Cilengitide induces cellular detachment and apoptosis in endothelial and glioma cells mediated by inhibition of FAK/src/AKT pathway

Human microvascular endothelial cells (HMEC-1), kind gift from Centre for Disease Control and Prevention, Atlanta, U.S.A., were grown in MCDB 131-medium (Gibco) supplemented with 5% fetal bovine serum (FBS, Gibco), 2 mM L-glutamine (Gibco), 10 ng/ml epidermal growth factor (ICN, Costa Mesa, CA, U.S.A.) and 1 μg/ml hydrocortisone (ICN), and maintained on uncoated dishes in a 5% CO₂/95% air atmosphere in a humidified incubator at 37°C. Porcine aortic endothelial cells stably transfected with KDR (PAE-KDR), provided by Shay Soker, Winston-Salem, NC, were maintained in F-12/HAM medium supplemented with 5% fetal bovine serum at 37°C in 5% CO₂/95% air. Commercial human umbilical vein endothelial cells (HUVECs) (Lonza) were cultured in EGM-2 medium (Clonetics) including 2% fetal calf serum. The human glioblastoma cell lines G28 and G44 [24, 25], kindly provided from the Department of Neurosurgery, University Hospital Hamburg-Eppendorf, were cultured in Modified Eagle's Medium supplemented with 10% fetal bovine serum on uncoated dishes. Cilengitide (CGT) was kindly provided by Merck Serono, Darmstadt, Germany. Stock solutions were diluted in sterile physiological saline solution at 20 mg/ml. Cells were incubated with cilengitide in final concentrations of 1, 5 and 50 μg/ml. Temozolomide (TMZ) was purchased from Bristol Myers Squibb, Munich. Stock solution was diluted in DMSO at 5 mg/ml. Cells were treated with temozolomide in a final concentration of 5 μg/ml. Texas Red-X phalloidin was from Invitrogen, mouse monoclonal anti phospho-Akt (Ser473) antibody was from Cell Signaling, rabbit polyclonal anti phospho-Src (Y418) antibody was from Biosource, mouse monoclonal anti phospho-Src (Y416) was from Biomol, mouse monoclonal anti Src antibody was from Upstate (NY, USA), mouse monoclonal anti phospho-FAK (Y397) was from BD Biosciences, rabbit polyclonal anti ZO-1, anti Erk1/2, mouse monoclonal anti phospho-Erk and anti β-actin antibodies were from Santa Cruz Biotechnology.

Tissue culture plates were incubated with a 12 mg/ml polyHEMA (Poly(2-hydroxyethyl methacrylate); Sigma Aldrich) ethanol solution at 37°C or with 10 μg/ml fibronectin (Harbor Bio-Products, Norwood, MA, USA) at 4°C overnight when indicated.

HMEC-1, G28 and G44 cells (1 × 10⁴ per well) were seeded on uncoated 48 well plates and incubated in serum free medium, medium containing 4% FCS or medium containing 4% FCS with cilengitide (1, 5 and 50 μg/ml) or/and temozolomide (5 μg/ml). For experiments with temozolomide, control cells were treated with medium containing 4% FCS and DMSO at the equivalent concentration used for the temozolomide stock solution. Each stimulation was performed in triplicate. After incubation for 24, 48 and 72 hours at 37°C cells were trypsinized and counted.

HMEC-1, G28 and G44 cells (5 × 10⁵) were incubated on uncoated dishes with and without cilengitide (1, 5 and 50 μg/ml) for 24 hours at 37°C. G28 and G44 cells (2 × 10⁵) were incubated with temozolomide (5 μg/ml) and cilengitide (5 μg/ml) in combination or separately for 48 hours at 37°C. For experiments with temozolomide, control cells were treated with medium containing 4% FCS and DMSO at the equivalent concentration used for the temozolomide stock solution. Apoptosis was assessed after staining with FITC-labeled annexin-V and PI (BD Pharmingen) by flow cytometric analysis. Positive staining with FITC-labeled annexin-V reflects a shift of phosphatidylserine from the inner to the outer layer of the cytoplasmatic membrane, which occurs early in apoptosis. Annexin-V-positive and PI-negative cells were scored as early apoptotic cells. Cells labeled by annexin V and PI have been determined as late apoptotic. Annexin negative and PI-positive events display necrotic cells.

HMEC-1 and G28 cells were plated on cover slips and treated with and without cilengitide (1, 5 and 50 μg/ml) for 1 hour at 37°C. Cells were fixed with 4% paraformaldehyde, permeabilized with methanol, and stained for ZO-1 (Santa Cruz Biotechnology) and actin (PE-phalloidin, Invitrogen). Immunofluorescence analysis was carried out with an Axioplan (Zeiss) inverted microscope.

Protein extracts were prepared with lysis buffer solution containing 50 mM Tris-HCl pH 7,4, 150 mM NaCl, 100 mM EGTA, 1% Nonidet P-40, 10% Na-deoxycholate, 1× protease inhibitor cocktail (Sigma Aldrich) and 1 mM sodium orthovanadate. Protein lysates were boiled in SDS-sample buffer before being applied into a 10% SDS-PAGE. After electrotransfer to nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany) and blocking in TBS-T buffer containing 5% non-fat milk overnight, blots were incubated with the appropriate primary antibody. The subsequent incubation with the peroxidase-conjugated secondary antibodies was followed by detection using ECL Western blotting detection reagents (Amersham). Specific bands were quantified by densitometric analysis using the GS-800 Calibrated Densitometer and Quantity-one software (BioRad).

Total cellular RNA from HMEC-1, G28 and G44 cells was extracted using RNeasy Kit from Qiagen. 3 μg of total RNA each were reverse transcribed into cDNA using the You-Prime First-Strand cDNA synthesis kit (Amersham). Following primers were used to amplify the genes encoding integrin subunits αv and β3: integrin αv (671 bp) forward (5'-cttcaacctagacgtggacagt-3') and reverse (5'-ttgaaatctccgacagccacag-3'), integrin β3 (123 bp) forward (5'-agaagagccagagtgtccca-3') and reverse (5'-gaattcttttcggtcgagga-3'). For amplification, touch down-PCR was performed (one cycle from 65°C-55°C and 25 cycles at 55°C). The amplification products were visualized on a 1% ethidium bromide-stained agarose gel.

DNA was extracted using QIAmp DNA Mini Kit from Qiagen. 2 μg genomic DNA was denaturated and chemical modificated via bisulfite treatment using the EZ DNA Methylation-Gold Kit from Zymo Research. For glioma cell lines (G28 and G44) and normal lymphocytes DNA was first amplified with flanking PCR primers as previously described [26]. The resulting fragment was used as a template for the MSP reaction. Subsequent PCR was performed with specific primers for either methylated or the modified unmethylated promotor region of MGMT gene. Primer sequences for the unmethlyated reaction were: 5'-TTT GTG TTT TGA TGT TTG TAG GTT TTT GT-3' (upper primer) and 5'-AAC TCC ACA CTC TTC CAA AAA CAA AAC A-3' (lower primer) and for the methylated reaction 5'-TTT CGA CGT TCG TAG GTT TTC GC-3' (upper primer) and 5'-GCA CTC TTC CGA AAA CGA AAC G-3' (lower primer). The annealing temperature was 59°C. Universal methylated human DNA Standard was used as a positive control for methylated alleles of MGMT and DNA from normal lymphocytes was used as a negative control. The PCR products were separated on a 4% agarose gel.

We first studied the effect of cilengitide on endothelial cell attachment in vitro. The human microvascular endothelial cell line HMEC-1 cultured in monolayer on uncoated dishes was incubated with and without cilengitide at concentrations of 1, 5 and 50 μg/ml. As shown in figure 1A, cilengitide induced a dose dependent detachment of HMEC-1 cells, accompanied by striking morphologic changes after 24 hours incubation.

Cilengitide, added at concentrations of 1, 5 and 50 μg/ml over a period of 72 hours, significantly decreased proliferation of HMEC-1 cells grown on uncoated dishes in vitro. We observed a dose-dependent reduction of endothelial cell counts, as shown in figure 1B. At a concentration of 1 μg/ml, cilengitide induced 33%, 59% and 44% inhibition after 24, 48 and 72 hours, respectively. In contrast, at concentrations of 5 and 50 μg/ml almost no proliferation of endothelial cells was observed comparable to the effect of serum starvation.

To investigate whether apoptosis was responsible for the decrease of adherent endothelial cells treated with cilengitide, we measured Annexin V/propidium iodide (PI) positive cells after incubation with and without cilengitide at varying concentrations. In HMEC-1 cells cilengitide had a significant pro-apoptotic effect, which was more profound with increasing concentrations (1, 5 and 50 μg/ml) after 24 hours incubation (figure 1C).

Cilengitide has been reported to inhibit glioblastoma growth via suppressing angiogenesis [16]. Since cilengitide acts as antagonist to integrin αvβ3 and αvβ5 and both integrins are expressed in glioma cells, especially on the periphery of high-grade gliomas [13], we asked whether cilengitide has a direct effect on glioma cells. Human glioma cell lines G28 and G44 expressing integrins αvβ3 and αvβ5 (data not shown) were incubated with increasing concentrations of cilengitide and changes were studied after 24 hours. Similar to endothelial cells, cilengitide inhibited adhesion of G44 and G28 glioma cells in a dose-dependent manner (fig. 2). In contrast to control cell cultures, where few cells detached (fig. 2A–2B), G28 (fig. 2C) and G44 (fig. 2D) cells treated with 1 μg/ml cilengitide for 24 hours showed striking cellular detachment. After 24 hours exposure to 5 μg/ml (fig. 2E–2F) and 50 μg/ml (fig. 2G–2H) cilengitide, almost all cells were rounded up, completely detached and seemed to form cell clusters.

Glioma cell lines G28 and G44 were treated with 1, 5 and 50 μg/ml cilengitide and cell counts were determined after 24, 48 and 72 hours. We observed an inhibitory effect on cell proliferation already at the lowest concentration (1 μg/ml) with significant differences becoming visible after 48 hours (fig. 3A–3B). Higher concentrations of cilengitide abolished cell proliferation and induced apoptosis already after 24 hours exposure to the drug. Annexin V/propidium iodide staining revealed the presence of 18% apoptotic G28 and 30% apoptotic G44 cells after treatment with 5 μg/ml cilengitide. At a concentration of 50 μg/ml, 35% – 50% of G28 and G44 cells respectively underwent apoptosis showing a dose dependent apoptosis induction in glioma cells (fig. 3C–3D).

In contrast to integrin-mediated cell death of adherent cells, apoptosis of cells loosing adherence is termed anoikis [27]. To compare cilengitide-induced apoptosis with anoikis, G28 cells were seeded on polyHEMA-coated surfaces that totally prevent cell adhesion. The rate of apoptotic cells when G28 were cultured on polyHEMA was identical to those observed with cilengitide treatment of adherent growing cells (fig. 3C). When cilengitide was added to non-adherent cells no increase in the rate of apoptosis was observed (fig. 4). Therefore, the mechanism of cilengitide-mediated apoptosis was indistinguishable from anoikis, suggesting that the anti-adhesive effect of cilengitide is responsible for apoptosis induction.

Cilengitide has been shown to inhibit proliferation and to induce apoptosis in endothelial cells, but its effects on integrin mediated signaling pathways are unknown. Therefore we studied the effect of integrin blockade by cilengitide on the phosphorylation and thus activation of p38 MAPK, p44/42 MAPK (Erk1/2), FAK, Src and Akt. After 24 h of starvation in serum-free medium HUVECs grown on uncoated dishes were treated with 20, 40 and 60 μg/ml cilengitide for 1 hour and the activation of signaling molecules was evaluated by Western blotting with phosphoprotein-specific antibodies. As shown in figure 5A, integrin blockade moderately reduced phosphorylation of FAK and Src dose-dependently, but not of Erk 1/2 and p38 (data not shown) under this conditions. Densitometric analyses were performed to quantify the effects on signalling events (figure 5B).

Integrins can physically interact with growth factor receptors to regulate a variety of biological processes. In particular, the interaction between αvβ3 integrin and KDR is important for downstream signaling of both receptor types [28, 29]. Given that cilengitide interacts with the extracellular domain of αvβ3, and might thereby interfere with the association of KDR with αvβ3 integrin, we examined whether cilengitide also affects downstream components of KDR signaling pathways. Stimulation of PAE-KDR cells (porcine aortic endothelial cell line stably transfected with VEGFR-2/KDR) with VEGF induced phosphorylation of KDR, FAK and Erk. Simultaneous treatment of PAE-KDR cells with VEGF and cilengitide (1, 5 and 50 μg/ml) did not alter KDR activation when comparing to total KDR protein but inhibited phosphorylation of FAK (figure 5C/D). Erk phosphorylation was decreased only at higher concentrations of cilengitide (50 μg/ml) after 10 minutes, which may be due to pathway crosstalk, since KDR activity was not inhibited under these conditions. These results suggest that cilengitide inhibits integrin-dependent signaling through FAK and Src while it does not influence VEGFR-2/KDR and its downstream pathways in endothelial cells.

We next studied the effect of cilengitide on integrin-mediated signaling pathways in the absence of VEGF in glioma cells. In order to determine the optimal time frame for signaling events G28 cells were lyzed after 30, 60 and 120 minutes of incubation with cilengitide (50 μg/ml) and analyzed for activation of FAK and Akt by Western blot using phospho-specific antibodies. As shown in figure 6A and 6B, inhibition of FAK phosphorylation by cilengitide was observed already at 30 minutes and this effect continued for at least until 1 hour. Additionally, inhibition of pAkt was noted after 60 minutes. Therefore, following experiments were performed with incubations of 1 hour using increasing concentrations of cilengitide. Incubation with cilengitide under these conditions caused inhibition of Src and Akt phosphorylation downstream of FAK in a dose dependent manner as shown in figure 6C. Results were quantified using densitometric analyses (figure 6D).

These results demonstrate that cilengitide inhibit identical pathways in glioma and endothelial cells explaining similar effects such as detachment and apoptosis induction observed in both cell types.

To analyze the effect of cilengitide on the distribution of the tight junction proteins and actin filaments, we performed immunofluorescent staining of endothelial (HMEC-1) and glioma cells (G28) for zona occludens (ZO-1) and phalloidin.

In control cells staining with ZO-1 highlighted tight junctions at cellular borders with continuous staining along cell-cell contacts on both HMEC-1 and G28 cells (fig. 7A, 7C). Cells exposed to cilengitide for 24 hours lost distribution of ZO-1 to intercellular junctions, showing irregular distribution with sporadic cytoplasmatic aggregates (figures 7E, 7I, 7G, 7K).

PE-phalloidin staining of actin cytoskeleton, revealed a disassembly of actin filaments in cilengitide treated endothelial and glioma cells compared to controls (figures 7B, 7D). With the disappearance of the actin fibers from the cell interior, we observed clustering of microfilaments along cell borders (figures 7F, 7J, 7H, 7L). Although effects were similar in both cell types, glioma cells appeared more sensitive for disassembly of filaments and cellular detachment. These observations highlight the profound changes on intercellular contacts and cytoskeleton caused by cilengitide similarly in endothelial and glioma cells.

Temozolomide (TMZ) is a DNA-methylating agent with activity as monotherapy in malignant gliomas [30‐32]. However the benefit from temozolomide treatment in glioblastoma is strongly associated with MGMT promoter methylation [33]. Therefore, we determined the MGMT promotor methylation status using a methylation-specific PCR assay. Cell lines G28 and 44 both had a methylated MGMT promotor as shown in figure 8. Lymphocytes from peripheral blood of healthy volunteer served as negative controls with unmethylated promotor, while a positive control was clearly methylated.

We next studied the effect of temozolomide (TMZ) in combination with cilengitide on glioma cellls with methylated MGMT promotor. G28 and G44 cells were incubated with cilengitide (5 μg/ml) and temozolomide (5 μg/ml) alone or in combination for 72 hours and changes were studied after 24, 48 and 72 hours. In contrast to cells treated with cilengitide, where many cells detached already after 24 hours (fig. 9C, 9D), G28 (left) and G44 (right) cells treated with TMZ alone did not show morphological changes or cellular detachment when compared to controls (fig. 9E, 9F). The combination of cilengitide and TMZ led to an increased detachment of glioma cells and cell cluster formation being more pronounced after 48 hours (fig 9G, 9H).

Glioma cell lines G28 and G44 were treated with 5 μg/ml cilengitide and 5 μg/ml temozolomide alone or in combination and cell counts were determined after 24 and 48 hours. As expected, an inhibitory effect on cell proliferation was observed on glioma cells treated with cilengitide already after 24 hours, whereas treatment with temozolomide alone showed only slight inhibition of proliferation in G44 and G28 glioma cells. When cilengitide was combined with temozolomide proliferation inhibition was slighty pronounced in both cell lines. Quantification of proliferation inhibition by either cilengitide or TMZ alone compared to the combination of both compounds suggested additive effects of cilengitide and TMZ in G44 cells (fig. 10A). The combination caused additional inhibiton in G28 cells. However the data are to limited to propose a synergism of the combination in these cells (fig. 10B). Differences observed between cell lines might be due to higher integrin expression in G44 compared to G28 cells (data not shown).

Annexin V/propidium iodide staining demonstrated similar apoptosis induction after incubation with either cilengitide or temozolomide in G44 cells. The combination of both compounds further increased the amount of apoptotic cells (fig. 10C). Apoptosis induction in G28 cells was less pronounced, but showed similar trends (fig. 10D).

Taken together, these results suggest additive activity of cilengitide combined with TMZ in glioma cells with methylated MGMT promotor.