n (%) N10 copies, n (%) Unknown, n (%) Shimada histopathologic classification Favorable, n (%) Unfavorable, n (%) Children's Oncology Group Risk Classification Low, n (%) Intermediate, n (%) High, n (%) Unknown, n (%) Sample type Primary tumor, n (%) Paired metastasis, n (%) Paired control normal tissue, n (%)
(N = 101) 52 (51.4) 49 (48.6) 30 (0-151) 30 (29.7) 53 (52.4) 18 (17.8) 60 (1-174) 64 (63.3) 16 (15.8) 7 (6.9) 22 (21.8) 48 (47.5) 8 (7.9) 67 (77.9) 19 (22.1) 15 49 (48.5) 52 (51.5) 19 (20.2) 21 (22.3) 54 (57.4) 7 101 39 (38.6) 56 (55.4)
Abbreviation: INSS indicates International Neuroblastoma Staging System.
Thioredoxin and AKT pathway in neuroblastoma (Table 2, Fig. 1A and B). These data were confirmed by Western blot analysis with pAKT expression found in 7 (87.5%) of the 8 tumors studied (Fig. 2A). PI3K was expressed in 61% of tumors. Among the membranous tyrosine kinase receptors studied, only IGF1R, TRKB, and PDGFRβ showed an intense and very frequent expression (Table 2, Fig. 1A and B). Among the VEGF receptors, the ligands VEGF (58% of tumors) and VEGFR1 (63%) were moderately expressed (median score of 1 for both), as opposed to VEGFR2 (27%; median score 0). The EGF (epidermal growth factor) family of receptors was rarely expressed. Among the downstream proteins, p-p70S6K (98% of tumors; median score 3) and 4EBP1 (95% of tumors; median score 3) were highly and frequently expressed, as opposed to p-mTOR (Table 2, Fig. 1A and B). IGF1R was more highly expressed in metastases than in primary tumors (P = .01). There was a strong correlation between PI3K and pAKT expression (ρ = 0.54; P b .0001) and between AKT and pAKT expression (ρ = 0.46; P b .0001). Among the tyrosine kinase receptors capable of upstream activation of AKT, there was a significant positive correlation between pAKT and VEGFR1 (ρ = 0.25; P = .01), IGF1R (ρ = 0.24; P = .02), and TRKB (ρ = 0.27; P = .008). Moreover, VEGF was highly correlated with pAKT (ρ = 0.33; P = .0008). The correlation between VEGF and its receptors was significant for VEGFR1 (ρ = 0.27; P = .007) and VEGFR2 (ρ = 0.46; P b .0001). Among downstream proteins, the correlation was high between pAKT and p-p70S6K (ρ = 0.43; P b .0001). Three proteins were expressed in the nucleus: pPTEN, pAKT, and Table 2
Protein expression in primary neuroblastoma
Protein
Median score (range)
% of positive tumors
Staining location
EGFR HER2 VEGF VEGFR1 VEGFR2 IGF1R TRKB PDGFRα PDGFRβ AKT pAKT PI3K PTEN pPTEN TRX-1 p-p70S6K p-mTOR 4EBP1
0 0 1 1 0 2 2 1 2 2 2 1 2.5 0 1 3 1 3
19 2 58 63 27 96 94 76 92 98 97 61 92 63 76 98 51 95
M M C M and C M and C M and C M M M C C and N C C C and N C and N C C C
(0-3) (0-0) (0-4) (0-4) (0-4) (0-3.5) (0-4) (0-3) (0-4) (0-4) (0-4) (0-3) (0-4) (0-3) (0-4) (0-4) (0-3) (0-4)
Immunostaining scores were established by semiquantitative optical analysis of samples containing more than 10 neuroblasts, assessing the percentage of positive cells in each sample: 0, none, all cells negative; 1+, up to 25% of cells were positive; 2+, 26% to 50%; 3+, 51% to 75%; 4+, more than 75%. Abbreviations: M indicates membrane; C, cytoplasm; N, nucleus; PDGFRα, platelet-derived growth factor receptor.
1733
Fig. 2 Activation of the AKT pathway by TRX-1. A, Western blot analysis of frozen neuroblastoma samples. Four samples from stage 1 tumors of infants younger than 1 year (tumors 1-4) and 4 from stage 4 tumors of patients older than 1 year (tumors 5-8) were immunoblotted with antibodies against pAKT, TRX-1, or β-actin (positive control) for 1 hour at room temperature. B, Western blot analysis of a neuroblastoma cell line. Almost 60% confluent SK-N-SH cells were incubated with hrTRX-1 10 μmol/L, RAD001 10 μmol/L, or LY294002 20 μmol/L for 3 hours at 37°C in a CO2 incubator and immunoblotted with antibodies against pAKT, p-mTOR, and β-actin (positive control). C, Cell line AKT kinase assay. SK-N-SH, SK-N-AS, and NB10 cells from cultures, almost 60% confluent, were incubated with LY294002 20 μmol/L, RAD001 20 μmol/L, or hrTRX-1 10 μmol/L for 3 hours at 37°C in a CO2 incubator. To assess TRX-1 activity, cells were incubated with LY294002 or RAD001 in combination with TRX-1. Means of 3 separate experiments are shown as well as SDs (error bars). ⁎P values b.05.
TRX-1 (5%, 40%, 75% of cells, respectively; Fig. 1B). This nuclear expression was not correlated with patient outcome, cell proliferation, or mitosis-karyorrhexis index.
3.3. Relationship between protein expression and clinical covariates On immunohistochemical studies, no relationship was found between the expression of pAKT or p-mTOR in the
1734 tumors and clinical variables. On Western blot analysis, however, the expression of pAKT was demonstrably lower in stage 1 than 4 (Fig. 2A), with 2 low expressions and 1 moderate in stage 1 versus 2 high and 3 moderate expressions in stage 4. Similarly, AKT was more highly expressed in metastatic versus nonmetastatic stage tumors (P = .001). Survival analysis found no correlation between patient outcome and tumor AKT expression (P = .1), but event-free survival tended to be lower in patients with a high expression of AKT (P = .04). TRKB expression was higher in older patients (P = .004) and in metastatic stages (P b .001). Survival analysis found that patients whose tumor expressed more TRKB (above a median score of 2) had a significantly worse survival and lower event-free survival rate than those whose tumors showed a low expression of TRKB (P = .004 and P b .001, respectively). The N-myc
H. Sartelet et al. amplification status showed no correlation with any of the proteins studied.
3.4. Regulation of PTEN activity in neuroblastoma PTEN was only expressed in the cytoplasm of tumors (92%; median score, 2.5). Expression was significantly correlated with that of pAKT (ρ = 0.33; P = .0009). pPTEN, an inactivated form of PTEN, had a very low and inconsistent expression (63% of tumors; median score, 0) (Table 2, Fig. 1A and B), whrereas TRX-1 expression was moderate and frequent (76% of tumors; median score, 1.3). The highly significant correlations between TRX-1 and PTEN (ρ = 0.26; P b .0001) and between TRX-1 and pAKT (ρ = 0.17; P b .0001) were confirmed by Western
Fig. 3 The cytotoxic effect of chemotherapeutic agents modified by TRX-1. A, The effect of specific inhibitors alone on cell viability. The effect of chemotherapy alone in neuroblasts from 3 cell lines (SK-N-SH, SK-N-AS, and NB10) was assessed using a standard MTT assay. Varying concentrations of doxorubicin, a chemotherapeutic agent commonly used in the treatment of neuroblastoma; AKT inhibitors LY294002 and deguelin; and mTOR inhibitor everolimus (RAD001) were added to the culture medium. Data are expressed as a percentage of control culture conditions. Means of 3 separate experiments are shown as well as SDs (error bars); ⁎P b .01 for the 3 cell lines. B, The combined effect of TRX-1 and cytotoxic agents. SK-N-SH, SK-N-AS, and NB10 cells were incubated for 24 hours with cytotoxic agents in combination with hrTRX-1 at progressively increased concentrations (0-16 μmol/L) or with hrTRX-1 alone. Doxorubicin, LY294002, and RAD001 were used at the half maximal inhibitory concentrations as determined from the experiments in A: 4, 20, and 10 μmol/L, respectively. Cell viability was measured by the MTT assay. Data are expressed as a percentage of control culture conditions. Means of 3 separate experiments are shown as well as SDs (error bars). ⁎P b .05 for the 3 cell lines.
Thioredoxin and AKT pathway in neuroblastoma
1735
Fig. 3
blot, with correlation between the expression of pAKT and TRX-1 (or lack thereof) in 7 of the 8 tumors studied (Fig. 2A).
3.5. Inhibition of the AKT pathway in neuroblastoma cell lines Of the 3 AKT inhibitors tested, 2 (LY294002 and RAD001) significantly reduced the activation of AKT; this was correlated with decreased kinase activity of AKT in cell
(continued).
lines (P b .05), the activation of mTOR being also decreased (Fig. 2B and C). Treatment with LY294002 and RAD001 also induced a significant decrease of viable cells in all 3 cell lines studied, as did doxorubicine, a chemotherapeutic agent often prescribed in the treatment of neuroblastoma (P b .01; Fig. 3A). The half maximal inhibitory concentrations for LY294002 and RAD001 for 24 hours of incubation were determined as 20 and 10 μmol/L, respectively (Fig. 3A). In SK-N-SH, SK-N-AS, and NB10 cell lines, the percentage of cells in S phase (synthesis) was significantly reduced when treated with LY294002 (8.62% ± 0.3% and 6.44% ± 0.3%,
1736 respectively) and RAD001 (5.04% ± 1.1% and 9.56% ± 0.4%, respectively), as compared with control medium (12.11% ± 0.2% and 13.4% ± 1.2%) (P b .01; Fig. 4). These observations suggest that the AKT inhibitors induced a G1 cell cycle arrest. Treatment with up to 252 μmol/L of deguelin showed no significant change in cell viability in any of the cell lines studied (Fig. 3A).
H. Sartelet et al. significant (Fig. 3B). When cell cycle analysis was performed in SK-N-SH, SK-N-AS, and NB-10, a significantly higher number of cells were found to be in the mitotic (G2/M) phase when treated with TRX-1 as compared with those without TRX-1 (9.1% ± 0.4%, 11.2% ± 0.2%, and 10.7% ± 0.8% versus 3.5% ± 0.5%, 4.2% ± 0.2%, and 5.1% ± 0.3%; P b .01) (Fig. 4). These results indicated a cell growth-induction effect by TRX-1.
3.6. Activation of the AKT pathway by TRX-1 TRX-1 significantly up-regulated AKT activation in neuroblasts, as demonstrated by an in vitro kinase assay performed on total cellular extracts after exposure to 10 μmol/L hrTRX-1 (P b .05; Fig. 2C). These data were confirmed through Western blot, by the increased level of the activated form of AKT (pAKT) after exposure (Fig. 2B). The hrTRX-1 dampened the down-regulation of AKT activity by LY294002 and RAD001 (Fig. 2C). TRX-1 decreased the cytotoxicity of both AKT inhibitors as well as that of doxorubicin (Fig. 3B). When used alone, TRX-1 induced a mild increase in cell viability, which was not statistically
4. Discussion Although substantial progress has been made in the treatment for children with low- and intermediate-risk neuroblastoma, the cure rate for high-risk patients remains poor. To identify novel therapeutic targets, it is important to uncover pathways critical to neuroblastoma tumorigenesis. The AKT pathway is of particular interest because it is associated with several tyrosine kinase receptors currently targeted by numerous anticancer drugs [17].
Fig. 4 Cell cycle analysis. SK-N-SH, SK-N-AS, and NB-10 cells were incubated with dimethyl sulfoxide (control) or with LY294002 20 μmol/L or everolimus (RAD001) 10 μmol/L or hrTRX-1 10 μmol/L. Percentages of cells in gap (G0/G1), synthesis (S), and mitosis (G2/M) phases are presented in histograms. The experiment was repeated 3 times. Data are shown as mean ± SD.
Thioredoxin and AKT pathway in neuroblastoma Our study confirmed that the AKT pathway was activated in neuroblastoma but failed to demonstrate a correlation between this activation and prognostic factors, in contrast to a previous study [16]. This difference may be explained in part by the different methodologies used, such as the number of core biopsies per tumor, doublecontrol analysis by independent pathologists, quantification of positive cells, and statistical design addressing the issue of clinical correlations. Nonetheless, in our study, the level of AKT protein expression was correlated with a poorer outcome, where event-free survival was significantly lower in patients displaying a high level of AKT. A significant correlation was observed between PI3K, an AKT activator, and pAKT, the activated form of AKT. Moreover, downstream proteins (p-p70S6K and 4EBP1) were present in more than 92% of primary tumors and metastases, a high expression confirming AKT pathway activation. Our data suggested that, among the tyrosine kinase receptors, TRKB, PDGFRβ, and IGF1R might represent targets of interest for specific therapeutic intervention. We also found that VEGF and VEGFR1 had moderate but frequent expression, the significant correlation between the molecule and its receptor strongly suggesting paracrine and autocrine activation [31]. With respect to the EGF receptor family, our results indicated that HER2 and EGFR expressions were very rare in neuroblastoma and showed no correlation with clinical findings, in concordance with a previous study [32] but contrary to others [33]. Of the 3 AKT inhibitors tested, only LY294002 and RAD001 (not deguelin) significantly decreased neuroblast survival and induced a G1 cell cycle arrest. RAD001 is a specific mTOR inhibitor; it likely blocks AKT activation by inhibiting the formation of mTOR complex 2; mTOR complex 2 is known to phosphorylate and activate AKT [34]. In neuroblastoma and acute myeloid leukemia, RAD001 also decreased cell survival. TRX-1, which activates the AKT pathway, partially reversed the action of RAD001, LY294002, and doxorubicin. Several studies have demonstrated that chemosensitivity to doxorubicin was regulated by the AKT pathway [35]. PTEN is a tumor-suppressor protein that negatively regulates the PI3K/AKT signaling pathway by dephosphorylating phosphatidylinositol 3′-kinase [6]. Although found in many malignancies [36], mutations in the PTEN gene are rare in neuroblastoma and may be responsible for malignant progression in only a limited percentage of cases [37]. In many cancers, the presence of molecular alterations of PTEN is often not significantly correlated with PTEN expression, as evidenced from immunohistochemical assays [38]. In our study, the monoclonal antibody assay for PTEN demonstrated only cytoplasmic staining and never nuclear expression. Despite an expression of PTEN in 92% of paired primary neuroblastomas, it is worth noting that pAKT and pp70S6K were still expressed in 97% and 98% of tumors, respectively, demonstrating continued activation of the AKT
1737 pathway. Further to a previous report of a positive correlation between the expression of PTEN and that of pAKT [38], we investigated pPTEN, which is the inactivated form of PTEN (phosphorylation inactivates PTEN by decreasing its phosphatase activity [21]), and TRX-1, a protein that inhibits dephosphorylation of phosphatidylinositol 3′-kinase by PTEN [21]. We observed an inconsistent presence of pPTEN and at low levels thereof. This finding in itself, therefore, cannot explain the high levels of PTEN expression in neuroblastoma without inactivation of the AKT pathway. Thioredoxin (TRX) is a key molecule for redox regulation. TRX-transgenic mice are more resistant to infection, inflammation, and ischemic diseases and survive longer than control mice [39]. TRX is an important regulator of the cell cycle in the G1 phase via cyclin D1 transcription and the ERK/AP-1 (extracellular signal-regulated kinase/ activator protein-1) signaling pathways [40]. However, TRX-1 was found to bind to the catalytic site of PTEN and to its C2 lipid membrane-binding domain [22]. Indeed, previous studies had suggested that the increased levels of TRX observed in human tumors could lead to a functional inhibition of PTEN tumor-suppressor activity [22]. In cancer cells, TRX-1 overexpression has been associated with a biologically aggressive cancer phenotype [25] and resistance to chemotherapeutic agents such as doxorubicin and cysplatin, drugs currently used in the treatment of neuroblastoma [26]. TRX was highly expressed in several neuroblastoma cell lines as well [23,28,29]. Our study was the first to show that TRX-1 was expressed in a large series of neuroblastomas from patients and that its expression was correlated with both PTEN and pAKT expressions. Hence, AKT activation despite a high level of PTEN was associated with the expression of TRX-1 in neuroblastoma. In in vitro [41,42] and in vivo studies [43], human recombinant TRX-1 enhanced cellular resistance to chemotherapy [41] and prolonged survival of cancer cells [42]. In this study, we demonstrated that hrTRX-1 induced AKT activation in neuroblastoma cell lines. Moreover, it partially inhibited the action of several chemotherapeutic agents, including AKT inhibitors; increased cell viability; and induced cell growth. Together, these data strongly suggest that specific inhibitors of TRX-1 alone or in combination with classical chemotherapy could be beneficial in the treatment of neuroblastoma. PX-12 (1-methylpropyl 2-imidazolyl disulfide) is a specific TRX-1 inhibitor that shows both excellent in vitro and promising in vivo antitumor activity [44]. A detailed knowledge of tumor cell genomics is essential in establishing the risk classifications in neuroblastomas [45]. We know that DNA copy number gain represents only one of the many mechanisms that can cause protein overexpression. The evaluation of the intensity of protein expression in tumor samples by immunohistochemistry, including tissue microarray and Western blot analysis, as was done in our study, represents a very efficient means of detecting new therapeutic targets in neuroblastoma and of providing patient follow-up.
1738 Our study confirmed that the AKT pathway was activated in primary and metastatic neuroblastomas and demonstrated a correlation between this AKT activation and the presence of TRKB, IGF1R, VEGFR1, and, in particular, TRX-1. TRX-1 seems to be a key player. These data suggest the feasibility of therapeutic intervention on the AKT pathway in neuroblastoma through a combination of targeted therapies.
Acknowledgment The authors thank Danielle Buch, our medical editor, for editing and revising the manuscript.
References [1] Spix C, Pastore G, Sankila R, et al. Neuroblastoma incidence and survival in European children (1978-1997): report from the Automated Childhood Cancer Information System project. Eur J Cancer 2006;42: 2081-91. [2] Escobar MA, Grosfeld JL, Powell RL, et al. Long-term outcomes in patients with stage IV neuroblastoma. J Pediatr Surg 2006;41:377-81. [3] Brazil DP, Park J, Hemmings BA. PKB binding proteins. Getting in on the Akt. Cell 2002;111:293-303. [4] Kandel ES, Hay N. The regulation and activities of the multifunctional serine/threonine kinase Akt/PKB. Exp Cell Res 1999;253:210-29. [5] Nicholson KM, Anderson NG. The protein kinase B/Akt signalling pathway in human malignancy. Cell Signalling 2002;14:381-95. [6] Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate. J Biol Chem 1998;273:13375-8. [7] Fang CX, Yang X, Sreejayan N, et al. Acetaldehyde promotes rapamycin-dependent activation of p70(S6K) and glucose uptake despite inhibition of Akt and mTOR in dopaminergic SH-SY5Y human neuroblastoma cells. Exp Neurol 2007;203:196-204. [8] Fleming JM, Desury G, Polanco TA, et al. IGF-I and epidermal growth factor receptors recruit distinct upstream signalling molecules to enhance AKT activation in mammary epithelial cells. Endocrinology 2006;147:6027-35. [9] Kyrylenko S, Roschier M, Korhonen P, et al. Regulation of PTEN expression in neuronal apoptosis. Brain Res Mol Brain Res 1999;73: 198-202. [10] Chesler L, Schlieve C, Goldenberg DD, et al. Inhibition of phosphatidylinositol 3-kinase destabilizes Mycn protein and blocks malignant progression in neuroblastoma. Cancer Res 2006;66: 8139-46. [11] Efron PA, Chen MK, Iyengar M, et al. Differential response of neuroblastoma cells to TRAIL is independent of PI3K/AKT. J Pediatr Surg 2006;41:1072-80. [12] Kim S, Kang J, Qiao J, et al. Phosphatidylinositol 3-kinase inhibition down-regulates survivin and facilitates TRAIL-mediated apoptosis in neuroblastomas. J Pediatr Surg 2004;39:516-21. [13] Ruffels J, Griffin M, Dickenson JM. Activation of ERK1/2, JNK and PKB by hydrogen peroxide in human SH-SY5Y neuroblastoma cells: role of ERK1/2 in H2O2-induced cell death. Eur J Pharmacol 2004;483:163-73. [14] Shankar SL, Krupski M, Parashar B, et al. UCN-01 alters phosphorylation of Akt and GSK3beta and induces apoptosis in six independent human neuroblastoma cell lines. J Neurochem 2004;90: 702-11. [15] Liu XH, Yu EZ, Li YY, et al. RNA interference targeting Akt promotes apoptosis in hypoxia-exposed human neuroblastoma cells. Brain Res 2006;1070:24-30.
H. Sartelet et al. [16] Opel D, Poremba C, Simon T, et al. Activation of Akt predicts poor outcome in neuroblastoma. Cancer Res 2007;67:735-74. [17] Sartelet H, Oligny LL, Vassal G. AKT pathway in neuroblastoma and its therapeutic implication. Expert Rev Anticancer Ther 2008;8:757-69. [18] Vlahos CJ, Matter WF, Hui KY, et al. A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002). J Biol Chem 1994;269:5241-8. [19] Bortul R, Tazzari PL, Billi AM, et al. Deguelin, A PI3K/AKT inhibitor, enhances chemosensitivity of leukaemia cells with an active PI3K/AKT pathway. Br J Haematol 2005;129:677-86. [20] Witzig TE, Kaufmann SH. Inhibition of the phosphatidylinositol 3kinase/mammalian target of rapamycin pathway in hematologic malignancies. Curr Treat Options Oncol 2006;7:285-94. [21] Gericke A, Munson M, Ross AH. Regulation of the PTEN phosphatase. Gene 2006;374:1-9. [22] Meuillet EJ, Mahadevan D, Berggren M, et al. Thioredoxin-1 binds to the C2 domain of PTEN inhibiting PTEN's lipid phosphatase activity and membrane binding: a mechanism for the functional loss of PTEN's tumor suppressor activity. Arch Biochem Biophys 2004;429: 123-33. [23] Andoh T, Chock PB, Chiueh CC. The roles of thioredoxin in protection against oxidative stress–induced apoptosis in SH-SY5Y cells. J Biol Chem 2002;277:9655-60. [24] Nakamura H, Nakamura K, Yodoi J. Redox regulation of cellular activation. Annu Rev Immunol 1997;15:351-69. [25] Noike T, Miwa S, Soeda J, et al. Increased expression of thioredoxin-1, vascular endothelial growth factor, and redox factor-1 is associated with poor prognosis in patients with liver metastasis from colorectal cancer. HUM PATHOL 2008;39:201-8. [26] Yokomizo A, Ono M, Nanri H, et al. Cellular levels of thioredoxin associated with drug sensitivity to cisplatin, mitomycin C, doxorubicin, and etoposide. Cancer Res 1995;55:4293-6. [27] Mukherjee A, Huber K, Evans H, et al. A cellular and molecular investigation of the action of PMX464, a putative thioredoxin inhibitor, in normal and colorectal cancer cell lines. Br J Pharmacol 2007;151:1167-75. [28] Li Y, Lu Z, Chen F, et al. Antisense bcl-2 transfection up-regulates anti-apoptotic and anti-oxidant thioredoxin in neuroblastoma cells. J Neurooncol 2005;72:17-23. [29] Farina AR, Tacconelli A, Cappabianca L, et al. Thioredoxin alters the matrix metalloproteinase/tissue inhibitors of metalloproteinase balance and stimulates human SK-N-SH neuroblastoma cell invasion. Eur J Biochem 2001;268:405-13. [30] Shapiro DN, Valentine MB, Rowe ST, et al. Detection of N-myc gene amplification by fluorescence in situ hybridization. Diagnostic utility for neuroblastoma. Am J Pathol 1993;142:1339-46. [31] Langer I, Vertongen P, Perret J, et al. Expression of vascular endothelial growth factor (VEGF) and VEGF receptors in human neuroblastomas. Med Pediatr Oncol 2000;34:386-93. [32] Langer I, Vertongen P, Perret J, et al. Expression of HER2/neu is uncommon in human neuroblastic tumors and is unrelated to tumor progression. Cancer Immunol Immunother 2003;52:116-20. [33] Zeng Z, Sarbassov dos D, Samudio IJ, et al. Rapamycin derivatives reduce mTORC2 signaling and inhibit AKT activation in AML. Blood 2007;109:3509-12. [34] Ho R, Minturn JE, Hishiki T, et al. Proliferation of human neuroblastomas mediated by the epidermal growth factor receptor. Cancer Res 2005;65:9868-75. [35] Gagnon V, Van Themsche C, Turner S, et al. Akt and XIAP regulate the sensitivity of human uterine cancer cells to cisplatin, doxorubicin and taxol. Apoptosis 2008;13:259-71. [36] Athanassiadou P, Athanassiades P, Grapsa D. The prognostic value of PTEN, p53, and beta-catenin in endometrial carcinoma: a prospective immunocytochemical study. Int J Gynecol Cancer 2007;17:697-704. [37] Moritake H, Horii Y, Kuroda H, et al. Analysis of PTEN/MMAC1 alteration in neuroblastoma. Cancer Genet Cytogenet 2001;125: 151-5.
Thioredoxin and AKT pathway in neuroblastoma [38] Pallares J, Bussaglia E, Martínez-Guitarte JL, et al. Immunohistochemical analysis of PTEN in endometrial carcinoma: a tissue microarray study with a comparison of four commercial antibodies in correlation with molecular abnormalities. Mod Pathol 2005;18:719-27. [39] Mitsui A, Hamuro J, Nakamura H, et al. Overexpression of human thioredoxin in transgenic mice controls oxidative stress and life span. Antioxid Redox Signal 2002;4:693-6. [40] Mochizuki M, Kwon YW, Yodoi J, et al. Thioredoxin regulates cell cycle via the ERK1/2-cyclin D1 pathway. Antioxid Redox Signal 2009;11:2957-71. [41] Ueda S, Nakamura T, Yamada A, et al. Recombinant human thioredoxin suppresses lipopolysaccharide-induced bronchoalveolar neutrophil infiltration in rat. Life Sci 2006;79:1170-7.
1739 [42] Sasada T, Nakamura H, Ueda S, et al. Secretion of thioredoxin enhances cellular resistance to cis-diamminedichloroplatinum (II). Antioxid Redox Signal 2000;2:695-705. [43] Nilsson J, Söderberg O, Nilsson K, et al. Thioredoxin prolongs survival of B-type chronic lymphocytic leukemia cells. Blood 2000; 95:1420-6. [44] Kirkpatrick DL, Kuperus M, Dowdeswell M, et al. Mechanisms of inhibition of the thioredoxin growth factor system by antitumor 2-imidazolyl disulfides. Biochem Pharmacol 1998;55:987-94. [45] Ambros PF, Ambros IM, Brodeur GM, et al. International consensus for neuroblastoma molecular diagnostics: report from the International Neuroblastoma Risk Group (INRG) Biology Committee. Br J Cancer 2009;100:1471-82.