Acetazolamide potentiates the anti-tumor potential of HDACi, MS-275, in neuroblastoma

Acetazolamide (AZ), cisplatin (CDDP), dimethyl sulfoxide (DMSO) and MS-275 were obtained from Sigma-Aldrich (Oakville, ON, Canada). Culture media, AMEM, DMEM/F12 and DMEM, and supplements, fetal bovine serum (FBS) and penicillin-streptomycin, were purchased from Gibco (Burlington, ON, Canada). Bovine serum albumin (BSA) was obtained from Invitrogen (Grand Island, NY, USA). Matrigel was purchased from BD Biosciences company (La Jolla, CA, USA). Methylcellulose was obtained from StemCell Technologies (Vancouver, BC, Canada) and phosphate buffered saline (PBS) from Multicell (St. Bruno, QC, Canada).

Cells were purchased from the American Type Culture Collection (ATCC) as follow: N-type, MYCN non-amplified SH-SY5Y (CRL-2266) and SK-N-SH (HTB-11) and MYCN amplified, SK-N-BE(2) (CRL-2271), and teratocarcinoma NT2/D1 (CRL-1973). Cells were cultured in AMEM (SH-SY5Y and SK-N-BE(2)) and DMEM/F12 (NT2/D1) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (Multicell, St. Bruno, Quebec) at 37 °C in a humidified atmosphere of 5% CO₂. Reference normal neuronal stem cell strains (NSC6539 and NSC6562) were kindly provided by Dr. Peter Dirks (SickKids) and maintained in Neurocult and Neuronal stem cell Expansion media-Human from Stemcell Technologies (Stemcell Technologies,Vancouver, BC, Canada) supplemented with 2 mM L-Glutamine, 75 μg/ml BSA, 10 ng/ml (B27, EGF and FGF), 2 μM/ml heparin and 1% penicillin/streptomycin (Multicell, St. Bruno, Quebec, Canada). The cells were grown on poly-L-ornithine and laminin (Sigma-Aldrich, Oakville, ON, Canada) coated plates at 37 °C and 5% CO₂. The cells were fed every 3–4 days.

Standard procedures were performed as described [17]. Briefly, cell viability was assessed by trypan blue exclusion assay by observing the number of trypan blue positive cells versus total cells counter per microscopic field.

Standard protocol was performed as described [17]. Percent survival vs. control (DMSO- 0.2x10⁻⁴μM) of cells when treated with AZ, MS-275 and AZ + MS-275 were observed using AlamarBlue agent (AbD Serotec, MorphoSys, Raleigh, NC, USA) agent (10% of total volume) was added to each well for 4 h before fluorometric detection. Fluorescence was measured using the SPECTRAmax Gemini Spectrophotometer (excitation 540 nm; emission 590 nm).

10⁵ cells were seeded into 96 well plates and treated for 48 h with 1.5 μM MS-275. Following treatment, cells were washed twice with PBS and fixed with 4% paraformaldehyde for 20 min. Cells were then permeabilized with 0.1% Triton-X 100 for 5 min and washed twice with PBS. Cells were blocked in 1x Odyssey blocking buffer (LI-COR, Guelph, ON, Canada) for 2 h at room temperature. Primary antibodies against p16 (1/50; Santa Cruz, Santa Cruz, CA, USA), p21 (1/20; Santa Cruz, Santa Cruz, CA, USA), p27 (1/20; Santa Cruz, Santa Cruz, CA, USA), BCL-2 (1/100; Cell Signaling Technology, Toronto, ON, Canada), cyclin D1 (1/10; Neomarkers/Lab Vision,ThermoScientific, Fremont, CA, USA), and CDK4 (1/100; Neomarkers/Lab Vision,ThermoScientific, Fremont, CA, USA) were diluted in Odyssey blocking buffer at indicated ratios and added to cells overnight at 4 °C. Pan-actin antibody (1/100; CEDARLANE, Burlington, Ontario, Canada) was also added in conjunction with the other antibodies to serve as a control of cell content. Cells were then washed with a 0.1% Tween 20 (Fisher Scientific Co, Markham, ON, Canada) solution for 5 min and repeated five times. Fluorescently labeled Li-COR secondary antibody (Goat-anti-Rabbit IRDye 680; LI-COR, Guelph, ON, Canada), (Goat-anti-Mouse IRDye 800; LI-COR, Guelph, ON, Canada) were then added at a dilution of 1/500 and cells treated for 1 h at RT. In wells where actin was not used as a cell content control, DRAQ5/Sapphire700 were added at 1 mM and 1/1000 dilution respectively. Cells were then again washed with 0.1% Tween 20 solution five times for 5 min. Cells were imaged on the Odyssey Infrared Imaging System at excitation of 700 nm and 800 nm. Total fluorescence was quantified and adjusted to cell content control of either actin or DRAQ5/Sapphire700.

Briefly, 2 × 10⁶ cells treated with AZ and/or MS-275 were lifted by citrate saline and fixed in 80% ice-cold ethanol for 48 h. Cells were then pelleted and re-suspended in 2 mg/mL RNase A (Sigma-Aldrich, Oakville, ON, Canada) for 5 min. A 0.1 mg/mL propidium iodide solution (Sigma-Aldrich, Oakville, ON, Canada) was added, incubated for 30 min at RT, and cells filtered through a cell-strainer into a 5 mL polystyrene tube. Labeled cells were analyzed on a BD FACSCAN flow cytometer. Data was fitted by the Watson-Pragmatic model on FlowJo Software (Tree Star, Ashland, OR, USA).

Standard protocol was performed as follows [17], cultures were trypsinized and single cells were suspended in methycellulose medium (Methocult; StemCell Technologies, Vancouver, BC, Canada). In this process, 1.2x10⁴ SH-SY5Y, 7.5 x 10³ SK-N-BE(2) and 2 × 10³ NT2/D1 cells/mL were placed into a 40% methycellulose solution supplemented with 10% FBS, 1% antibiotics and 49% culture medium. MS-275 concentrations ranging from 10nM to 3 μM were added to the methycellulose. Cells in methylcellulose were gently vortexed and distributed into non-adherent 35 mm tissue culture dishes with a blunt end 16 gauge needle. Samples were placed in a 37 °C incubator in 5% CO_2. After 2 weeks colonies were photographed and counted on a phase contrast microscope using a grading dish. Clonogenicity was determined as the average of number of colonies per dish for each group of cells.

Briefly, 10⁶ cells/ml were lifted with citrate saline (0.05 M) and incubated for 1.5 h with 5 μg/mL Hoechst 33342 (bisbenzimide trihydrochloride); (Sigma-Aldrich, Oakville, ON, Canada) in a 37 °C water bath. Negative controls were prepared by prior addition of 50 μM Verapamil HCl (Sigma-Aldrich, Oakville, ON, Canada), calcium channel blocker. Cells were washed, counterstained with 1 μg/mL propidium iodide (Sigma-Aldrich, Oakville, ON, Canada) and analyzed on a BD LSRII flow cytometric analyzer.

For flow cytometry, 10⁵ cells/ml were lifted with trypsin and were blocked in cold 5% BSA/PBS solution at 4 °C for 15 min. Next, cells were treated with anti-ABCG2 conjugated to phycoerythrin (R&D Systems, Mineapolis, MN, USA) for 45 min. After being washed three times with cold PBS, cells were resuspended in PBS solution containing 7-AAD (BD Pharmingen, San Jose, CA, USA) and then analyzed on a BD LSRII flow cytometric analyzer. Cells negative for 7-AAD were gated to exclude non-viable cells. Gating was determined from the negative trypsin controls.

Adherent cells were lifted by trypsin, washed and fixed with 4% paraformaldehyde (Canemco, St. Laurent, Quebec, Canada) in PBS. 3 × 10⁶ cells were permeabilized with 0.1% Triton-X in PBS, washed twice with PBS and blocked with 5% BSA/PBS solution for 1 h at RT. Cells were incubated overnight at 4 °C in primary antibody against OCT4 (1/200; Cell Signaling, Danvers, MA, USA), SOX2 (1/200; R&D Systems, Mineapolis, MN, USA) or Nanog (1/200; Cell Signaling Technology, Toronto, ON, Canada) in 5% BSA/PBS. Cells were subsequently washed three times with PBS and incubated with a chicken-anti-rabbit Alexa Fluor-488 (1/3500; Invitrogen, Carlsbad, CA, USA) or goat-anti-mouse R-Phycoerythrin (1/500; Caltag, Burlingame, CA, USA) secondary antibody in 5% BSA/PBS, for 1 h in room temperature. Cells were washed and analyzed on a BD FACSCAN flow cytometer.

Cells were grown on glass coverslips until 75% confluent and then treated with MS-275. Immunofluorescence was performed [24] with primary antibodies to OCT4 (1/300; Cell Signaling Technology, Toronto, ON, Canada), SOX2 (1/250; R&D Systems, Minneapolis, MN, USA) and Nanog (1/200; Cell Signaling Technology, Toronto, ON, Canada) followed by incubation in AlexaFluor secondary antibodies (Invitrogen, Grand Island, NY, USA), and mounting in PBS/glycerol.

Cells were lysed with RIPA extraction buffer (MBiotech, Seoul, Korea) supplemented with a CompleteMini protease inhibitor tablet (Roche, Indianopolis, IN, USA). 100ug of protein was loaded for SH-SY5Y lysates, and 20 μg for NT2/D1 lysates. OCT4, SOX2 (Cell Signaling Technology, Toronto, ON, Canada) and Nanog (Cell Signaling Technology, Toronto, ON, Canada) antibodies were used at 1/1000 dilution. Secondary horseradish peroxidase conjugated antibodies (Jackson Immunoresearch, West Grove, PA, USA) were used at a dilution of 1/6000 and signal was detected with the Supersignal chemiluminescence detection system (Pierce Biotechnology, Rockford, Il, USA).

SH-SY5Y cells were seeded in a 48-well plate on glass cover slips and allowed to adhere overnight at a density of 10⁵ cells/well in 500 μl culture medium in triplicate. Wells were marked with a straight black line on the bottom for orientation. At the time of 90% confluence, cell monolayers were scratched with a 200 μl pipette tip using the marker guide. Loosened non-adherent cells were washed off with medium. Fresh medium was added to the cultures with additions of AZ (10 μM, 20 μM, 40 μM) and MS-275 (0.75 μM, 1.5 μM and 3 μM) and cultured for 48 h. After the 48 h period cells were washed with PBS and fixed in 4% paraformaldehyde. After three washes in PBS, cells were stained with 1% Crystal violet in 20% methanol. Phase contrast light microscopic images (10x original magnification) were taken at time points of 0, 48 and 72 h of treatment. Migrated cells were counted manually to quantify numbers of cells migrated to wound area using NIH Image J program. Each experiment was conducted three times in triplicate and one representative assay is shown.

For the in vivo xenograft study, 4–6 weeks-old female NOD/SCID mice were obtained from the animal facility at The Hospital for Sick Children. The animal use protocols were approved by the Animal Care Committee, Sickkids Research Institute. Animals were treated per guidelines of Canadian Council on Animal Care (CCAC). Subcutaneous xenograft tumors were developed by injecting SH-SY5Y cells (2 × 10⁶) into the inguinal fat pad of NOD/SCID mice. When tumor diameter reached 0.5 cm, the mice were randomized into four groups (5 mice per group). The control and treatment groups received intraperitoneal injections of vehicle (PBS) or AZ (40 mg/kg), MS-275 (20 mg/kg) or the combination, respectively, every day for 2 weeks. Experiments were terminated when tumor sizes exceeded 2 cm³ in volume or animals showed signs of morbidity. Tumor diameters were measured on a daily basis until termination. The long (D) and short diameters (d) were measured with calipers. Tumor volume (cm³) was calculated as V = 0.5 × D × d². After euthanizing the mice, tumors were resected, weighed and fixed in 10% neutral-buffered formalin at room temperature and processed for histopathology. For the in vivo serial heterotransplantation analysis, 2x10⁶ untreated and pretreated AZ + MS-275 cells, manually and enzymatically dissociated from treated tumors, were injected subcutaneously to NOD/SCID mice. Growth rates were measured 2–3 times per week. On the 38th day, the animals were sacrificed, after which tumors were removed and weighed.

Tumor fragments were fixed in 4% formaldehyde and 1% glutaraldehyde in phosphate buffer, pH 7.4, and post fixed in 1% osmium tetroxide. Tumor tissues were then dehydrated in a graded series of acetone from 50 to 100% and subsequently infiltrated and embedded in Epon-Araldite epoxy resin. The processing steps from post fixation to polymerization of resin blocks were carried out in a microwave oven, Pelco BioWave 34770 (Pelco International, Clovis, CA, USA). Ultrathin sections were cut with a diamond knife on the Reichert Ultracut E (Leica Inc., Vienna, Austria). Uranyl acetate and lead citrate were used to stain the sections before being examined in the JEM-1011 (JEOL USA Inc., Peabody, MA, USA). Digital electron micrographs were acquired directly with a 1024 × 1024 pixels CCD camera system (AMT Corp., Danvers, MA, USA) attached to the ETM (1200 EX electron microscope).

Standard protocol was performed as described [17]. Immunohistochemistry (IHC) was performed on paraffin sections where slides underwent a series of deparaffinization and rehydration washes which were further processed for antigen retrieval and blockage of endogenous peroxidase activity. Sections were then incubated with secondary antibody broad-spectrum poly horseradish peroxidase, and incubation with DAB. The percentage of positive cells was calculated by using the formula [X (6 low power fields of positive staining)/Y(total count per 6 fields) × 100]. The level of IHC of the positive cells was also examined by ImageJ64 software.

The TUNEL assay was performed on 5 μm sections prepared from formalin-fixed, paraffin-embedded xenografts, using the In Situ Cell Death Detection Kit (Roche, Indianopolis, IN, USA) and the protocol suggested by the manufacturer, except that the positive control was treated with 500 units/ml DNaseI (Roche, Indianopolis, IN, USA) before adding the TUNEL reaction buffer. The peroxidase reaction was carried out with stable DAB solution (Invitrogen, Grand Island, NY, USA). Finally slides were counterstained with haematoxylin and examined under light microscopy.

The data are presented as mean +/− SD. Statistical analysis of variance was run based on triplicate experiments and performed with Graphpad Prism 5.0 software (Hearne Scientific Software, Chicago, IL, USA) using 2-tailed Student’s t-test or 2-tailed paired t-test. Asterisks denote significance (*) p ≤ 0.05; (**) p ≤ 0.01; (***) p ≤ 0.001. Coefficient of Drug Interaction (CDI) was used for the assessment of drug interaction (antagonistic, synergistic and additive). CDI was calculated by formula AB/AxB; where AB, A and B are the cytotoxicity ratio of the combination, AZ single agent and MS-275 single agent, respectively. Cytotoxicity ratio at a certain drug concentration is the ratio of %viability of cells at that concentration to % viability of untreated cells. CDI equal to 1, < 1 and > 1 is considered as additive, synergistic and antagonistic, respectively.

To determine the effect of AZ, MS-275 and AZ + MS-275 treatments on the growth of NB SH-SY5Y, SK-N-SH and SK-N-BE(2) cells, we used the AlamarBlue and trypan blue assays. As reference normal controls, we included the neuronal stem cell strains, NSC6539 and NSC6562, which represent neural lineage derived stem cells albeit from the central nervous system [25]. We chose clinically acceptable concentration ranges for AZ (0-160 μM) [17, 26] and MS-275 (0–3 μM) [7‐9, 27]. It should be noted that AlamarBlue also indicates effects on oxidative phosphorylation as it is a substrate for the last step in oxidative phosphorylation. Thus it can also reflect metabolic mitochondrial effects.

Figure 1a-c shows that both AZ and MS-275 had a more moderate concentration-dependent inhibitory effect on neuronal stem cells while the effects of MS-275 and the AZ + MS-275 combination were significantly enhanced on all three tumor lines. The reduction in cell viability and IC50 values of NSC6539, NSC6562, SH-SY5Y, SK-N-SH and SK-N-BE(2) by AZ, MS-275, and AZ + MS-275 (48 h) shows the highest percent reduction with AZ + MS-275 in achievable plasma concentration (Tables 1 and 2). A significant difference in IC50 values for SH-SY5Y, SK-N-SH and SK-N-BE(2), indicates the potentiation of MS-275 effect by AZ. Interestingly, CDI analysis for the combination of AZ and MS-275 on different cell lines reveal that the combination is antagonistic at all concentrations on NSC6539 cells (CDI = 1.14–1.42) and additive on NSC6562 cells at concentrations above 20 μM AZ (CDI = 1). On NB cell lines (SH-SY5Y, SK-N-SH and SK-N-BE(2)) the combination was found to be additive at 40 μM and 80 μM of AZ (CDI = 1) and synergistic at 160 μM of AZ (CDI < 1). Since SH-SY5Y showed moderate resistance to AZ and/or MS-275, average concentration-response and IC50 compared to the other two NB cell lines, we chose to focus on SH-SY5Y cell line for the rest of the study. The AlamarBlue assay results also raised the question of effects on apoptosis and cell cycle as a measure of possible growth arrest and/or toxicity.

We characterized the effect on apoptosis and cell cycle using a propidium iodide (PI) based FACS analysis. Using a dose less than the IC50 dose, AZ (40 μM vs. 45 μM), MS-275(1.5 μM vs. 2.36 μM) and AZ + MS-275(40 μM + 0.75 μM) it was found that AZ, MS-275 and AZ + MS-275 treatments increase entry of SH-SY5Y cells into SubG0-phase (0.6%, %57 and %61) with decrease into S-phase (13%, 9% and 4%) and G2/M-phase (6%, 2% and 3%), significantly following a 48 h dosage of AZ(40 μM), MS-275(1.5 μM) and AZ + MS-275(40 μM +0.75 μM), respectively (Fig. 2a-b). In addition, western blot analysis of cell cycle inhibitor, p21, shows that MS-275 and AZ + MS-275 treatments (48 h) cause of 4.5 and 5.5 induction of p21 (p < 0.01) of SH-SY5Y cells compare to control, respectively (Fig. 2c-d). Results would suggest that AZ, MS-275 and AZ + MS-275 treatments are inducing apoptosis and preventing entry into S and G2/M-phases. This was further characterized by in-cell western and western blot analysis of cell cycle checkpoints, demonstrating a 2.5 fold induction of p16 (85% ± 0.19%; p < 0.001), a 1.97 fold induction of p21 (50% ± 0.64%; p < 0.001), a 1.48 fold induction of p27 (34% ± 0.31%; p < 0.05), cyclin dependent kinase inhibitors (Cdki), and downregulation of CDK4 (33% ± 0.15%; p < 0.01) and cyclin D1 (31% ± 0.35%; p < 0.001) following a 48 h dosage of 1.5 μM MS-275, respectively (Fig. 2e-f). These results were confirmed by western blot analysis (Fig. 2 g-h). Furthermore, western blot analysis of the expression of the Ki67 proliferative marker showed a down regulation of Ki67 (Fig. 2 g-h). Ki67 is expressed at all points during the cell cycle, except in G0. It should be noted that inductions of Cdki p21 occurred at the lower concentration of MS-275 but decreased somewhat at the higher concentrations suggesting a more complex epigenetic regulatory mechanism. Overall Cdki inductions varied except for p16. We also conducted a PI cell cycle analysis to determine the percentage of sub-G1 cell.

Since cell death due to mechanical damage is also accounted for in sub-G1 cell analysis, we also assessed Annexin/7-AAD and cleaved-caspase 3 expressions by FACs. Annexin stains cells in the early stage of apoptosis, while positive stain for 7-AAD indicates loss of membrane integrity. Cells at end stage of apoptosis are characterized by double staining for Annexin and 7-AAD. The results of FACs analysis demonstrated that MS-275 induces both early (10.7%) and late stage apoptosis (7.66%). The parallel assay with a clinically used chemotherapeutic 1 μM Etoposide (the IC50 dose) showed 2.51% and 8.58% induction of early and late stage apoptosis, respectively (Fig. 3a and Table 3). In addition, using the apoptotic indicator, cleaved caspase 3, 0.75 μM and 1.5 μM MS-275 after 48 h treatment yielded significantly 1% and 3% expression of cleaved-caspase 3, respectively (p = 0.0003 and p < 0.0001 respectively as compared to control and p = 0.0001 when comparing doses) (Fig. 3b-c). The pro-apoptotic effect of MS-275 was further investigated by western blot analysis of BCL2 and BAX expression. The results demonstrated that BCL2/BAX ratio is reduced (19.6% of control) by 1.5 μM MS-275 treatment (48 h);(Fig. 3d), indicating a potent apoptotic effect. MS-275 increases the expression of apoptotic protein BAX, hence decreasing the BCL2/BAX ratio, and coordinate with decrease in survival. Western blot analysis also demonstrated that survivin, a potent inhibitor of apoptosis, is reduced in expression following MS-275 treatment (Fig. 3e). Here NT2/D1 is a neuronal subclone of the teratomas NT2 cell line used in comparison, and as it is more stem cell like.

Previous studies had shown that both AZ and MS-275 reduce migration capacity in bladder and liver cancer cells [27, 28]. In the current study, we show the reduction of migration capacity in AZ (40 μM), MS-275 (1.5 μM) and AZ + MS-275 (40 μM + 0.75 μM) treated NB SH-SY5Y cells in Table 4. In Fig. 4a it is evident that AZ alone has an obvious inhibitory effect, and greatly enhanced by MS-275 at 48 h. The difference between MS-275 alone and the combo treatment was statistically significant (p < 0.001) at both 48 h and 72 h. AZ therefore enhances the inhibitory effect of MS-275 on SH-SY5Y cell migration capacity (Fig. 4a-b). In this assay the cytotoxicity of the agents is demonstrated overtly because of the initial cell confluence and may predict subsequent in vivo effects.

A concentration of 24.5 mg/kg MS-275 reduced growth of NB KCNR cell line as orthotopic xenografts [8]. A 14-day treatment protocol with 40 mg/kg AZ and/or 20 mg/kg MS-275 was devised based on the observed IC50 results and previously used safe doses in other preclinical studies [8, 29]. The inhibition of tumor volume is presented in Table 5. Figure 5a grossly shows a dramatic reduction in tumor growth and volume after MS-275 (20 mg/kg) and significantly greater when AZ (40 mg/kg) was added to the treatment. Figure 5b shows the changes in histology suggesting phenotypic alterations while Fig. 5c and d graphically reflect large reductions of volumes and weights, over the 14 days treatment period. The extirpated tumors also revealed grossly that both MS-275 and AZ + MS-275 markedly reduced the hematogeneous appearance of the tumors suggesting significant loss of vascularization (Fig. 5a-b). The significant reductions in tumor growth and weight were greatest with AZ + MS-275 (Table 5). The significant anti-tumor growth potentiation effect of AZ on MS-275 suggested an additive effect for this combination. IHC results revealed a possible explanation for the dramatic reduction in tumor volumes in that there was a strong inhibition of angiogenesis as revealed with staining for the angiogenesis marker (CD31) (Fig. 5e). Compared to the untreated group, expression of the CD31 was most significantly reduced with the combination AZ + SFN (Table 5; Fig. 5f).

The initial histopathological assessment of the residual tumors (Fig. 5a-e) revealed reductions in cell density and size, presence of pyknotic nuclei and reduced nuclear size, most prominently in the case of AZ + MS-275. This suggested that apoptosis could account for cell loss. To further confirm the histological changes, we performed electron microscopy on the tumor xenografts. Ultrastructural analysis revealed cells with degradative cytoplasmic changes [8, 30, 31] and nuclear fragmentation (pyknotic cells) indicative of apoptosis in SH-SY5Y xenografts. AZ + MS-275 treated cells had the highest number of pyknotic cells (Table 6; Fig. 6a-b), further supported the histological finding that AZ potentiated the apoptotic effect of MS-275. Next, we assessed apoptosis with the TUNEL assay on the xenografts (Table 6). We found that AZ + MS-275 treated cells had the highest percent TUNEL positive cells (Fig. 6c-d), and significantly greater than MS-275 alone. To further confirm the apoptotic process, we performed IHC to study the effect of AZ, MS-275 and AZ + MS-275 treatments on expression of the apoptotic marker (cleaved PARP) in the SH-SY5Y xenografts. Compared to the untreated group, expression of the apoptotic marker (cleaved PARP) was significantly induced, versus untreated controls (Table 6; Fig. 6e-f). Thus, AZ showed an overt potentiation of the anti-tumor effect of MS-275 in the SH-SY5Y xenografts.

The results from the apoptosis assessment of treated xenografts suggested strong effects on tumor cell proliferation. We therefore performed IHC to study the effect of AZ, MS-275 and AZ + MS-275 treatments on expression of mitotic (phosphohistone-H3; pHH3) and proliferation (Ki-67) markers in the SH-SY5Y xenografts. Compared to the untreated group, expression of the mitotic marker (pHH3) was moderately reduced after AZ treatment alone, significantly after MS-275 treatment and further enhanced with the combination AZ + MS-275 (Table 7; Fig. 7a-b). In a similar manner expression of the proliferative marker, Ki67 was significantly reduced by all treatments and the most by AZ + MS-275 (Table 7; Fig. 7c-d). Thus, the large reductions in tumor growth produced by MS-275 and the AZ + MS-275 combination are additionally reflected in significant reductions in mitosis and proliferation paralleled by increased apoptosis. It is noteworthy that the potent inhibitory effect of AZ alone was revealed using these markers, and potentiation of MS-275 was further confirmed.

Given the remarkable reductions in vascularization it might be surmised that the tumors would experience enhanced hypoxia under the treatments and thereby increased hypoxia induced gene expression. However, we performed IHC on the xenografts and did observe that control untreated tumors significantly expressed HIF1-α and its downstream target CAIX (Table 7; Fig. 8a-d). We further found that the number of HIF1-α positive cells decreased significantly after all treatments and markedly after AZ + MS-275 (Table 8; Fig. 8a and b). Additionally, a markedly enhanced reduction in CAIX expression (membrane localization) after AZ + MS-275 (Table 8; Fig. 8c and d) paralleled that of HIF1-α. The major reduction in CAIX staining after AZ + MS-275 may reflect much more than a loss of viable cells supporting the concept of AZ potentiation of the epigenetic alterations in expression.

The ultimate test for loss of tumorigenic potential is whether treatments abrogate formation of tumors, accomplished by conducting serial heterotransplantation. To test this, after the initial round of treatments with the most effective AZ + MS-275 combination we serially heterotransplanted a second set of 3x10⁴ cells into NOD/SCID mice and in comparison to a similar number of untreated cells. We found that the combination of AZ + MS-275 caused a 69% ± 0.84% (p = 0.0002) inhibition in resulting tumor volumes and weights (81% ± 0.78%; p = 0.0001) respectively after 38 days (Fig. 9a-d). Note however that only 3/5 injections yielded any tumor mass. A marked phenotypic alteration in morphological appearance (Fig. 9b) resembled the treated tumors in Fig. 5a-e. These results suggest that the AZ + MS-275 combination might permanently alter cell phenotype and affect the presumptive CSCs in the tumor.

Our serial heterotransplantation results suggested that the CSC fraction could be targeted by a potentiated effect of AZ + MS-275. In parallel studies, we had examined the effect of MS-275 on another MYCN amplified NB cell line, SK-N-BE(2), and in comparison to the teratocarcinoma cell line, NT2 with a prevalent stem cell phenotype. MS-275 significantly inhibited clonogenic potential of SH-SY5Y and SK-N-BE(2) cells (Fig. 10a-c). Here we first established an effective concentration range for abrogating clonogenicity in methycellulose, the reduction in the presumptive tumor initiating SP cell fraction (Fig. 11a-b), compared MS-275 with TSA and found MS-275 to be more potent (Fig. 11c-d) while cisplatin (CDDP) was ineffective and reduction in expression of stem cell markers OCT4 and SOX2 by FACs and western blot analysis (Fig. 12a-h). These studies established the ability of MS-275 to reduce the tumor initiating potential of NB MYCN NB amplified cell lines. However, to expand upon these studies using immunophenotyping to determine the role of specific stemness markers (OCT4, SOX2, Nanog), we applied IHC to our SH-SY5Y xenografts after AZ, MS-275 and AZ + MS-275 treatments. We found that the AZ + MS-275 treatment produced the highest reduction in OCT4 expression, number of SOX2 positive cells, and Nanog immunopositive cells (Table 9; Fig. 13a-f). It is interesting to note that OCT4 and SOX2 expressions were most affected by AZ treatment, and relative to Nanog, proportionally greatest after AZ + MS-275 treatment, again supporting the idea of AZ potentiation of MS-275.