Purpose: Vinblastine and rapamycin displayed synergistic inhibition of human neuroblastoma-related angiogenesis. Here, we studied the antitumor activity of vinblastine and rapamycin against human neuroblastoma.
Experimental Design: Cell proliferation, cell cycle progression, and apoptosis were evaluated by measuring 3H-thymidine incorporation, bromodeoxyuridine uptake, and phosphatidylserine exposure, respectively. The in vivo sensitivity of neuroblastoma cells to vinblastine and rapamycin was determined in orthotopic neuroblastoma-engrafted mice. Angiogenesis was assessed by the chick embryo chorioallantoic membrane assay.
Results: Each compound alone was able to induce a dose-dependent significant inhibition of cell proliferation, with a dramatically enhanced antiproliferative effect for the drugs used in combination. A marked G2-M cell cycle arrest with a nearly complete depletion of S phase was associated. The combined treatment triggered an increased apoptosis compared with either drug tested alone. A significant inhibition of tumor growth and microvessel area was obtained in neuroblastoma-bearing mice when treated with vinblastine or rapamycin alone, and a more dramatic effect with the combined treatment, compared with control mice. The therapeutic effectiveness, expressed as increased life span, was statistically improved by the combined therapy, compared with mice treated with either drug tested separately. Histologic evaluation of primary tumors showed that the combined treatment inhibited proliferation and angiogenesis and induced apoptosis. Combined treatment of neuroblastoma cells and neuroblastoma-bearing mice with vinblastine and rapamycin induced the down-modulation of both vascular endothelial growth factor production and vascular endothelial growth factor receptor 2 expression. In the chorioallantoic membrane assay, angiogenesis induced by human neuroblastoma biopsy specimens was significantly inhibited by vinblastine and rapamycin.
Conclusions: These results may be relevant to design new therapeutic strategies against neuroblastoma.
Neuroblastoma is the most common extracranial solid tumor of infancy (1). Despite of aggressive treatment strategies, such as high-dose chemotherapy and bone marrow transplantation, the prognosis for patients suffering from advanced stage disease has not been improved in a satisfactory manner and neuroblastoma continues to present a formidable clinical challenge (2). Therefore, development of new treatment approaches has been the focus, in the last few years, of many neuroblastoma studies.
The Vinca alkaloid vinblastine is a microtubule inhibitor belonging to an important class of chemotherapeutic agents that affect polymerization and stability of microtubules (3). Vinca alkaloids prevent mitotic spindle formation, resulting in a G2-M phase cell cycle arrest and subsequent apoptotic cell death (4). Vinblastine is mainly used, in combination with other drugs, for therapy of several lymphoid malignancies and some solid pediatric tumors (5, 6).
Rapamycin is a lipophilic macrolide antibiotic, originally identified as an anti-fungal agent. Many studies have, however, showed that rapamycin and its derivatives present both immunosuppressant (7, 8) and antitumor properties (9). Rapamycin acts by specifically inhibiting the mammalian target of rapamycin protein kinase. The inhibition of mammalian target of rapamycin pathways determines alterations in cell cycle progression, resulting in the blockage of cells in G1 phase (10–12).
In the last decade, several cytotoxic drugs, already used in the clinical setting, have been investigated for their potential antiangiogenic effects (13). In this respect, it has been recently reported that both vinblastine and rapamycin, tested at low doses, exerted antiangiogenic activity in vitro and in vivo, through two different mechanisms (14, 15). Furthermore, we have lately shown that vinblastine and rapamycin used in combination, at very low doses, inhibit synergistically human neuroblastoma-related angiogenesis (16). Moreover, proteomic analysis of vinblastine- and rapamycin-treated or untreated endothelial cells allowed us to show that many proteins involved in cell proliferation, migration, apoptosis, and angiogenesis were modulated only by the combined treatment (17).
Vinblastine has been used for chemotherapy of human neuroblastoma, but dose-limiting side effects (i.e., vascular-associated toxicity) have been observed (18). Furthermore, it has been shown that rapamycin inhibited in vitro human neuroblastoma cell proliferation (19).
Based on these data, the primary goal of this study was to achieve proof-of-principle for the hypothesis that combined administration of two drugs targeting different phases of the cell cycle results into therapeutic effects superior to those achieved upon treatment with either drug alone. Thus, herein, we have investigated the antitumor activity of both vinblastine and rapamycin tested alone or in combination in vitro on a panel of neuroblastoma cell lines and in vivo on an orthotopic xenograft mouse model of human neuroblastoma.
Materials and Methods
Chemotherapeutics. Vinblastine (Lilly France SA) was solubilized in PBS and stored at −20°C. Rapamycin (ICN Biomedicals, Inc.) was dissolved in DMSO and kept at −20°C. At the time of use, the drugs were freshly prepared and diluted stepwise, until the desired concentrations, with the culture medium.
Cell lines and culture conditions. The following human neuroblastoma cell lines were used: GI-LI-N (20, 21), HTLA-230 (22), SH-SY5Y (23), ACN (24). All cell lines were used between passages 50 and 75. All cells were grown in complete DMEM, as previously described (25).
Human umbilical vein endothelial cells (HUVEC) were obtained from the American Type Culture Collection. HUVEC were maintained in endothelial cell basal medium-2 (Cambrex Bio Science), as previously described (14).
Cell proliferation assay. All neuroblastoma cell lines were plated in 96-well plates (at 3-8 × 103 per well) in complete medium and cultured for 24 h. HUVEC were plated in 96-well plates (at 3 × 103 per well) precoated with collagen and then cultured for 24 h. The medium was removed and replaced with fresh complete medium that had been supplemented with different concentrations of either vinblastine (0-4 nmol/L) or rapamycin (0-100 nmol/L), alone or in combination for an additional 72 h. Cells were then pulse-labeled for 18 h with 0.5 μCi (0.0185 MBq) 3H-thymidine (Amersham Biosciences), as previously described (25). Cell proliferation was tested by measuring DNA synthesis as a function of 3H-thymidine uptake.
We further assessed if the association of the two drugs could have produced synergistic effects by calculating the combination indexes. Combination indexes were calculated as follow: CIx = (D1 combined / D1 alone) + (D2 combined / D2 alone) + [(D1 combined × D2 combined) / (D1 alone × D2 alone)], where D1 combined and D2 combined represent the amount of drug 1 and drug 2, respectively, that caused 50% inhibition of cell growth. Both D1 and D2 alone displayed 50% inhibition for the mutually exclusive case where both drugs had different modes of action.
Analysis of DNA synthesis by pulse labeling with bromodeoxyuridine. Exponentially growing neuroblastoma cells (SH-SY5Y and GI-LI-N; 1.5-3 × 106 per 25-cm2 flask) were treated for 24 h with either 1 or 2 nmol/L vinblastine (GI-LI-N and SH-SY5Y, respectively) and 10 nmol/L rapamycin, given alone or in combination. At the end of the treatment, cells were pulse-labeled with 10 μmol/L bromodeoxyuridine (BrdUrd; Sigma) for 30 min, as previously described (16). BrdUrd uptake was detected by staining the cells with 20 μL FITC-conjugated mouse monoclonal anti-BrdUrd antibody (Becton Dickinson) at a final concentration of 5 μg/mL for 30 min at room temperature. The cells were washed and resuspended in PBS containing 5 μg/mL propidium iodide. Bivariate distributions of BrdUrd amounts (FITC) versus DNA content (propidium iodide) were assessed by flow cytometry using a FACSCalibur device (Becton Dickinson), as previously described (16, 26). The gates represent the different phases of the cell cycle (R1, sub-G1 phase; R2, G1 phase; R3, S phase; R4, G2-M phase).
Cell viability assay. Neuroblastoma cells (SH-SY5Y and GI-LI-N) were plated in 12-well plates (1.8-2.5 × 105 per well) and treated for 24 h as described above. Then, cells were harvested, washed with complete medium, and incubated with trypan blue (0.04%; Sigma), as previously described (26). The proportion of dead (or living) cells was calculated by dividing the number of dead (or living) cells by the total number of cells per field.
Phosphatidylserine detection. Phosphatidylserine exposure was assessed with a human Annexin V-FITC kit (Bender MedSystems). Briefly, cultured neuroblastoma cells (GI-LI-N and SH-SY5Y) were treated with vinblastine and rapamycin, administered alone or in combination as already mentioned in the previous paragraphs. Cells were then collected, washed, and incubated for 10 min with 5 μL Annexin V-FITC; washed once with PBS; resuspended in 190 μL pre-diluted binding buffer plus 10 μL of a 20 μg/mL propidium iodide stock solution; and examined by two-color flow cytometry using a FACSCalibur device, as described (16, 26).
Western blot analysis. Protein lysates were prepared, as previously described (26), from GI-LI-N and SH-SY5Y cells treated for 24 h with 1 nmol/L vinblastine (GI-LI-N), 2 nmol/L vinblastine (SH-SY5Y), and 10 nmol/L rapamycin, either alone or in combination. The protein lysates (40 μg per lane) were resolved on SDS/8% polyacrylamide gels; pre-stained molecular weight markers were run in parallel on each gel. The resolved proteins were transferred to nitrocellulose membranes; the membranes were then incubated with a rabbit monoclonal anti–phosphorylated p70 S6 kinase antibody (Thr389; Upstate). The rabbit monoclonal anti–p70 S6 kinase antibody (Upstate) was used as a loading and transfer control. Peroxidase-conjugated goat anti-rabbit antibodies were used as secondary antibodies (Upstate and Chemicon International, respectively). Immune complexes were visualized with the use of an enhanced chemiluminescence system (Amersham International) according to the manufacturer's instructions.
Flow cytometry evaluation of vascular endothelial growth factor receptor 2. Neuroblastoma cells (ACN, SH-SY5Y, HTLA-230, and GI-LI-N) were plated in six-well plates (3-7.5 × 105 per well). The day after plating, neuroblastoma cells were treated for 24 h with 1 nmol/L vinblastine and 10 nmol/L rapamycin, administered alone or in combination. Cells were then collected by scraping and centrifuged at 1,100 rpm for 8 min. Thus, cells were processed for flow cytometry analysis. Briefly, 5 × 105 cells per point were incubated for 30 min at 4°C with an anti–vascular endothelial growth factor receptor 2 (VEGF-R2) mouse monoclonal antibody (clone ab9530, Abcam) at a concentration of 10 μg/mL. As negative control, cells were incubated with an isotype-matched monoclonal antibody (monoclonal mouse IgG1, Ancell Corp.). After incubation, cells were washed twice with 1% fetal bovine serum in PBS and then stained at 4°C for 20 min with a FITC-conjugated goat anti-mouse immunoglobulin (Caltag Laboratories). After washing, cells that stained positive for VEGF-R2 were assessed by a FACSCalibur device, using Cell Quest Software. VEGF-R2 expression has been indicated as mean ratio fluorescence intensity, defined as the ratio between the fluorescence intensity of untreated and treated cells and the fluorescence intensity of their isotype-matched control antibody.
Quantitative evaluation of VEGF production. Neuroblastoma cells (ACN, SH-SY5Y, HTLA-230, and GI-LI-N) were plated and treated as described in the previous paragraph. Supernatants derived from each treatment point were then collected, sequentially centrifuged at 3,000 and 15,000 × g for 10 min, and stored at −80°C until use, as previously reported (27). Secretion of VEGF, by neuroblastoma cells, was determined using a human VEGF detection ELISA kit (R&D Systems GmbH), according to manufacturer's instruction. Because of the possible cytotoxic effects of vinblastine and rapamycin on these cells, the raw data obtained for VEGF concentrations were normalized for the number of viable cells.
In vivo therapeutic studies in mice. All experiments involving animals were reviewed and approved by the licensing and ethical committee of the National Cancer Research Institute, Genoa, Italy and by the Italian Ministry of Health. All in vivo experiments were done using eight mice per group and were repeated twice.
For these studies, we settled a very aggressive orthotopic neuroblastoma mouse model. Four-week-old female athymic (Nude-nu) mice were purchased from Harlan Laboratories and housed under specific pathogen-free conditions. Mice were anesthetized with ketamine (Imalgene 1000, Merial Italia S.p.A.), subjected to laparotomy, and injected with GI-LI-N cells (1.5 × 106 cells in 15 μL of saline solution) in the capsule of the left adrenal gland, as described (28, 29). No mice died as a result of this treatment. Tumors were allowed to grow from the injected cells for 14 days. At this time, mice were randomly assigned to four groups and treated with vinblastine and rapamycin, given alone or in combination. Rapamycin (1.5 mg/kg) was administered i.p. five consecutive days per week; vinblastine (0.5 mg/kg) was injected i.p. every 3 days. For the combined treatment, vinblastine and rapamycin were admixed together and injected as a single bolus. Treatment continued for a total of 5 weeks. Body weight and general physical status of the animals were recorded daily, and mice were killed by cervical dislocation after being anesthetized with xilezine (Xilor 2%, Bio98 Srl), when they showed signs of poor health, such as abdominal dilation, dehydration, or paraplegia.
Survival time, defined as the time (in days) between tumor cell inoculation and euthanasia of mice (n = 8 per group) due to evidence of poor health, was used as the main criterion for determining treatment efficacy. Measurement of tumor volumes has been even used as further confirmation of treatment efficacy. At various times after cell inoculation, mice (n = 3 per time; total, n = 18 per group) were randomly selected and killed by cervical dislocation after being anesthetized with xilezine, and their tumors were measured with calipers. Tumor volumes were calculated using the formula π/6 [w1 × (w2)2], where w1 represents the largest tumor diameter, and w2 represents the smallest tumor diameter.
Histologic analysis of mouse orthotopic tumors. Histologic evaluation of primary tumors was done 30 days after the beginning of the treatment. Briefly, orthotopic tumor-bearing mice (two mice per group) were anesthetized with xilezine and killed by cervical dislocation. Tumor masses were collected, split in two samples, and then processed for either paraffin or optimum cutting temperature compound (Miles Chemical Co.) embedding, respectively, as previously described (26).
Tissue sections (5 μm thick) were examined after staining with Mayer's H&E (Sigma Chemical Co.). Paraffin-embedded tissue sections were deparaffinized by the xylene-ethanol sequence, rehydrated in a graded ethanol scale and in TBS (pH 7.6), and then processed for antigen retrieval using a microwave, by boiling tissue sections in 0.1 mol/L EDTA (pH 8) for 10 min. The sections were then washed twice in PBS and saturated with 2% bovine serum albumin in PBS before staining with primary antibodies against CD31 that recognize endothelial cells (goat anti-mouse, clone SC-1506, Santa Cruz Biotechnology); the proliferation antigen Ki-67 (mouse anti-human Ki-67, clone MIB-1, DAKO); the endothelial cell marker CD34 (rat anti-mouse, clone MEC 14.7, DAKO); the 57-kDa human neuroblastoma-specific antigen, NB84a (mouse anti-human neuroblastoma, clone NB84a, DAKO); the VEGF (mouse anti-human VEGF, clone 26503, R&D Systems); the endothelial growth factor receptor 2 (mouse anti-human VEGF-R2, clone 89115, R&D Systems); and for terminal deoxynucleotidyl transferase–mediated nick-end labeling (TUNEL) analysis using the In situ Cell Death Detection kit (Roche), according to manufacturer's instructions. Binding of the primary antibodies was detected with horseradish peroxidase–conjugated horse anti-goat immunoglobulin G (Vector Laboratories) for immunohistochemical analysis and red phycoerythrin-conjugated rabbit anti-mouse (DAKO) or goat-anti-rat (Southern Biotech) or green FITC-conjugated goat anti-mouse (Caltag Laboratories) immunoglobulin for immunofluorescence analysis. Sections were analyzed for CD31 positivity by staining with diaminobenzidine substrate and, simultaneously, for apoptosis and for staining with the primary antibodies with a Nikon E-1000 fluorescence microscope equipped with specific filters for FITC, tetramethyl rhodamine isothiocyanate, and 4′,6-diamidino-2-phenylindole, as described (26, 30).
For VEGF and VEGF-R2 quantification, sections were observed under an Olympus photomicroscope (Olympus Italia) equipped for fluorescence and digital images acquisition. The intensity of VEGF- and VEGFR-2–related immunofluorescence (FITC label) from each section was quantified using the Image Analysis software (Olympus Italia). Red, green, and black images were obtained decomposing each original section, and the size of FITC-fluorescence brightness was determined by an accurate “thresholding” procedure on the green images. The results were expressed as percentage of FITC-labeled area referred to total area of the section.
Relative quantitative expression analysis of VEGF and VEGF-R2 in adrenal neuroblastoma xenograft in nude mice. Total RNA was isolated from tumor samples of adrenal neuroblastoma xenograft in nude mice, untreated and treated with rapamycin, vinblastine, or rapamycin + vinblastine, using the RNeasy Mini kit (QIAGEN) with an on-column DNase treatment. The concentration and quality control of all RNAs were assessed with the Agilent 2100 BioAnalyzer (Agilent Technologies). Expression analysis was done by a two-step reverse transcription-PCR. The first-strand cDNA was obtained by reverse transcription of 5 μg total RNA using a random primer-based synthesis (High-Capacity cDNA Archive kit, Applied Biosystems) according to the manufacturer's instructions. Relative quantitative expression analysis was carried out by duplex real-time reverse transcription-PCR using FAM-labeled Taqman Gene Expression Assays for each target gene (VEGF and VEGF-R2) and a VIC-labeled assay for the human 18S rRNA housekeeping gene as endogenous control (reference gene), which resulted to be homogeneously and uniformly expressed in series of neuroblastoma cell lines and tumors from patients in a preliminary analysis.5
Patrizia Perri, personal communication.
Microvessel area. Microvessel area in tumor sections that were stained with the CD31 antibody was assessed independently by two investigators (B.N. and D.R.) with the use of a Quantimet 5000 computerized image analysis system (Leica), as previously described (26, 30).
Patients and collection of tumor samples. Ten neuroblastoma specimens collected from 1987 to 1998 were retrieved from the Italian Neuroblastoma Tissue Bank at the onset of the Ethical Committee of the G. Gaslini Children's Hospital, and all patients or their parents gave informed consent. Tumor cell content was always ≥80%. Disease extension was classified according to the International Neuroblastoma Staging System criteria (31). The specimens were from seven patients with stage IV, two patients with stage I, and one with localized stage II tumors.
Chick embryo chorioallantoic membrane assay. Chick embryo chorioallantoic membrane assays have been done according to Ribatti et al. (32). Briefly, growing chorioallantoic membranes (10 eggs per group) were treated by (a) overlaying them with 1-mm3 sterilized gelatin sponges (Gelfoam Upjohn) that had been loaded with 1 μL PBS (negative control), 1 μL PBS containing 500 ng of recombinant fibroblast growth factor-2 (R&D Systems; positive control); (b) grafting fresh biopsy specimens from patients with neuroblastoma that had been collected under sterile conditions and minced in RPMI 1640 to obtain 1- to 2-mm3 fragments onto the chorioallantoic membrane and then treated with PBS or 2 pmol/L vinblastine and 5 pmol/L rapamycin singly or in combination. These doses have been chosen based on our previous findings (16). The chorioallantoic membranes were examined daily until day 12 of incubation and then processed for light microscopy (26, 32). The angiogenic response was assessed by using a planimetric method of point counting (33). Mean values for microvessel area, with 95% confidence intervals (95% CI), were determined for each analysis.
Statistical analysis. All in vitro data are from at least three independent experiments, and the results are expressed as mean values with SD. To determine whether drug combinations exhibited synergistic antiproliferative activity, isobolographic analysis was done. From the dose-response curves, the concentration at which neuroblastoma cell proliferation was inhibited to 75% and to 50% of control level was calculated to generate the isobolograms. The diagonal line shows the theoretical line indicating additive effects: the points below the line indicate synergistic activity, whereas those above the line represent sub-additivity. Different combinations of vinblastine and rapamycin were added to neuroblastoma cells, and the values obtained were plotted in the isobolograms.
The statistical significance of differences between experimental and control groups was determined by ANOVA with Tukey's multiple comparison test using GraphPad Prism 3.0 software (GraphPad Software, Inc.). Survival curves were constructed by using the Kaplan-Meier method. All in vivo experiments were done twice with similar results, and the data are expressed as mean values with 95% CI. Survival in different treatment groups was compared by using Peto's log-rank test in StatsDirect 0.1 statistical software (CamCode). All statistical tests were two sided, and P < 0.05 was considered statistically significant.
Combined treatment with vinblastine and rapamycin inhibits neuroblastoma cell proliferation. We investigated the antiproliferative effects of vinblastine and rapamycin on four neuroblastoma cell lines and HUVEC in vitro, by measuring 3H-thymidine uptake. Both drugs showed a dose-dependent inhibition of cell proliferation on all neuroblastoma cell lines examined as well as on endothelial cells, after a 72-h treatment (Fig. 1A and B). The dose-response curves plotted in Fig. 1 showed that HUVEC were more sensitive to vinblastine and rapamycin (drug concentration that caused IC50 = 200 pmol/L for vinblastine and 75 pmol/L for rapamycin), with respect to neuroblastoma cells (mean IC50 = 1.1 ± 0.4 nmol/L for vinblastine and mean IC50 = 5.5 ± 2 nmol/L for rapamycin).
In our previous report (16), we observed that vinblastine and rapamycin produced, on endothelial cells, a synergistic antiproliferative effect, defined as a combination of two drugs, showing a greater efficacy than that of the expected sum of the effects of each drug.
To determine whether vinblastine and rapamycin in combination had synergistic activity also on neuroblastoma cells, different combinations of concentrations of vinblastine and rapamycin were added to neuroblastoma cells. From the dose-response curves, the concentrations at which GI-LI-N and SH-SY5Y cells proliferation was inhibited to 75% and to 50% of control were calculated. Figure 1C and D shows the isobolograms generated from these data, respectively. When compared with the theoretical line representing additive effect, only the values from the combination treatment obtained at the IC75 fall below the line indicating synergistic activity for this end point, whereas their effects likely indicated summation when combined to reach a 50% inhibition of cell growth. Similar results have been obtained also for HTLA-230 and ACN cells (data not shown). Nevertheless, the combined treatment produced, on HUVEC, synergistic antiproliferative effects at both IC75 and IC50, as expected from our previous work (16).
To have proof-of-principle data regarding the efficacy of the doses herein used of rapamycin alone or in combination with vinblastine on neuroblastoma cells, Western blot analysis has been done on cells treated with these drugs, and S6K protein expression was evaluated. P-S6K (Thr389) was found to be constitutively expressed in neuroblastoma cells, and rapamycin alone and in combination with vinblastine was found to suppress phosphorylated S6K levels, without affecting S6K protein levels (Supplementary Fig. S1).
Vinblastine and rapamycin induce cell cycle arrest in neuroblastoma cells. To investigate whether the antiproliferative effect, displayed by the combination of vinblastine and rapamycin, was associated to an inhibition of cell cycle progression, SH-SY5Y and GI-LI-N cells were treated for 24 h with either 2 or 1 nmol/L vinblastine, respectively, and 10 nmol/L rapamycin, given alone or in combination. These drug concentrations were chosen because they represent likely the IC50 on the two cell lines. Cells were then pulse-labeled with BrdUrd to examine DNA synthesis. BrdUrd uptake versus total cellular DNA content was then evaluated by densitometric fluorescence-activated cell sorting analysis.
As shown in Fig. 2, treatment with rapamycin alone induced, as expected, a G1 phase cell cycle arrest (R2), in both neuroblastoma cell lines analyzed. Vinblastine treatment led to a marked G2-M phase cell cycle arrest (R4), associated with a pronounced decrease of cells actively synthesizing DNA (R3). Furthermore, the combined treatment resulted in a more dramatic block of neuroblastoma cells in G2-M phase [vinblastine, 56 ± 5% of GI-LI-N cells in G2-M phase; rapamycin, 15 ± 2%; vinblastine + rapamycin, 69 ± 6% (P < 0.05 and P < 0.001 versus vinblastine and rapamycin, respectively); vinblastine, 26 ± 3% of SH-SY5Y cells in G2-M phase; rapamycin, 19.5 ± 2%; vinblastine + rapamycin, 39.9 ± 4% (P < 0.001 and P < 0.001 versus vinblastine and rapamycin, respectively)] and a complete arrest of the S phase [vinblastine, 7.2 ± 0.75% of GI-LI-N cells in S phase; rapamycin, 24.5 ± 3%; vinblastine + rapamycin, 1.5 ± 0.2% (P < 0.05 and P < 0.001 versus vinblastine and rapamycin, respectively); vinblastine, 12.9 ± 1.5% of SH-SY5Y cells in S phase; rapamycin, 21.7 ± 2.3%; vinblastine + rapamycin, 4 ± 0.3% (P < 0.01 and P < 0.001 versus vinblastine and rapamycin, respectively)], suggesting that when used together, the two drugs, work to increase the effect produced by vinblastine given alone. These results could explain the dramatic antiproliferative effects produced by the combined treatment with vinblastine and rapamycin.
Vinblastine and rapamycin lead to an induction of cell death in neuroblastoma cells. Next, we investigated the effects of vinblastine and rapamycin on SH-SY5Y and GI-LI-N cells, through the evaluation of cell death induction. Neuroblastoma cells were treated for 24 h with either 1 nmol/L (GI-LI-N) or 2 nmol/L vinblastine (SH-SY5Y) and 10 nmol/L rapamycin, delivered alone or in combination. At the end of the treatment, cells were collected and analyzed for both cell viability (trypan blue assay) and apoptosis (phosphatidylserine detection assay).
As shown in Fig. 3A and B, rapamycin produced an induction of cell death, with respect to control cells, even if not statistically significant. On the other hand, vinblastine treatment resulted in a statistically significant induction of cell death, on both cell lines examined (GI-LI-N, A; SH-SY5Y, B), with a further significant increase of trypan blue–positive cells caused by the combined treatment (**, P < 0.01, vinblastine + rapamycin versus vinblastine; ***, P < 0.001, vinblastine + rapamycin versus rapamycin).
In subsequent experiments, we investigated whether the observed cell death could be due to apoptosis induction. Phosphatidylserine exposure detection assays showed that, similarly to the results obtained in the aforementioned cell viability experiments, treatment with rapamycin had a slight proapoptotic effect. In contrast, after vinblastine treatment, we observed a statistically significant increase in the percentage of Annexin-positive cells that was significantly enhanced by the two drugs used in combination (Fig. 3C and D: *, P < 0.05, vinblastine + rapamycin versus vinblastine; ***, P < 0.001, vinblastine + rapamycin versus rapamycin). These results suggest that the vinblastine and rapamycin induced cell death could be due to apoptosis.
Combined therapeutic effects of vinblastine and rapamycin in human neuroblastoma orthotopic mouse model. To investigate whether the combined treatment with vinblastine and rapamycin could have antitumor effectiveness against human neuroblastoma in vivo, we settled an aggressive and highly angiogenic orthotopic mouse model by injecting GI-LI-N cells in the adrenal gland of nude mice. Different criteria were adopted to evaluate the therapeutic efficacy of our treatments. First, we assessed the antitumor effect produced by both the drugs, used alone or in combination, in terms of inhibition of tumor growth. As reported in Fig. 4A, at 30 days after the beginning of the treatment, mice treated with vinblastine and rapamycin, given alone, had statistically significant smaller primary tumor masses than control mice [mean tumor volume of vinblastine-treated versus untreated group: 503 versus 2,200 mm3 (difference, 1,697 mm3; 95% CI, 1,344-2,050 mm3), P < 0.001; mean tumor volume of rapamycin-treated versus untreated group: 553 versus 2,200 mm3 (difference, 1,647 mm3; 95% CI, 1,294-2,000 mm3), P < 0.001]. Noteworthy, mice treated with the combination of the two drugs had even smaller tumor masses [mean tumor volume of vinblastine + rapamycin-treated versus untreated group: 218 versus 2,200 mm3 (difference, 1,982 mm3; 95% CI, 1,629-2,335 mm3), P < 0.001]. Moreover, at 50 days after start of treatment, when all of control mice had died, the volumes of primary tumor masses from mice treated with the drug combination were statistically smaller than those of tumors from mice treated with either drug administered separately [mean tumor volume of vinblastine + rapamycin-treated versus rapamycin-treated group: 280 versus 810 mm3 (difference, 530 mm3; 95% CI, 323.7-736.3 mm3), P < 0.001; mean tumor volume of vinblastine + rapamycin-treated versus vinblastine-treated group: 280 versus 980 mm3 (difference, 700 mm3; 95% CI, 493.7-906.3 mm3), P < 0.001]. A further evidence, for the efficacy of our treatments, came from the count of CD31-positive blood vessels within primary tumor masses of treated mice. Figure 4B shows that vinblastine treatment led to a statistically significant decrease of CD31-positive blood vessels, with respect to control; this reduction was even more significant in tumor masses derived from mice treated with rapamycin. Finally, the combined treatment caused the maximum decrease of blood vessel count that was statistically different from the single drugs (**, P < 0.01, vinblastine + rapamycin versus rapamycin; ***, P < 0.001, vinblastine + rapamycin versus vinblastine).
As a second criterion to evaluate the therapeutic effectiveness of our treatments, we used the survival time of treated mice. Both the drugs administered alone triggered to a statistically significant increase in life span, with respect to saline-injected control animals (vinblastine versus CTR, P = 0.04; rapamycin versus CTR, P = 0.0015). Moreover, the combined treatment led to a superior increase in life span that was statistically significant with respect to control mice (vinblastine + rapamycin versus CTR, P = 0.0002) and, noteworthy, statistically significant with respect to either vinblastine- or rapamycin-treated mice (vinblastine + rapamycin versus vinblastine, P = 0.002; vinblastine + rapamycin versus rapamycin, P = 0.0162).
Vinblastine and rapamycin, administered alone or in combination, gave only minimal side effects, as shown by recording of mice body weight (in grams) plotted in Fig. 4D.
Combination therapy with vinblastine and rapamycin increases anti-vascular effects, tumor growth inhibition, and apoptosis in vivo. Once shown the therapeutic efficacy of the combined treatment against human neuroblastoma in vivo, we decided to elucidate the mechanisms responsible of the aforementioned effects. With this purpose, histologic and immunofluorescence experiments on paraffin-embedded tissue sections derived from orthotopic tumor-bearing mice have been done. As shown in Fig. 5, treatment with either vinblastine or rapamycin led to a destruction of tumor vasculature, evidenced by the lack of CD31 staining that was more evident in tissue sections derived from rapamycin-treated mice. Moreover, the combined treatment resulted in an almost complete loss of tumor blood vessels, in agreement with the results presented in Fig. 4B. The effect of our treatments was also very important in terms of inhibition of cell proliferation, as assessed by the detection of Ki-67 antigen (Fig. 5). In this regard, tissue sections derived from mice treated with vinblastine and rapamycin, given alone, showed a marked inhibition of cell growth, with respect to control; the combined treatment showed once again the significant capability to almost completely arrest tumor cell proliferation. Indeed, the percentage of Ki-67–positive cells was also quantified in four random fields from two independent experiments: control, 87.5 ± 9.5% Ki-67+ cells; vinblastine, 50 ± 7.5% (P < 0.001 versus control); rapamycin, 37.5 ± 5% (P < 0.001 versus control); vinblastine + rapamycin, 5.5 ± 1.5% (P < 0.001 versus control, P < 0.001 versus vinblastine, P < 0.01 versus rapamycin). Next, we investigated the potential of our combined treatment in terms of apoptosis induction. As shown in Fig. 5, tumors derived from mice treated with the combination of the two drugs presented a pronounced number of apoptotic cells, both neuroblastoma and endothelial cells, as assessed by the double staining with TUNEL plus NB84a and TUNEL plus CD34, respectively. These apoptotic features were clearly evident with respect to untreated mice and, statistically relevant, also compared with tumors derived from either vinblastine- or rapamycin-treated mice individually. Indeed, the percentage of TUNEL+-NB84a+ cells was quantified in four random fields from two independent experiments: control, 1 ± 1% TUNEL+-NB84a+ cells; vinblastine, 31.5 ± 5% (P < 0.001 versus control); rapamycin, 12 ± 3%; vinblastine + rapamycin, 51 ± 7.5% (P < 0.001 versus control, P < 0.01 versus vinblastine, P < 0.001 versus rapamycin). Finally, the percentage of TUNEL+-CD34+ cells was quantified in four random fields from two independent experiments: control, 1 ± 1% TUNEL+-CD34+ cells; vinblastine, 30 ± 4.5% (P < 0.001 versus control); rapamycin, 18 ± 3.5% (P < 0.01 versus control); vinblastine + rapamycin, 44 ± 5% (P < 0.001 versus control, P < 0.01 versus vinblastine, P < 0.001 versus rapamycin). Noteworthy, vinblastine + rapamycin–induced apoptosis was detected in tumor tissues but not in normal tissues, such as heart, lung, kidneys, liver, and spleen (data not shown).
Finally, we did histologic studies after only 1 week from the beginning of treatment, when the sizes of the tumors in the different branches were similar. The results presented in Supplementary Fig. S2 are superimposable to those presented in Fig. 5.
Combined administration of vinblastine and rapamycin down-modulate VEGF and VEGF-R2 in vitro and in vivo. Because we showed that part of the observed therapeutic effects, obtained through the combined treatment, depend, at least in part, on an antiangiogenic activity, we did experiments to investigate whether our treatment could also affect, in vitro and in vivo, the production of VEGF molecule, as well as modulate the expression of its receptor KDR (VEGF-R2), being the main proangiogenic loop involved in neuroblastoma-induced angiogenesis (13, 34, 35).
To this purpose, we determined the physiologic production of VEGF on a panel of neuroblastoma cells (SH-SY5Y, HTLA-230, GI-LI-N, and ACN). Although all the neuroblastoma cells analyzed produced VEGF, even if at different levels (data not shown), we focused our attention on GI-LI-N cells, being those chosen for the orthotopic in vivo model. Figure 6A shows that either vinblastine or rapamycin down-modulated significantly the production of VEGF, compared with control cells (2-fold reduction). Furthermore, the combined treatment caused a more evident decrease in the secretion of VEGF (roughly, a 4-fold decrease), with respect to control cells. To rule out the possibility that the decreased VEGF production could simply be a reflection of cytotoxicity of vinblastine and rapamycin, the raw data were normalized for the number of viable cells. Moreover, the apoptogenic drug Fenretinide was not able to consistently modulate VEGF production and secretion, confirming our previous findings (ref. 27; data not shown). In regard to the expression of VEGF-R2, two of four neuroblastoma cell lines analyzed (GI-LI-N and SH-SY5Y) expressed the receptor. Also in this case, we focused on GI-LI-N cells. As shown in Fig. 6B, either vinblastine or rapamycin treatment led to a statistically significant down-modulation of VEGF-R2 expression, compared with control cells. Noteworthy, the down-modulation became more statistically significant, reaching a complete switching off of receptor expression, when the cells were exposed to the combination of the two drugs. The relevance of these results has been further confirmed by in vivo assaying the effects of the combined treatment on VEGF and VEGF-R2 expression. In this case, tissue sections of primary tumor masses, derived from untreated or treated mice, were excised on day 25 from the beginning of the treatment and then stained for both VEGF (Fig. 6C) and VEGF-R2 (Fig. 6D) expression. As shown in Fig. 6C and D, either vinblastine or rapamycin induced a decrease of VEGF and VEGF-R2, more marked in primary masses excised from rapamycin-treated animals. Again, the simultaneous administration of both the drugs resulted in a stronger effect, leading to an almost complete disappearance of the aforementioned proangiogenic molecules. Indeed, the percentage of VEGF-positive cells were quantified in four random fields from two independent experiments: control, 58 ± 7.5% VEGF+ cells; vinblastine, 44 ± 6% (not significant versus control); rapamycin, 25 ± 4.5% (P < 0.01 versus control); vinblastine + rapamycin, 12.5 ± 2.5% (P < 0.001 versus control, P < 0.001 versus vinblastine, P < 0.01 versus rapamycin). Moreover, the percentage of VEGF-R2–positive cells were quantified in four random fields from two independent experiments: control, 55 ± 7.5% VEGF-R2+ cells; vinblastine, 35 ± 5.5% (P < 0.05 versus control); rapamycin, 18 ± 3% (P < 0.001 versus control); vinblastine + rapamycin, 3 ± 0.5% (P < 0.001 versus control, P < 0.001 versus vinblastine, P < 0.01 versus rapamycin).
To support the in vivo relevance of these findings, relative quantitative reverse transcription-PCR analysis has been also done on tumor samples derived from orthotopic-engrafted animals, untreated and treated with the various drugs for 1 week. In accordance to the in vitro (see Fig. 6A and B) and in vivo findings (Fig. 6C and D), both VEGF and VEGF-R2 expression levels resulted to be down-regulated by either the single drugs or the combination of vinblastine + rapamycin (Supplementary Fig. S3).
Finally, to further rule out that the antitumor efficacy was only due to the effect on tumor vasculature, we did double staining also on in vivo VEGF and VEGF-R2 expression regulation. As clearly observed in Supplementary Fig. S4, both molecules were co-expressed on both endothelial and neuroblastoma cells and were down-regulated by the combination of vinblastine + rapamycin in both cells.
Taken together, these results seem to suggest that the therapeutic efficacy of the combined treatment could be explained through a double mechanism of action: the first one that directly affects tumor cells and the second one that causes damage to endothelial tumor cells, with consequent effects on cancer cells.
Combination therapy with vinblastine and rapamycin inhibits angiogenesis induced in vivo by human neuroblastoma biopsy specimens. Chick embryo chorioallantoic membrane assays were done to further assess the effectiveness of vinblastine and rapamycin in inhibiting angiogenesis in vivo. The morphometric evaluations of microvessel area were done on day 12 of incubation with the sponges or the neuroblastoma biopsies. Microscopic examination of the microvessel area of chorioallantoic membranes revealed highly vascularized tissue among the trabeculae of the sponges loaded with the positive control fibroblast growth factor-2 (3.30 × 10−2 mm2; 95% CI, 2.5-4.4 × 10−2 mm2) and those loaded with human neuroblastoma biopsy specimens alone (3.1 × 10−2 mm2; 95% CI, 2.65-3.55 × 10−2 mm2). Addition of vinblastine significantly reduced the microvessel area (1.75 × 10−2 mm2; 95% CI, 1.45-2.1 × 10−2 mm2; P < 0.01 versus biopsy specimens alone). Similar results were obtained even after addition of rapamycin (1.55 × 10−2 mm2; 95% CI, 1.35-1.75 × 10−2 mm2; P < 0.01 versus biopsy specimens alone). The two drugs used in combination displayed a significantly greater angiostatic activity (0.77 × 10−2 mm2; 95% CI, 0.65-0.9 × 10−2 mm2; P < 0.001 versus biopsy specimens alone) then the single drugs (P < 0.05 versus vinblastine; P < 0.05 versus rapamycin). These results further confirm the implication of an antiangiogenic activity, concordant with our previous data (16), in the therapeutic efficacy of vinblastine and rapamycin.
In this study, we have shown for the first time that vinblastine and rapamycin, used in combination, exhibit a greater antitumor effectiveness with respect to either drug used separately, in vitro against a panel of human neuroblastoma cell lines, and in vivo against a biologically and clinically relevant, highly angiogenic neuroblastoma mouse model.
One of the major goals in cancer research is to develop new therapeutic approaches having limited side effects and good efficacy. This aim is terribly pregnant in neuroblastoma research because, in spite of the application of aggressive treatment strategies, the great outcome for this disease is quite poor (2). The rationale to use vinblastine and rapamycin in combination stems from three major reasons. First, vinblastine and rapamycin act by affecting cell cycle progression in different phases (4, 10); thus, the effects produced when used alone would, reasonably, lead to a greater result if used together. Second, it has been reported that the mammalian target of rapamycin inhibitor rapamycin sensitizes several tumor cells to chemotherapeutics (36, 37). Third, vinblastine and rapamycin are well-known drugs already used in clinical practice; thus, a new approach based on the combination of the two molecules could be easily transferred to clinical therapy.
Recently, we showed that vinblastine and rapamycin exerted synergistic inhibition of human neuroblastoma-related angiogenesis (16). Here, we found that combination of vinblastine and rapamycin resulted in a marked inhibition of neuroblastoma and endothelial cell growth, leading to a drastic antiproliferative result. The inhibition of cell proliferation induced by these two drugs has been associated to an impaired cell cycle progression in other tumor models (38, 39). Accordingly, our findings showed that also in neuroblastoma cells vinblastine and rapamycin, used together, triggered a dramatic G2-M phase cell cycle arrest, coupled to an almost complete depletion of cells actively synthesizing DNA. This result suggests that the effect obtained by the single administration of vinblastine (G2-M phase block) is likely increased through the addition of rapamycin, according to previous studies (40, 41) in which rapamycin was shown to enhance the efficacy of several chemotherapeutics.
Vinblastine and rapamycin used in combination led to a stronger cell death induction compared with that obtained by either drug tested alone. Furthermore, our data suggested that the observed cell death could be due to apoptosis, as already shown for the activity of vinblastine in other tumor cells (42). However, although our findings suggest that apoptosis activation has a role in vinblastine plus rapamycin triggered cell death, the upstream signals that lie between drug exposure and programmed cell death are still unknown as is whether mechanisms of cell death other than apoptosis, such as mitotic catastrophe and autophagy, are involved in the function of vinblastine plus rapamycin (43, 44). Nevertheless, our recent data (17) showed that the combined treatment of endothelial cells with vinblastine and rapamycin led to the modulation of molecules involved in the apoptotic pathway.
We observed that vinblastine and rapamycin, given in combination, inhibited neuroblastoma tumor growth and angiogenesis, leading to a statistically significant increased life span in mice orthotopically implanted with a highly angiogenic and MYCN-amplified human neuroblastoma cell line (35, 45). This model is biologically and clinically relevant, mimicking a condition of disease with primary tumor and disseminated metastases, similarly to our already described neuroblastoma orthotopic model (29) obtained by injecting the MYCN single copy neuroblastoma cell line SH-SY5Y, even if with a more aggressive metastatic spreading. The MYCN-amplified GI-LI-N cells have been chosen because of their highly angiogenic features; indeed, as previously reported by us (35), GI-LI-N cells present angiogenic behavior in vitro and in vivo. Furthermore, we have also shown that in vivo angiogenic activity of neuroblastoma correlates with MYCN oncogene overexpression (45). The aggressiveness and the metastatic spreading as well as the highly angiogenic properties of our orthotopic neuroblastoma mouse model gave us the opportunity to elucidate the mechanisms of action that underlie to the obtained therapeutic results. Immunohistochemical analyses of orthotopic neuroblastoma tumors showed a statistically significant decrease in microvessel density and cell proliferation, associated to an increase in apoptotic neuroblastoma and endothelial cells. The above data could be consistent with the hypothesis that the therapeutic efficacy of the combined treatment is due, at least in part, to an antiangiogenic effect, consistently with our previous findings (16). Through these results, we can postulate that the combined therapy works by a double mechanism of action. The first one that damages directly tumor cells by inhibiting tumor cell proliferation and inducing apoptosis; the second one that produces injuries to endothelial cells, with consequent effects on cancer cells. This conclusion is consistent with the concept that a strategy affecting both tumor vasculature and tumor cells themselves may be more efficacious (30, 46, 47) than approaches directed against only one of the two therapeutic targets.
Because we showed that antiangiogenesis had an important role in our therapeutic results, we investigated the potential effect of the combined treatment in the modulation of VEGF and its KDR receptor (VEGF-R2). Indeed, VEGF plays a critical role in tumor angiogenesis and is expressed in a number of human tumors (48, 49), including neuroblastoma (50). It has also been envisaged the possibility that VEGF could take part to neuroblastoma progression by directly regulating neuroblastoma cell growth (51). Interestingly, in neuroblastoma tumors, VEGF, as well as VEGF-R2, expression is associated with poor prognosis (34, 52). Here, we observed that the administration of vinblastine and rapamycin in combination produced a statistically significant decrease in the secretion of VEGF molecule, as well as VEGF-R2 expression, improving the effects obtained by either drug used as monotherapy. Concordantly, it has been already shown that the functional block of VEGF-R2 and, hence, of VEGF itself in neuroblastoma xenograft models led to therapeutic effectiveness (53). Moreover, Stephan et al. showed that the combination of rapamycin and anti-VEGF antibodies inhibited primary and metastatic tumor growth in an orthotopic model of human pancreatic cancer (46).
The therapeutic relevance of the antiangiogenic aspect has been further underlined through the effectiveness of vinblastine and rapamycin in inhibiting angiogenesis induced in vivo by human neuroblastoma biopsy specimens grafted to chick embryo chorioallantoic membrane. The angiostatic activity was greatly increased by the combined therapy, with respect to monotherapy, in agreement with our previous data (16).
In conclusion, the association of vinblastine and rapamycin results in an increased antitumor effectiveness against neuroblastoma compared with either drug administered separately. The observed antitumor efficacy can be ascribed to inhibition of tumor cell proliferation and angiogenesis as well as apoptosis induction. These preclinical results can be relevant to design new therapeutic approaches against neuroblastoma.
Grant support: Associazione Italiana per la Ricerca sul Cancro, Fondazione Italiana per la lotta al Neuroblastoma, and Ministry of Health.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).
C. Brignole is a recipient of a Fondazione Italiana Ricerca Cancro fellowship. F. Pastorino, G. Pagnan, L. Longo, and P. Perri are recipients of a Fondazione Italiana per la Lotta al Neuroblastoma fellowship.
No sponsor(s) have a role in the design of the study; the collection, analysis, and interpretation of the data; the writing of the manuscript; and the decision to submit the manuscript for publication.
D. Marimpietri and C. Brignole contributed equally to this work and should be considered joint first authors.
We thank V. Pistoia for suggestions and revisions, F. Parodi for expert technical assistance, and C. Bernardini for editing.