Resveratrol abrogates the Temozolomide-induced G2 arrest leading to mitotic catastrophe and reinforces the Temozolomide-induced senescence in glioma cells

TMZ (3,4-dihydro-3-methyl-4-oxoimidazo [5,1-d]-as-tetrazine- 8-carboxamide), Rsv, 3-methyladenine (3MA) and the fluorescent dye acridine orange (AO) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). TMZ and Rsv were dissolved in dimethyl sulfoxide (DMSO) (Acros Organics, NJ, USA). 3MA and AO were dissolved in water. All culture materials were obtained from Gibco Laboratories (Grand Island, NY, USA).

Human GBM cell lines U87-MG (p53^wt, PTEN^mut, p14ARF/p16^del, low MGMT levels), U-138 MG (p53^mut, PTEN^mut, p14ARF/p16^del, high MGMT levels) and U251 (p53^wt, PTEN^null, p14ARF/p16^del, low MGMT levels), described hereafter only as U87, U138 and U251, were obtained from American Tissue Culture Collection (ATCC, Rockville, MD). Cell lines were cultured in DMEM low glucose, while primary cultures were maintained in DMEM high glucose, both supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin and 0.1% amphotericin B at 37°C and 5% CO₂ in a humidified incubator. The inhibitor 3-MA was used at he concentration of 2 mM, in a pre-incubation of 1 h before the treatments with Rsv and TMZ. The concentration of the vehicle DMSO did not exceed 0.5% (v/v). Cells were counted in a hemocytometer and viability was accessed by measuring PI incorporation as described [32]. Primary GBM culture was established from a biopsy of a GBM tumor following the ethical procedures approved by the Ethical Committee of PUC-RS number 07/03562.

Autophagosome formation: cells were transfected with the expression vector pEGFP-LC3 (Microtubule-associated protein 1 light chain 3 (MAP1-LC3), which localizes at the autophagosome membranes after processing [33]). Cells were imaged with a Zeiss Axiovert 200, using the 40x objective and at least 100 green cells per treatment were counted and the percentage of cells with at least 5 clear green dots in the cytosol was determined [34].

Acidic vacuolar organelles (AVOs) quantification: acridine orange (AO) is a marker of AVOs that fluoresces green in the whole cell except in acidic compartments (mainly late autophagosomes), where it fluoresces red. Cells were plated at 2 × 10⁴ cells per well in a 24-wells plate, followed by treatments as indicated. After this, cells were incubated with 2.7 μM of AO for 15 min at room temperature, followed by visualization in a fluorescence microscope. Images were analyzed using Image J software. To quantify the percentage of cells with AVOs (i.e. red marked cells) and the intensity of red fluorescence (i.e. the intensity of AVOs formation), treated cells were detached from the plate, marked with AO as cited above and analyzed by flow cytometry, using a flow cytometer GUAVA EasyCyte and GUAVA software ExpressPlus (Guava Technologies, Hayward, CA).

Apoptosis induction was quantified by Annexin V-FLUOS Apoptosis Kit (Roche, Germany) according to manufacturer’s instructions with minor modifications, as described [24].

For cell cycle analysis, cells were plated at 2 × 10⁴ cells per well in a 24-wells plate, followed by treatments as indicated. After treatments, cells were harvested and fixed in ice-cold ethanol 70% (v/v in PBS) for at least 2 h. Fixed cells were washed with PBS and marked with a solution containing 50 μg/ml PI, 0.1% Triton X-100 and 50 μg/mL RNAse for 30 min, in the dark, at room temperature. Marked cells were analyzed using the flow cytometer GUAVA EasyCyte software to evaluate DNA content of cells and, thus, cell cycle distribution of samples.

TMZ, Rsv and cotreatment-induced DNA damage was quantified using the alkaline comet assay, as described by Singh et al., with minor modifications [35‐37]. Cells were plated at 5 × 10⁴ cells per well in a 24-wells plate, followed by treatments for 20 and 48 h, as indicated. Cells were embedded in 0.75% low-melting agarose and placed onto a glass microscope slide pre-coated with a thin layer of 1% normal melting point agarose. Slides were then incubated in ice-cold lysis solution [2.5 M NaCl, 10 mM Tris, 100 mM EDTA, 1% Triton X-100 and 10% DMSO, pH 10.0] at 4°C for at least 1 h. After, slides were incubate with fresh alkaline buffer (300 mM NaOH, 1 mM EDTA, pH. 13.0) and followed by electrophoresis. Slides were then neutralized (0.4 M Tris, pH 7.5), washed with water, and stained using a silver staining protocol as described by Nadin et al.[38]. One hundred nuclei were scored blindly according to the amount of DNA present in the tail and the tail length. Each nuclei received an arbitrary value range from 0–4 (0, undamaged; 4, maximally damaged) [39], and 100 nuclei per slide were evaluated.

Analysis of protein expression and phosphorylation was performed as described previously with minor modifications [15, 24]. Primary antibodies used were: cyclin D1 (1:1000), phospho-Rb (S807/811)(1:1000), phospho-Cdk1 (Tyr15)(1:1000), cyclin B (1:250), phospho-ATM (Ser1981)(1:1000), phopho-Chk2 (T68)(1:1000), gammaH2AX (1:1000), phospho-Wee-1 (S642) (1:500) and phospho-H3 (Ser10) (1:1000) (Cell Signal ling, Beverly, MA). Optical density of the bands was obtained with Bio-Rad software (Quantity One; Hercules, CA).

For clonogenic assay, cells were treated with Rsv, TMZ or cotreatment for 48 h, followed by medium removal. Cells were washed twice with PBS, harvested and plated at a density of 10² cells/well in a 6-wells plate. After 14 days, colonies were fixed with methanol, followed by staining with 0.1% crystal violet. The number of colonies was counted and single colonies were photographed for analysis.

For senescence measurement, cells were treated with Rsv, TMZ or cotreatment for 48 h, followed by medium removal. Cells were washed twice with PBS and replated at a density of 20 × 10³ cells/well, in a 12-wells plate. After seven days, cells were tested for senescence as described [40], with minor modifications. Briefly, cells were washed with PBS, fixed with 2% paraformaldehyde for 30 min at room temperature and incubated with fresh SA-beta-gal staining solution (1 mg/mL X-gal (Sigma), 40 mM citric acid/sodium phosphate (pH 6.0), 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 150 mM NaCl, and 2 mM MgCl) for 8–12 h at 37°C. Then, cells were marked with a solution containing 300 nM DAPI and 0.1% triton X-100 (v/v in PBS) for 30 min at room temperature. Results are presented as ratio of SA-beta-gal-positive cells to total cells.

The analysis of nuclear morphometry was performed using a tool recently developed by our group [41]. Briefly, cells were treated as described in SA-beta-gal assay and, at day 7, cells were fixed with 2% paraformaldehyde (v/v in PBS) for 30 min at room temperature, and kept in PBS. Next, fixed cells were marked with a solution containing 300 nM DAPI and 0.1% triton X-100 (v/v in PBS) for 30 min at room temperature, followed by quantification of the images obtained with DAPI staining using the Software Image Pro Plus 6.0 (IPP6 - Media Cybernetics, Silver Spring, MD) or Image J plugin available at http://​www.​ufrgs.​br/​labsinal/​nma. Data is presented as a plot of Area versus Nuclear Irregularity Index (NII), which separates nuclei considering its morphometric phenotype. The percentage of normal, irregular, large and regular, large and irregular, small, small and regular and small and irregular nuclei were determined as described [41].

To measure the levels of reactive species, we performed the DCF assay. The fluorescein derivative DCF (Sigma-Aldrich) is a non-fluorescent compound which is converted to a highly fluorescent DCF upon oxidation by oxygen or nitrogen reactive species. To this, 5 × 10⁴ cells were plated in 24-well plates, followed by treatments as indicated. Cells were harvested, washed once with PBS 1× and incubated with 10 μM (in PBS 1×) for 30 min at 37°C prior to analysis by flow cytometry.

TMZ and Rsv individually reduced the number of glioma cells (Figure 1), as shown by others [6, 7, 24]. When given together, Rsv potentiated the cytotoxic effect of TMZ on human glioma cell lines U87, U138 and U251 (Figure 1 and Additional file 1: Figure S1a). A primary GBM culture which was resistant to TMZ was sensitive to Rsv and to RT (Additional file 1: Figure S1b). Rsv 30 μM plus TMZ 100 μM (hereafter referred to as RT) did not induce loss of membrane integrity, large morphological alterations, increase in annexinV/PI-staining or ROS when cells were treated for 48 h (Additional file 2: Figure S2), indicating that apoptosis, necrosis or a major imbalance in ROS levels are not involved in the reduction in cell number induced by RT.

TMZ, Rsv or RT induced autophagy (Figure 2A, Additional file 1: Figure S1B and Additional file 3: Figure S3) both in primary GBM cells and cell lines, which can be partially inhibited with 3MA, an inhibitor of PI3K class III, a protein whose activity is necessary for the early formation of autophagosomes (Figure 2A). RT induced higher levels of autophagy in an additive manner with respect to Rsv and TMZ alone (Figure 2A, Additional file 1: Figure S1 and Additional file 3: Figure S3), however, an efficient inhibition of autophagy with 3MA did not reduce toxicity of treatments, even increasing the toxicity of Rsv, suggesting a small protective role of autophagy (Figure 2B).

As expected, TMZ induced a robust G2/M cell cycle arrest in U87 cells, while Rsv did not alter cell cycle distribution after 48 h of treatment. Interestingly, the presence of Rsv completely abrogated TMZ-induced cell cycle arrest (Figure 3A - mid graph). The addition of Rsv right after a 48 h treatment with TMZ did not alter the arrest induced by TMZ, suggesting that Rsv inhibits rather than reverses the cell cycle arrest induced by TMZ (Figure 3A – right). In U251 cells, Rsv induced a significant S arrest and slightly reduced TMZ-induced G2/M arrest, although the effect is less clear due to the S phase-arrest induced by Rsv. In U138, Rsv induced a S-phase arrest, but no clear TMZ-induced arrest was observed, which may be due to the high expression of MGMT (methyl-guanine methyl transferase) in these cells [42] (Additional file 4: Figure S4).

Since Rsv modulates proteins involved on DNA repair and interacts directly with DNA in vitro[43], we asked whether Rsv was abrogating cell cycle arrest due to a block of TMZ-induced DNA damage or DDR signaling. Quantification of DNA damage by comet assay indicated that Rsv did not reduce the TMZ-induced DNA damage (Figure 3B) nor DDR signaling, as indicated by the phosphorylation levels of H2AX, ATM and Chk2 (Figure 3C). Actually, RT treatment induced higher levels of gammaH2AX, pATM and pChk2 when compared to single treatments.

DDR signals to mitosis by activating Wee1 and inhibiting CDC25A, which are, respectively, negative and positive regulators of Cdk1 (Cdc2) [9, 44]. Wee1 phosphorylates whereas CDC25 dephosphorylates tyrosine 15 of Cdk1, and dephosphorylation of this site is required for the passage of G2 to M [45]. The increased levels of DDR signaling induced by RT (Figure 3C) does not translate into increased levels of pWee1, in which RT-treated cells have a lower level when compared to Rsv- or TMZ-treated cells, suggesting that Rsv acts as an inhibitor of the phosphorylation and activation of Wee1, leading to a reduced level of pCdk1 (Y15), (Figure 3D). Serine 10 of histone H3 is phosphorylated during mitosis and cells arrested at G2/M by TMZ showed an increase in the phosphorylation of this site whereas this was significantly reduced in RT treated cells. Surprisingly, levels of Cyclin B1, which has to be degraded for mitotic release, were found elevated in RT treated cells (Figure 3D). Noteworthy also is the observation that phospho(S807/811)-Rb (pRb), which is phosphorylated at the G1 to S transition [46] and dephosphorylated at the completion of M [47], is elevated with RT (Figure 3D). This reduction of pCdk1 (Y15) to mitotic permissive levels together with increased levels of Cyclin B1 leads to a molecular state which can lead to failed mitosis [48].

Cells with DNA damage that bypass cell cycle checkpoints tend to undergo MC [10, 49]. We observed that RT-treated cells presented alterations that suggest MC (Figure 4A and 4B), i.e. aberrant chromosome condensation and mitotic phenotype, micronuclei and nuclear fragmentation, and abnormal/triple mitosis.

We developed an objective morphometric analysis of nuclei size and irregularity [41], which showed that RT treatment increased the percentage of irregular nuclei, characteristic of MC, when compared to the individual treatments (Figure 4D), in agreement with direct counting of irregular nuclei (Figure 4C). It is also important to notice that the percentage of large nuclei, a feature of senescence induction, as previously described for another glioma cell line [24], was increased by TMZ, Rsv and RT treatments (Figure 4C and D).

To test the long term effect of RT treatment, which more closely correlates with in vivo data [50], U87 cells were treated for 48 h followed by Drug-Free Medium (DFM). After fourteen days, Rsv and TMZ reduced the number of colonies formed in 40% and 90%, respectively, in relation to control (Figure 5A, left graph). Furthermore, colonies formed from Rsv or TMZ-treated cells were much smaller when compared to untreated colonies. In RT-treated cells, in turn, no colonies were observed, but only individual, senescent-like cells (Figure 5A, right images).

TMZ induced senescence in around 50% of cells after 7 days in DFM, with the remaining cells showing normal morphology (Figure 5B). These cells may be responsible for the colonies observed after TMZ treatment (Figure 5A). RT induced a much higher proportion of SA-beta-gal-positive cells than treatments alone in both U87 (Figure 5B) and in primary GBM cells (Figure 5C). SA-beta-gal-negative cells were not observed in RT treated cultures (Figure 5B, insert), suggesting that Rsv could reduce the appearance of TMZ resistance.