Glioblastoma (GBM) is one of the most frequent malignant cancers occurring in the human brain. The association of temozolomide (TMZ) with radiotherapy is the ideal first-line therapy for GBM [1
]. Despite of this hard line treatment, and the latest improvements obtained with the combination of chemotherapy, surgery, and radiations, GBM regularly recurs and clinical outcome is poor [2
]. Recently, fotemustine (FTM) has shown a significant action on recurrent GBM in phase II trials [4
]. FTM is a highly lipophilic nitrosourea that may cross the blood–brain barrier (BBB) [6
]. FTM also showed a significant antitumor activity on human GBM in preclinical models [8
]. In spite of a wide range of recent specific treatments, GBM therapy remains a challenge. The BBB is the main obstacle for the advances in the delivery of new drugs in the central nervous system (CNS). Additional progress towards a new therapeutic strategy for GBM need to better understand the mechanisms at the basis of cell growth and chemo-resistance of these cancers. Several reports have suggested that GBM are characterized by high expression of 5-lipoxygenase (5-LO), an enzyme responsible for the biosynthesis of inflammatory molecules that are involved in different pathological processes [9
]. In particular, 5-LO is implicated in the growth and survival of tumor cells, such as different brain tumors, suggesting that inhibition of 5-LO activity could represent a valid drug-target for these cancers. To confirm the implication of 5-LO in the pathogenesis of cancer, researchers have used several medications such as 5-LO inhibitors (zileuton, ZYflo, ABT-761), FLAP inhibitors (MK-886) or other related drugs (zafirlukast and montelukast) in inhibiting cell proliferation and inducing cell death in vitro and in vivo [11
]. Our group has synthesized and characterized several agents such as quinone-based derivatives with anti-cancer and anti-inflammatory properties [12
]. These studies on the development of quinone-based compounds with antitumor activity have led to investigate the natural compound embelin and its analogues known as inhibitors of 5-LO and able to suppress proliferation of GBM cells [24
]. In this light, we showed that quinone derivative RF-Id, a novel 5-LO inhibitor [16
], was able to inhibit the inhibitor of apoptosis proteins (IAPs) and to induce cell death in GBM cells. IAPs are often overexpressed in cancer and play a key role in inhibiting apoptosis blocking caspase activation. A recent study has demonstrated that the hydroquinone-based derivative 3-tridecyl-4,5-dimethoxybenzene-1,2-diol (TDD, EA100 Red) is a very potent direct 5-LO inhibitor [26
]. The main aim of this work was to provide a new strategy for GBM treatment. In the light of the potent activity of this novel compound as 5-LO inhibitor, we studied the potential molecular mechanisms at the basis of its anti-tumor effects.
GBM is the furthermost frequent neoplasia occurring in brain, with unfortunate clinical outcome [2
]. In spite of the aggressive conventional treatments, and the latest obtained improvements, GBM regularly recurs and median survival is 14–15 weeks. Additional advances towards a new therapeutic strategy for GBM need to better understand the mechanisms at the basis of cell growth and chemoresistance of these cancers. In the last decade, several approaches have been investigated, such as target-therapies against tumor growth factor receptors, epigenetic regulation, angiogenesis, and metastasis, but there are no evidences that demonstrate patient outcome improvement for this disease [3
]. Several benzoquinones have shown antitumor activity on several types of cancer; in particular they are able to regulate apoptosis, cell cycle, production of ROS. Unfortunately, their use is limited by their cardiotoxicity; therefore, our challenge is to develop quinone compounds effective in treating tumors with fewer side effects. A recent work showed that natural benzoquinone derivatives, such as primin and irisoquin show a significant anti-proliferative activity [27
]. Other important benzoquinones like embelin and its derivatives have shown anti-inflammatory, antioxidant and anti-tumor effects [32
]. A previous study has demonstrated that the orthoquinone EA-100C and its reduced form EA-100C red, synthesized with a C13 n
-alkyl chain lacking hydroxyl groups, showed IC50
values of 10 and 60 nM, respectively, in cell-free assays representing the most potent 5-LO inhibitors [26
Our study provides the proof of a complex cross-talk between apoptosis, autophagy, ER stress, and NF-kappaB pathway in GBM cells, which regulates the destiny of cancer cells and senses changes in tumor microenvironment. In response to different stresses, the ER triggers the so called unfolded protein response (UPR), an adaptive response that restores ER homeostasis. If the stress is too long, ER stress activates cell death mechanisms [37
]. Further progress towards a therapy for malignant gliomas requires an improved understanding of the mechanisms underlying the switch between pro-survival and pro-death UPR signals and the crosstalk of the UPR itself with other signaling pathways. In this study, we studied the effects of the hydroquinone derivative EA-100C red on growth inhibition of GBM cells. In particular, EA-100C red showed IC50
values of 75,5 and 21 μM in U87MG and LN229, respectively, therefore, EA-100C red was more potent on LN229. GBM cells have cancer-specific genetic alterations as described in [38
] that could explain this difference in response to the drug.
U87MG, and less LN229, are fragile cells and during detachment (with trypsin or any other method) phosphatidylserine residues in the membrane can translocate and give false positive results; therefore, we always found a 75–80% of live cells even if we used accutase, reduced trypsyn time or scraped cells. According to this, as one can see in the MTT assay that it does not need cell detachment, cell viability is very high. On these bases, we can confirm that low viability does not affect the response to the drug because when cells are gently treated, after detachment we found few false positives. However, considering false positives, apoptosis was significantly increased (two times) with the compound was compared to untreated cells. EA-100C red induced a significant increase in apoptosis and autophagy associated to the disruption of MMP on both cell lines. Probably, the difference of the effects of EA-100C red on the mitochondrial potential is due to genetic differences between the two cell lines and to the different molecular context. In particular, the LN229 cell line seems less sensitive to the effects on ROS production due to its higher antioxidant capacity and expression of enzymes that have a key role in redox homeostasis [39
] compared to U87MG. In order to better understand the molecular mechanisms driving its anti-proliferative effects, we assessed the effects of EA-100C red on the expression of genes involved in cell death mechanisms. In particular, EA-100C red increased the levels of NFKBIA expression and decreased NFKB1 expression compared to control, in LN229. On the other hand, in U87MG, EA-100C red up-regulated NFKBIA, Bcl3, but down-regulated CASP8. The difference of caspase 8 expression among the two cell lines is probably not related to a difference in the induction of apoptosis via intrinsic and/or extrinsic pathway; in ER stress both caspases 8 and 9 are often activated. We probably found the difference of caspase 8 expression among the two cell lines because of the different p53 status of the cells.
It is reported that caspase 8 can regulate the NF-kappaB pathway independently from its activity as a pro-apoptotic protease [40
]. Probably, EA-100C red induced apoptosis by modulating the NF-kappaB pathway. In this light, we studied NF-kappaB-related pathways such as endoplasmic reticulum stress pathway. In particular, we investigated the effects of compound on the expression of CHOP and XBP1, that are involved in ER-stress mechanisms. In details, EA-100C red induced an up-regulation of XBP1 and at a greater extent of CHOP. Up-regulation of CHOP, triggered by ATF4 induces a persistent state of stress that leads the cells to programmed cell death [37
]. Apoptosis and autophagy induced by EA-100C red may be related to induction of endoplasmic reticulum stress. Based on the obtained data, we selected temozolomide-resistant cell line LN229 that resulted more sensitive (IC50
= 21 μM) to the drug; this was made in order to better study the molecular mechanism at the basis of the cell death induced by this new compound. It is known that high ROS overproduction leads to ER stress in the cells but it seems that also low levels of ROS can induce protein kinases and phosphatases activation, mobilize deposits of Ca2+, regulate transcription factors, resulting in apoptosis [41
]. To investigate if apoptosis induced by EA-100 red could be mediated by ER stress even if ROS levels in mithocondria were low, we selected LN229. Our aim is to find a potential candidate for GBM treatment, for which at least 50% of the patients do not respond to temozolomide and this compound seems to be a useful tool, at least based on our results. We evaluated the effects of EA-100C Red on the main pathways of signal transduction, the cascade of caspases and the NF-kappaB pathway. We studied the p65 subunit of NFKB, and it is not surprising that it is absent in the control by considering that this cell line is p53 mutated and p53 indirectly regulates p65 transcription [42
EA-100C red induced apoptosis through the caspase cascade and led to the activation of IRE-1, JNK, NF-kappaB and the increase of Beclin-1 which are involved in the cell death mechanisms related to the ER-stress. Data coming from literature suggest that an increase of the misfolded proteins can activate ER stress and lead to apoptosis paralleled by autophagy in cancer cells. These effects are regulated by IRE1/JNK/Beclin-1 pathway. Based upon these data, we specifically evaluated the latter pathway [37
]. In ER stress conditions, the activation of TRAF2-IRE1 would lead to JNK and NF-kappaB activation that induce apoptosis and autophagy through the activation of caspases and Beclin-1 up-regulation [37
]. In conclusion, EA-100C red was probably able to induce ER stress-mediated apoptosis and autophagy associated to the disruption of mitochondrial membrane potential by activating IRE-1 pathway and up-regulating CHOP levels.
4. Materials and Methods
Analytical grade reagents were used in this study and purchased from Sigma Aldrich (Milan, Italy). Carlo Erba silica gel 60 (230–400 mesh; Carlo Erba, Milan, Italy) were used for flash chromatography. We purchased plates coated with silica gel 60F 254 nm from Merck (Darmstadt, Germany) to carry out TLC. We recorded the 1H- and 13C-NMR spectra using an AC 300 instrument (Bruker, Billerica, Massachusetts, USA).
EA-100C red was synthesized as reported in our previous works [13
4.3. Cell Culture
U87MG and LN229 cells were a gift of Dott. Carlo Leonetti (Regina Elena Cancer Institute, Rome, Italy). LN229 and U87MG were cultured in DMEM and RPMI, respectively, (Life Technologies, Carlsbad, CA, USA) increased with 10% FBS (fetal bovine serum), 1% L-glutamine, streptomycin and penicillin (Lonza Group Ltd., Basel, Switzerland). Cells were grown in a 5% CO2 -95% air incubator at 37°C.
4.4. Cell Viability Assay
U87MG and LN229 were plated in 96-well plates (2 × 103
cells/well) and treated with increasing concentrations of EA-100C red (0.8–100 μM) for 24 h, 48 h and 72 h. Cell proliferation was evaluated by MTT assay as formerly described [25
]. At least three separate experiments were performed.
4.5. Analysis of Apoptosis by FACS
Apoptosis was identified by using FITC Annexin V Apoptosis Detection Kit I (BD Biosciences Pharmingen, Heidelberg, Germany). GBM cells were plated in 6-multiwell plates at the density of 2 × 105 cells/well and treated after 24h with EA-100C red. After 72 h, cells were incubated as described by the manufacturers. BD Accuri™ C6, (Becton Dickinson, San Jose, CA, USA) was used to perform FACS analysis. We acquired about 20,000 events for each sample, in at last three separate experiments. Annexin V-FITC and PI fluorescence were detected using the FL1 and FL3 channels, respectively.
4.6. FACS Analysis of Mitochondrial Potential
LN229 and U87MG cells were seeded in 6-multiwell plates (2 × 105
cells/well) and treated for 72h. Subsequently, cells were collected and stained with Mitotracker Red probe as described in our previous work [25
]. After labeling with the probe, cells were analysed by FACScan, BD Accuri™ C6 and red fluorescence was collected through FL2 channel. We acquired about 20,000 events in at last three separate experiments for each sample.
4.7. Flow Cytometric Analysis of Autophagy
GBM cells were plated in 6-multiwell plates (2 × 105
cells/well). Cells were treated for 72 h and then incubated with MDC as described in our previous work [25
]. After MDC staining (Sigma, Milan, Italy), cells were collected and analyzed by flow cytometry (FACScan, BD Accuri™ C6). For all the samples MDC fluorescence was collected through the FL1 channel and 20,000 events were acquired in at last three separate experiments. The formula (MFI treated/MFI control) was used to calculate the mean fluorescent intensity.
4.8. Taqman Human Apoptosis Array-Real-Time-PCR
After treatment, total RNA was extracted according to the mirVana PARIS (Ambion, Life Technologies) manufacturer’s instructions. Then RNA quantity was quantized by NanodropND-1000 Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). Reverse transcription was performed according to the QuantiTect Reverse Transcription Kit, (Qiagen, Hilden, Germany) instructions. Taqman human apoptosis array including 93 apoptosis-related genes and 3 housekeeping controls (18S, ACTB, GAPDH) was performed by using ViiA7™ Real time PCR system (Applied Biosystems, Darmstadt, Germany). The comparative cross-threshold (Ct) method was performed to calculate the relative expression of the transcripts. Untreated control was used to normalize treated samples.
4.9. Immunofluorescence with ER Tracker
The ER stress was identified by staining cells with the ER-specific dye ER Tracker™ Blue/White 1 µM in a solution of PBS, for 30 min at 37 °C. Tunicamycin 2 µM for 16 h represents the positive control. After 72 h treatment with EA-100C red at IC50 concentration, cell fixation was performed with 4% paraformaldehyde solution, then permeabilization with 0.1% Triton X/PBS solution, finally blocking was performed with a 1% BSA/FBS solution for 1 h at RT. Images were collected under a fluorescence microscope (LSM 510, X63, Zeiss, Oberkochen, Germany) after MOVIOL counterstaining.
4.10. Western Blot Analysis
After 72 h of treatment with the concentration inhibiting 50% of cell growth (IC50) of EA-100C Red, cells were collected and lysated for 30 min at 4 °C in 1 mL of lysis buffer (1% Triton, 0.1 NaCl, 0.5% sodium deoxycholate, 10 mM Na2HPO4, pH 7.4, 1mM EDTA, pH 7.5, 10 mM PMSF, 1 mM leupeptin, 25 mM benzamidin, 0.025 units/mL aprotinin). Total proteins were electrotransferred to a nitrocellulose membrane by using Trans blot turbo (BioRad, Hercules, CA, USA). TBST (150 mM NaCl, 10 mM Tris, pH 8.0, 0.05% Tween 20) was used to wash the membranes. Then, membranes were incubated with specific Abs. The rabbit antibodies raised against, caspase-3, caspase-8, p-IKK, IKB, p-p65, p65, Beclin 1, 5-LO and the mouse antibodies raised against capase-9, p-IKB, IKK were acquired from Cell Signaling Technology (Denvers, MA, USA). The mouse antibodies for p-JNK, JNK and p-IRE were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Enhanced chemiluminescence detection reagents ECL (Thermo Fisher Scientific, Rockford, IL, USA) was used to develop the blots. Blots were analyzed by using Quantity One software (BioRad Chemi Doc). At least three separate experiments were performed.
4.11. Statistical Analysis
All data are expressed as mean ± SD. Analysis of variance (ANOVA) together with Neumann-Keul’s multiple comparison test or Kolmogorov-Smirnov test were performed to obtain statistical information.