Cucurbitacin-I inhibits Aurora kinase A, Aurora kinase B and survivin, induces defects in cell cycle progression and promotes ABT-737-induced cell death in a caspase-independent manner in malignant human glioma cells

Abstract

Because STAT signaling is commonly activated in malignant gliomas as a result of constitutive EGFR activation, strategies for inhibiting the EGFR/JAK/STAT cascade are of significant interest. We, therefore, treated a panel of established glioma cell lines, including EGFR overexpressors, and primary cultures derived from patients diagnosed with glioblastoma with the JAK/STAT inhibitor cucurbitacin-I. Treatment with cucurbitacin-I depleted p-STAT3, p-STAT5, p-JAK1 and p-JAK2 levels, inhibited cell proliferation, and induced G2/M accumulation, DNA endoreduplication, and multipolar mitotic spindles. Longer exposure to cucurbitacin-I significantly reduced the number of viable cells and this decrease in viability was associated with cell death, as confirmed by an increase in the subG1 fraction. Our data also demonstrated that cucurbitacin-I strikingly downregulated Aurora kinase A, Aurora kinase B and survivin. We then searched for agents that exhibited a synergistic effect on cell death in combination with cucurbitacin-I. We found that cotreatment with cucurbitacin-I significantly increased Bcl^-2/Bcl^-xL family member antagonist ABT-737-induced cell death regardless of EGFR/PTEN/p53 status of malignant human glioma cell lines. Although >50% of the cucurbitacin-I plus ABT-737 treated cells were annexin V and propidium iodide positive, PARP cleavage or caspase activation was not observed. Pretreatment of z-VAD-fmk, a pan caspase inhibitor did not inhibit cell death, suggesting a caspase-independent mechanism of cell death. Genetic inhibition of Aurora kinase A or Aurora kinase B or survivin by RNA interference also sensitized glioma cells to ABT-737, suggesting a link between STAT activation and Aurora kinases in malignant gliomas.

Keywords:

• Aurora kinases
• caspase-independent cell death
• cell cycle arrest
• Glioma

Abbreviations

BSA	=	bovine serum albumin
DMSO	=	dimethyl sulfoxide
EGFR	=	epidermal growth factor receptor
MTS	=	3-[4, 5-dimethylthiazol- 2yl]-5-[3-carboxymethoxyphenyl]-2-[4-sulfophenyl]-2H, tetrazolium
FITC	=	fluorescein isothiocyanate
NF-кB	=	nuclear factor кB
PI3K	=	Phosphatidylinositol 3-Kinase
PBS	=	phosphate-buffered saline
PAGE	=	polyacrylamide gel electrophoresis
PDGFR	=	platelet derived growth factor receptor
TBS	=	Tris-buffered saline
TRAIL	=	tumor necrosis factor–related apoptosis inducing ligand
PI	=	propidium iodide

Introduction

Malignant gliomas are aggressive, highly invasive, morphologically heterogeneous tumors that have proven largely refractory to conventional treatment modalities such as surgery, irradiation, and cytotoxic chemotherapy. Despite recent research efforts in cancer therapy, the prognosis of patients with gliomas remains poor and their therapeutic resistance has not been fully understood.^1-3 These tumors arise from a sequential accumulation of genetic aberrations and dysregulation of growth-factor signaling pathways, encompassing mutations and amplification of epidermal growth factor receptors (EGFR), deletion of the phosphatase and tensin homolog on chromosome 10 (PTEN), TP53 mutations, and p16 deletion and, less frequently, MDM2 amplification. These alterations activate critical downstream mediators that contribute to the neoplastic phenotype, and may constitute promising therapeutic targets.^4,5

The epidermal growth factor receptor (EGFR) is frequently deregulated in malignant glioma, as a result of EGFR gene amplification, activating mutation, or both.^6-11 EGFR variant III (EGFRviii), a common EGFR mutation, results from an in-frame deletion of exons 2 to 7, strongly and persistently stimulates several key signaling pathways, including phosphatidylinositol 3-kinase (PI3K)/Akt, extracellular regulated kinase (ERK), signal transducer and activator of transcription 3 (STAT3), BCL^-xL, and nuclear factor ĸB (NF-ĸB).^12-19 Although both EGFR and EGFRviii are associated with tumor growth and progression and are, thus, a logical target for anticancer therapy, the efficacy of EGFR-targeted small-molecule inhibitors or monoclonal antibodies in gliomas has been modest.

In studies using a large-scale short interfering RNA (siRNA) screening approach to identify critical “nodes” for cell death signaling, we identified several targets, including nuclear factor kappa B (NF-ĸB) and the proteasome, as well as Akt and Bcl^-2 family members that, when inhibited, promoted apoptotic signaling in glioma cells.^20-24 Bcl^-2 antagonists can induce apoptosis as single agents only in cancer cells dependent on Bcl^-2 and/or Bcl^-xL for survival. As a single agent ABT-737, a novel Bcl^-2/Bcl^-xL inhibitor triggers apoptosis in various types of human cancers including multiple myeloma, leukemia, lymphoma and small cell lung cancer.^25-29 However, we^30-32 and others³³ have demonstrated that ABT-737 minimally inhibits cell growth in glioma when used as a single agent. The complexity of the interactions between cell surface receptors and downstream signaling targets has called into question the clinical utility of blocking any target in isolation, and the results of single agent- based strategies have to date been disappointing in patients with glioma.^2,4 Combination approaches can enhance ABT-737-induced glioma cell death, including use with the proteasomal inhibitor bortezomib,³⁴ the survivin inhibitor YM-155,³⁰ the PI3K/Akt inhibitor NVP-BKM120,³¹ the histone deacetylase inhibitor vorinostat,³⁵ the alkylating agent temozolomide,³⁶ or the death receptor ligand TRAIL.³⁷ We performed a screen of ABT-737 in combination with other signaling inhibitors used at clinically achievable doses; combination of ABT-737 and cucurbitacin-I promoted apoptosis in malignant human glioma cell lines.

Signal-transducer-and-activator-of-transcription-3 (STAT3) regulates the transcription of several genes that are involved in cell cycle progression, anti-apoptosis, cell survival, and angiogenesis. STAT3 is activated in 60% of primary high-grade/malignant gliomas³⁸ and closely linked with a variety of human malignancies, including leukemia,³⁹ head and neck squamous cell carcinoma,⁴⁰ and breast.⁴¹ Inhibition of JAK2/STAT3 signaling with cucurbitacin-I (JSI-124),⁴² a natural plant product, has been shown to have potent anti-cancer activities in hematological malignancies and solid tumors.^43-47 Several lines of evidence have indicated that cucurbitacin-I promotes cell cycle arrest and/or apoptosis in glioma,^48,49 lung,^42,50 breast,⁵¹ and other malignancies.^52-55 Previously we observed that blockade of the JAK/STAT pathway by cucurbitacin-I enhanced the efficacy of SRC family kinase (SFK) inhibition with dasatinib in glioma cells.⁴⁸ Chumbalkar et al performed a tyrosine-directed search for signaling events downstream of ΔEGFR (EGFRviii) and identified STAT5 phosphorylation at Y699 as a key event.⁵⁶ Later, Latha et al demonstrated that EGFRviii cooperated with STAT5 to regulate the Bcl^-xL promoter.⁵⁷ This prompted us to investigate the role of JAK/STAT inhibitor cucurbitacin-I and the Bcl^-2/Bcl^-xL antagonist ABT-737 in glioma cell lines. In this study, we report that cotreatment with cucurbitacin-I and ABT-737 induced G2/M arrest and apoptosis regardless of EGFR, PTEN, and TP53 mutation status in malignant human glioma cell lines.

Results

Cucurbitacin inhibits growth in EGFR-overexpressing glioma cell lines and primary cultured glioma cells but not non-neoplastic human astrocytes

In our previous study we have demonstrated that cucurbitacin can induce cytotoxicity in a variety of glioblastoma derived cell lines.⁴⁸ Because the epidermal growth factor receptor (EGFR) plays a prominent role in many tumors including glioblastoma and overexpression is observed in >50% of GBM patients,^58-60 we used isogenic U87 cell lines expressing EGFR-WT and EGFRviii (Fig. S1A and S1B). U87, U87-EGFR-WT and U87-EGFRviii cultures were exposed to varying concentrations of cucurbitacin and MTS cell proliferation assay was performed after 24, 48 and 72 h of treatment. Vehicle-treated (DMSO) cells served as control. Cucurbitacin inhibited cell proliferation in a dose and time-dependent manner (Fig. 1A). The cucurbitacin IC50 (72 h) was 385.1 nmol/L for U87, and 295 and 266.4 nmol/L for U87-EGFR (WT) and U87-EGFRviii respectively.Figure 1B presents the effect of cucurbitacin in primary cultured glioma cells (GBM-1, GBM-8, GBM-9 and GBM-13) established from cells collected from 4 patients with histologically confirmed glioblastoma. The non-neoplastic human astrocytes (HA) showed no growth inhibition within the dose range (>2.0 μmol/L) tested, suggesting that cucurbitacin acts selectively against tumor cells (Fig. 1C).

Figure 1. Anti-proliferative activity of cucurbitacin in glioma cell lines and non-neoplastic astrocytes. EGFR overexpressing human glioma cell lines U87-EGFR-WT, U87-EGFRviii and isogenic control, U87 (A), primary cultured glioma cells, GBM-1, GBM-8, GBM-9 and GBM-13 (B) and non-neoplastic human astrocytes (C) were grown on 96-well plates in growth medium and, after an overnight attachment period, were exposed to selected concentrations of inhibitor or vehicle (0, DMSO) for 24, 48, or 72 h (Day 1, Day 2 or Day 3). Cell proliferation inhibition was assessed semiquantitatively by spectrophotometric measurement of MTS bioreduction. Points represent the mean of 3 measurements carried out in triplicate.

Cucurbitacin inhibits Aurora kinase A, B and survivin

Previously we have shown that activated JAK1, JAK2, and STAT3 were inhibited with high-dose cucurbitacin in established glioma cell lines.⁴⁸ In this study, we evaluated the protein expression of the phosphorylated forms of Akt, Bcl^-2, XIAP as well as STAT target gene products such as Bcl^-xL, Mcl^-1, and cyclin D1 in the EGFR overexpressing U87-EGFRviii cell line. Modest or no inhibition of p-AKT, cyclin D1 and BCL^-xL was observed after 24 h (Fig. 2A). No significant difference in the amount of proteins was noted (ERK, p38, JNK, nuclear forms of NF-ĸB, data not shown). However, levels of phospho-Stat-3 (at Tyr705), phospho-Stat-5 (at Tyr694), phospho-JAK-1 (at Tyr1022/1023), phospho-JAK-2 (at Tyr1007/1008) were reduced by cucurbitacin (Fig. 2B) in U87 and U87-EGFRviii cell lines. Interestingly, nanomolar concentrations of cucurbitacin led to a substantial reduction of Aurora kinase A, Aurora kinase B, and survivin in U87, U87-EGFR-WT, U87-EGFRviii (Fig. 2C, upper panel), LN229, T98G, and LNZ308 cell lines (Fig. 2C, lower panel). Then we carried out real time PCR on total RNA isolated from glioma cell lines. As shown in Figure 2C, a significant decrease in the mRNA expression of Aurora kinase A (2.5–4.5 fold), Aurora kinase B (20–25 fold), and survivin (25–30 fold), was observed in U87 cells treated with cucurbitacin (250 nmol/L) for 24 h. Similar results were observed in U87-EGFRviii, LNZ308, A172 and LN229 cell lines (data not shown).

Figure 2. Cucurbitacin inhibits Aurora kinase A, Aurora kinase B and survivin. U87 and U87-EGFRviii (A and B) cells were treated with cucurbitacin (indicated concentration) for 24 h. (C), A panel of 4 glioma cell lines (U87, LN229, T98G and LNZ308) along with EGFR over expressing U87-EGFR-WT (U87-EGFR) and U87-EGFRviii cells were treated with cucurbitacin (indicated concentration) for 24 h. Cell extracts were subjected to Western blot analysis with indicated antibody. (D). U87 cells were treated with cucurbitacin (indicated concentration) for 24 h, after which total RNA was extracted and Aurora kinase A, Aurora kinase B and survivin mRNA levels were quantified by real time PCR as described in the Materials and Methods.

Cucurbitacin induces G2/M arrest in human glioma cell lines

Impaired regulation of Aurora kinase-A, Aurora kinase-B and survivin expression has been noted in proliferating cells to cause chromosomal abnormality and instability.⁶¹ Because preferential ablation of Aurora kinase A, Aurora kinase B or survivin by RNA interference^62-64 produces a mitotic arrest, we investigated whether cucurbitacin influences the cell cycle profile in glioma cell lines. U87, U87-EGFRviii and other established cell lines such as T98G. LN18, LN229, A172, and LNZ308 were cultured with the indicated concentration of cucurbitacin for 24 h and analyzed by flow cytometry. A representative DNA profile (Fig. 3A) and a histogram obtained from multiple FACS analyses are shown in Figure 3B. Compared to the control (DMSO treated cells), cells accumulated in the G2/M phase with cucurbitacin treatment in a dose-dependent manner irrespective of EGFR or TP53 or PTEN status (see Table 1 for genetic features of the glioma cell lines). This was accompanied by the reduction of cells in the G1 and S phases of the cell cycle. After prolonged drug incubation (48 h, Fig. 3C upper panel; 72 h, Fig. 3C lower panel), there was an increase in the sub-G1 phase of cells, typical for late stages of apoptosis.

Table 1. The genotypic features of the glioma cell lines used in this study

Cell line	p53-gene status	PTEN-gene status
1. LN18	Heterozygous, WT/Mut238	Wild type
2. U87	Homozygous, WT	Deleted
3. T98G	Homozygous, Mut237	Mutant
4. LN229	Heterozygous, WT/Mut164	Wild type
5. LNZ308	Heterozygous, deleted/translocated	Deleted
6. A172	Heterozygous, WT/WT	Deleted

Figure 3. Cucurbitacin induces G2/M arrest regardless of EGFR/PTEN/p53 status in malignant human glioma cell lines. U87, U87-EGFRviii, T98G, LN18, LNZ308, A172 and LN229 cells were seeded at 60% confluence, allowed to attach overnight, and treated with cucurbitacin (indicated concentration) for 24 h. Control cells received DMSO (0). Cell cycle analysis using propidium iodide (PI) staining was performed as described in the Materials and Methods Section. Representative results of 3 independent experiments are presented here (A). Histogram represents the average of 3 experiments (B). LN18, A172 and LN229 cells were seeded at 60% confluence, allowed to attach overnight, and treated with cucurbitacin (indicated concentration) for 48 or 72 h. Control cells received DMSO (0). Cell cycle analysis was performed as described in the Materials and Methods Section (C).

Cotreatment with cucurbitacin and ABT-737 potentiates glioma cell death in a caspase-independent manner

Because survivin ³⁰ signaling inhibited the clinical efficacy of ABT-737 and cucurbitacin affected survivin both at mRNA and protein levels (Fig. 2C and 2D), we hypothesized that cotreatment with cucurbitacin and ABT-737 may be a useful strategy for the treatment of glioma. To systematically compare the effect of ABT-737 and cucurbitacin, cells were treated with ABT-737 or cucurbitacin or the combination of both for 72 h and MTS assay was performed as described in the Materials and Methods. Cotreatment of ABT-737 and cucurbitacin was able to reduce proliferation of glioma cells significantly. Lower concentrations of ABT-737 and cucurbitacin sufficed to inhibit proliferation (significant reduction in U87, U87-EGFR-WT, U87-EGFRviii, T98G, LN18, and LNZ308 cell lines after 72 h, P < 0.005, t-test; Fig. S2). Cucurbitacin significantly increased the sensitivity of established and primary cultured glioma cells to ABT-737 treatment compared with cells treated with ABT-737 alone. Simultaneous treatment of ABT-737 plus cucurbitacin also had little or no effect on cell proliferation of non-neoplastic astrocytes, suggesting selectivity against glioma cells.

We next quantified the effects of the inhibitor combinations on apoptosis. U87, U87-EGFR-WT, U87-EGFRviii, A172, LN229 and LNZ308 cells treated with ABT-737 or cucurbitacin or the combination of both for 24 h were stained with annexin V and PI and analyzed by flow cytometry. Three experiments were performed in duplicate with similar results. A representative annexin V binding histogram (Fig. 4A and 4B) and a bar chart representing 3 independent experiments is shown in Figure 4C. Single agent ABT-737 or cucurbitacin resulted in only minimal or modest annexin V/PI staining. On the other hand, cotreatment with ABT-737 plus cucurbitacin enhanced annexin V/PI sensitivity (Fig. 4A–C). To understand whether the mechanism of cell death was caspase-dependent, an irreversible pan-caspase inhibitor (z-VAD-fmk) was used. U87, U87-EGFR-WT and U87-EGFRviii cells were treated with z-VAD-fmk, for 2 hours prior to treatment with cucurbitacin or ABT-737 or the combination of both for 20 h. Annexin V/PI analysis showed that cucurbitacin plus ABT-737-induced cell death was unaffected by the presence of z-VAD-fmk, suggesting a possible caspase-independent cell death pathway (Fig. 4D). We then used Western blot analysis to validate the results. LN18 cells treated with TRAIL (Tumor necrosis factor related apoptosis-inducing ligand) served as a positive control. In response to cucurbitacin or ABT-737 as a single agent or the combination of both, the 32- kDa procaspase-3 was not cleaved to a -p20, -p17 and –p12 kDa “active” form; nor were other caspase-processing events (Fig. 4E, F). Activation of caspase-3 (appearance of cleaved 19, 17, and 12 kDa fragments), caspase-7 (appearance of cleaved 35, 30, and 20 kDa fragments), caspase-8 (appearance of cleaved 43, 41, and 18 kDa fragments), caspase-9 (appearance of cleaved 37 and 17 kDa fragments) and PARP (appearance of 89 kDa fragment) was observed in LN18 cells treated with TRAIL (lane 11, Fig. 4E, F).

Figure 4. Cotreatment with cucurbitacin and ABT-737 potentiates glioma cell death in a caspase-independent manner. EGFR overexpressing human glioma cell lines U87-EGFR-WT, U87-EGFRviii and isogenic control, U87 (A), A172, LN229 and LNZ308 (B) were seeded at 60% confluence, allowed to attach overnight, and treated with cucurbitacin (C) or ABT-737 (A) or the combination of both (C + A) for 24 h. Control cells received an equivalent amount of DMSO. Apoptosis was analyzed by flow cytometry. The percentages of cells in each quadrant are indicated. (C) Bar chart represents data from 3 independent experiments (** P < 0.005 vs. untreated control or single agent). (D) U87, U87-EGFR-WT and U87-EGFRviii cells were pretreated with pan-caspase inhibitor zVAD-fmk (25 μmol/L) for 2 h followed by combination of cucurbitacin plus ABT-737 (C + A) for 24 h. Cell death was analyzed by flow cytometry. Data are representative of triplicate studies from 3 independent experiments (NS, not significant). U87 and U87-EGFRviii (E) LN229, A172 and LNZ308 (F) were seeded at 60% confluence, allowed to attach overnight, and treated with cucurbitacin (indicated concentration) with or without ABT-737 (500 nmol/L) for 24 h. Control cells received an equivalent amount of DMSO. Cell extracts were subjected to Western blot analysis with indicated antibody. LN18 treated with TRAIL (25 ng/ml for 3 h) served as positive control. G. U87, U87-EGFR-WT, U87-EGFRviii and LNZ308 cells were grown on chamber slides in growth medium, and, after an overnight attachment period, were exposed to cucurbitacin (C, 100 nmol/L) or ABT-737 (A, 1.0 μmol/L) or the combination of both (C + A) or vehicle (DMSO) for 72 h. Cells were washed once with PBS, fixed with 3.7% formaldehyde for 30 min and stained with DAPI for 2 h at room temperature. The slides were then washed in PBS, mounted, and examined under a fluorescent microscope. Cells with multi-lobed nuclei are presented in the bar graph. Points represent the mean of 3 measurements ± S.D (**, P < 0.005).

Because Aurora kinase A, Aurora kinase B and survivin are essential for successful mitotic transition, particularly in cytokinesis,^65-68 and mitotic catastrophe is characterized by the appearance of enlarged micro- and/or multinucleated cells, we examined the morphology of nuclei of cucurbitacin and ABT-737-treated cells. At least 100 cells were analyzed and the results are presented in Figure 4G. Staining of cucurbitacin and ABT-737-treated U87, U87-EGFR-WT, U87-EGFRviii and LNZ308 cells with DAPI revealed the presence of fragmented or lobulated nuclei in association with micronuclei, characteristics consistent with mitotic catastrophe. Cucurbitacin induced the accumulation of cells with multi-lobed nuclei, whereas multinucleated cells were rarely or minimally detected in the control or ABT-737 treated groups. Seventy-two hours after treatment with cucurbitacin, about 20% of the EGFR overexpressing U87 cells showed multi-nucleated cells. The fraction of multinucleated cells was significantly increased with the combination of cucurbitacin and ABT-737 (Fig. 4G).

Targeting Aurora kinase A, Aurora kinase B and survivin by RNA interference enhanced ABT-737-induced cytotoxicity

To determine the effects of Aurora kinase A, Aurora kinase B and survivin on ABT-737-induced cytotoxicity, U87 cells were transfected with the expression vector containing shRNA targeted against Aurora kinase A, Aurora kinase B or survivin. Western blot analysis was used to evaluate the levels of Aurora kinase A, Aurora kinase B and survivin protein expression. As shown in Figure 5 (left panel), abundant levels of Aurora kinase A, Aurora kinase B and survivin expression were seen in non-target shRNA transfected control U87 cells, whereas genetic interference with shRNA inhibited >50% of the respective proteins. To determine whether the inhibition of Aurora kinase A, Aurora kinase B or survivin expression results in enhanced sensitivity, cells were treated with varying doses of ABT-737. Annexin V/PI analysis clearly demonstrated that the suppression of Aurora kinase A, Aurora kinase B or survivin expression enhanced sensitivity to ABT-737 (Fig. 5 right panel).

Figure 5. Targeting Aurora kinase A, Aurora kinase B and survivin by RNA interference enhanced ABT-737-induced cytotoxicity. U87 cells were transfected with non-target or Aurora kinase A, Aurora kinase B or survivin shRNA as described in the Materials and Methods. Forty-eight hours post-transfection, cells were treated with the indicated concentrations of ABT-737 for 24 h, and viability was assessed by annexin V/PI apoptosis assay (right panel; ** P < 0.005 versus non-target shRNA control). In parallel, cell lysates were collected and protein was subjected to Western blot analysis using indicated antibodies. Immunoblots were stripped and reprobed with β-actin (left panel). Data are representative of triplicate studies from 3 independent experiments (**, P <0.005).

Discussion

STAT signals downstream of EGFR are activated in 60% of glioblastoma patients and correlates inversely with survival. The EGFR/JAK/STAT pathway has generated significant interest as a key driver of tumor cell survival, proliferation, and invasion in GBM.³⁸ Recently, using subcellular fractionation, Fan, et al ⁶⁹ demonstrated that receptor phosphorylation of EGFRviii led to nuclear transport of EGFRviii and enhanced the formation of a complex between EGFRviii and STAT5. This complex serves as a transcriptional co-activator for a series of tumor-promoting genes such as cyclin D1, iNOS, B-myb, Aurora kinase and Cox-2.^57,70 In this report we evaluated the effect of the JAK/STAT inhibitor, cucurbitacin on malignant human glioma cells in vitro. Application of cucurbitacin to glioma cell lines inhibited cell proliferation and induced cell-cycle arrest regardless of mutational status of p53, PTEN and EGFR. We observed that cucurbitacin diminished Aurora kinase A, Aurora kinase B and survivin expression. To the best of our knowledge, these observations have not been previously reported. In light of this observation, it is possible that cucurbitacin-induced STAT5 down regulation might reduce the transcription of Aurora kinases.

Increasing evidence indicates that aberrant expression of Aurora kinase A, Aurora kinase B and survivin results in chromosomal instability, aneuploidy and malignant transformation. Aurora kinase A encodes a protein that localizes to the spindle poles, contributes to separation and maturation of the centrosomes during mitosis, plays a pivotal role in G2/M cell cycle progression and is frequently amplified in several types of cancer.^71,72 Aurora kinase B encodes a chromosomal passenger protein that is required for proper mitosis and cytokinesis.⁷³ Here, we have demonstrated that glioma cells treated with cucurbitacin develop spindle abnormalities and multinucleated cells at a high frequency, consistent with the results of previous reports on Aurora kinase A, Aurora kinase B and survivin inhibition in a variety of experimental systems. Thus, it is plausible that the mechanism by which Aurora kinase A, Aurora kinase B and survivin inhibition leads to cell death is related to its ability to induce polyploidy.^72,74

In our initial screen of cucurbitacin with other signaling inhibitors; simultaneous treatment of cucurbitacin with ABT-737 at clinically achievable doses in vitro resulted in a synergistic interaction. However, we were unable to detect the active forms of caspases-3, -7, -8 -9 and PARP in cucurbitacin and ABT-737 treated cell lines. Pan-caspase inhibitor zVAD-fmk did not block cucurbitacin and ABT-737-induced cell death. This could be due to the involvement of a number of non-caspase proteases (such as lysosomal cathepsins) that are capable of executing cell death. It has been suggested that mitotic catastrophe is not a genuine cell death executioner mechanism, but a pathway that precedes cell death, in which caspase activation and other markers of classical apoptosis are completely absent.⁷⁵ Our study clearly demonstrates that the combination of ABT-737 and cucurbitacin promoted mitotic spindle failure, G2/M arrest which was accompanied by an increase in the sub-G1 fraction of cells, and eventual cell death. Genetic knockdown of Aurora kinase A, aurora kinase B or survivin dramatically increased ABT-737 induced glioma cell death. This is in agreement with our previous report in which we demonstrated that cotreatment with the survivin inhibitor YM-155 potentiated ABT-737-induced cytotoxicity in a synergistic manner.³⁰ In summary, our data reveal that the treatment of glioma cells with cucurbitacin potentiated ABT-737-induced cytotoxicity by down-regulating Aurora kinase A, Aurora kinase B and survivin regardless of EGFR/PTEN/p53 status, suggesting the utilization of these agents for GBM.

Materials and Methods

Cell lines

Primary cultures were obtained from freshly resected tissues (histologically confirmed glioblastoma) after surgical removal under an Institutional Review Board-approved protocol for acquisition and the use of tumor tissue collected at the time of tumor resection. We also obtained primary glioma cells from Conversant Biologics (Huntsville, AL). Established malignant human glioma cell lines such as T98G, A172 and LN229 were obtained from the American Type Culture Collection (Manassas, VA). LN18 and LNZ308 were provided by Dr. Nicolas de Tribolet (Lausanne, Switzerland). The establishment of the parental human glioblastoma cell line, U87 and its derivatives, which overexpress exogenous wild-type EGFR (U87-EGFR-WT), or constitutively active EGFR with a genomic deletion of exons 2–7 (U87-EGFRviii) has been described elsewhere.⁷⁶ The cell lines were kindly provided by Dr. Shi-Yuan Cheng (Northwestern University Feinberg School of Medicine, Chicago, IL). PTEN and TP53 status (Table 1) of these glioma cell lines have been characterized elsewhere.^77,78 Human astrocytes (HA) and growth media were obtained from ScienCell Research Laboratories (Carlsbad, CA). Cell culture conditions of these cell lines were as previously described.^7,30,31,79

Reagents and antibodies

ABT-737 and cucurbitacin-I (referred as cucurbitacin in this manuscript) were purchased from Chemie Tek (Indianapolis, IN) and Sigma (St. Louis, MO) respectively. All antibodies were purchased from Cell Signaling Technology (Beverly, MA).

Cell proliferation analysis

Cells (5 × 10³/well) were plated in 96-well microtiter plates (Costar, Cambridge, MA) in 100 μl of growth medium, and after overnight attachment, exposed for the indicated days to inhibitors or vehicle (DMSO). After the treatment interval (24, 48, and 72 h) at 37°C, cells were washed in medium, and the number of viable cells was determined using a colorimetric cell proliferation assay (CellTiter96 Aqueous NonRadioactive Cell Proliferation Assay; Promega, Madison, WI)⁸⁰ essentially by incubating cells in the MTS solution for 2 h and by determining the absorbance at 490 nm wavelength, as reported previously.³¹ Dose response curves and IC50 values were obtained using GraphPad Prism (GraphPad Software, Inc.., La Jolla, CA) by taking the absorption values from an MTS assay.

Annexin V apoptosis assay

Apoptosis induction in vehicle- or inhibitor-treated cells was assayed by the detection of membrane externalization of phosphatidylserine using an Annexin V assay kit (Molecular Probes, Invitrogen) as described previously.³¹ 2 × 10⁵ cells were harvested at various intervals after treatment, washed with ice-cold phosphate-buffered saline (PBS) and resuspended in 200 μl of binding buffer. Annexin V-FITC and 1 μg/ml propidium iodide were added and cells were incubated for 15 min in a dark environment. Labeling was analyzed by flow cytometry with a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA).

Cell cycle analysis

The effect of varying concentrations of inhibitors on cell cycle distribution was determined by flow cytometric analysis of the nuclear DNA content as previously described.³¹ Briefly, cells grown exponentially to 50–60% confluency were exposed to the inhibitors or DMSO for a range of intervals, harvested, washed in ice-cold PBS, and fixed in 70% ethanol. DNA was stained by incubating the cells in PBS containing propidium iodide (50 μg/ml) and RNase A (1 mg/ml) for 60 min at room temperature, and fluorescence was measured and analyzed using a Becton Dickinson FACScan and Cell Quest software (Becton Dickinson Immunocytometry Systems, San Jose, CA).

Western blotting analysis

Treated and untreated cells were washed in cold PBS and lysed in buffer containing 30 mM HEPES, 10% glycerol, 1% Triton X-100, 100 mM NaCl, 10 mM MgCl2, 5 mM EDTA, 2 mM Na3VO4, 2 mM β-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 1 mM 4-(2-aminoethyl) benzenesulfonyl fluoride, 0.8 μM aprotinin, 50 μM bestatin, 15 μM E-64, 20 μM leupeptin, and 10 μM pepstatin A for 15 min on ice. Samples were centrifuged at 12,000 g for 15 min, supernatants were isolated, and protein was quantified using Protein Assay Reagent (Pierce Chemical, Rockford, IL). Equal amounts of protein were separated by SDS PAGE and electrotransferred onto a nylon membrane (Invitrogen). Nonspecific antibody binding was blocked by incubation of the blots with 4% bovine serum albumin in Tris-buffered saline (TBS)/Tween 20 (0.1%) for 1 h at room temperature. The blots were then probed with appropriate dilutions of primary antibody overnight at 4°C. The antibody-labeled blots were washed 3 times in TBS/Tween 20 for 15 min and then incubated with a 1:2000 dilution of horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology, Inc..) in TBS/Tween 20 at room temperature for 1 h. After additional washing in TBS/Tween 20, the proteins were visualized by Western Blot Chemiluminescence Reagent (Cell Signaling Technology Inc., Beverly, MA). Where indicated, the blots were reprobed with antibodies against β-actin to ensure equal loading and transfer of proteins.

Quantitative real-time PCR array

Changes in the gene expression were determined using real-time PCR (RTPCR) assay. Cells (5 × 10⁵/well) were plated in 6-well plates (Costar, Cambridge, MA) in 2.0 ml of growth medium, and after overnight attachment, exposed to inhibitors or vehicle (DMSO) for 24 h. After the treatment, cells were washed in PBS and total RNA was isolated using RNeasy mini kit (Qiagen) as per the manufacturer's protocol. RNA concentration and purity were checked by UV spectrophotometry, and RNA integrity was determined by electrophoresis. Primers were purchased from Qiagen that enable gene expression analysis using SYBR Green-based quantitative real-time PCR. cDNA synthesis was performed using QuantiTect Reverse Transcription Kit (Qiagen # 205311). The RT-PCR reactions were run on the MyIQ (Bio-Rad) and the results, normalized to GAPDH, were determined using iQ5 software (Bio-Rad).

Transient transfection

Optimal 29mer-pRS-shRNA constructs were obtained from Origene (Rockville, MD). Sequences specific for human Aurora kinase A, Aurora kinase B or survivin and non-target control shRNA sequences were used for this study. Cells were seeded in 6-well plates and allowed to reach 70% to 80% confluence. Transfection of targeting or control (non-targeting) shRNA was performed by using FuGene 6 according to the manufacturer's recommendations (Roche Applied Science, Indianapolis, IN). One μg of shRNA in 100 μL Opti-MEM medium was mixed with 2 μL of FuGene 6. After the mixture was incubated at room temperature for 20 min, complete medium was added to make the total volume up to 2 mL. After 48 h, medium was changed and cells were incubated with inhibitors for 24 h. Cell viability (annexin V/propidium iodide binding) or Western blot analysis were carried out as described above.

Fluorescence microscopy

Cells were grown on chamber slides (Nalge Nunc, Naperville, IL) in growth medium, and, after an overnight attachment period, were exposed to selected concentrations of inhibitor or vehicle (DMSO) for various intervals. Cells were washed once with PBS, fixed with 3.7% formaldehyde for 30 min and stained with DAPI for 2 h at room temperature. The slides were then washed in PBS, mounted, and examined under a fluorescent microscope. Morphological changes in response to inhibitor treatment were evaluated by microscopic inspection.

Statistical analysis

Unless otherwise stated, data are expressed as mean ± SD. The significance of differences between experimental conditions was determined using a 2-tailed Student's t test. Differences were considered significant at p values <0.05.

Authorship contributions:

Participated in research design: Daniel R. Premkumar and Ian F. Pollack. Conducted experiments: Esther P. Jane and Daniel R. Premkumar. Performed data analysis: Esther P. Jane, Daniel R. Premkumar, and Ian F. Pollack. Wrote or contributed to the writing of the manuscript: Daniel R. Premkumar and Ian F. Pollack

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Jay Venkatachari, Robert Lacomy and Alexis Styche for technical assistance.

Funding

This work was supported by National Institutes of Health [Grant P01NS40923] (to I.F.P); by the Walter L. Copeland fund of The Pittsburgh Foundation (to D.R.P); by a grant from Ian's Friends Foundation and Connor's Cure Foundation (to I.F.P).