Abstract
Glioblastoma (GB) is the most aggressive and common form of primary brain tumor characterized by fast proliferation, high invasion and resistance to current standard treatment. The average survival rate post-diagnosis is 14.6 months, despite the aggressive standard post-surgery radiotherapy concomitant with chemotherapy with temozolomide (TMZ). Currently, efforts are being endowed to develop new and more efficient therapeutic approaches capable to overcome chemoresistance, inhibit tumor progression and improve overall patient survival rate. Abnormal microRNA (miRNA) expression has been correlated with chemoresistance, proliferation and resistance to apoptosis, which result from their master regulatory role of gene expression. Altered cell metabolism, favoring glycolysis, was identified as an emerging cancer hallmark and has been described in GB, thus offering a new target for innovative GB therapies. In this work, we hypothesized that a gene therapy-based strategy consisting of the overexpression of a miRNA downregulated in GB and predicted to target crucial metabolic enzymes might promote a shift of GB cell metabolism, decreasing the glycolytic dependence of tumor cells and contributing to their sensitization to chemotherapy with TMZ. The increase of miR-200c levels in DBTRG cells resulted in downregulation of messenger RNA of enzymes involved in bioenergetics pathways and impaired cell metabolism and mobility. In addition, miR-200c overexpression prior to DBTRG cell exposure to TMZ resulted in cell cycle arrest. Overall, our results show that miR-200c overexpression could offer a way to overcome chemoresistance developed by GB cells in response to current standard chemotherapy, providing an improvement to current GB standard treatment, with benefit for patient outcome.
Introduction
The major therapeutic challenge of glioblastoma (GB), classified by the World Health Organization as a grade IV glioma and the most frequent malignant primary brain tumor, accounting for >50% of all gliomas (1), relies on location in the brain, which renders this type of tumor extremely difficult to remove without causing severe damage to the patient. The infiltrative nature of GB cells further enhances this difficulty, due to the formation of a peritumoral region, which comprises both tumor and normal brain cells. In parallel, the extremely fast growth rate of GB cells and the nearly universal relapse after tumor excision are features that contribute to low patient survival rates after GB diagnosis (2). These characteristics should be accounted for in the design of a suitable therapeutic approach to tackle GB. Currently, the standard treatment strategy for newly diagnosed GB starts with the maximum surgical resection of tumor tissue. However, due to GB infiltrative capacity, and the nearly impossible removal of the totality of tumor cells, the treatment also includes radiotherapy with concomitant or adjuvant chemotherapy with temozolomide (TMZ), along with symptomatic treatment (2). The current standard of care for GB hardly increases patient survival and typically results in death within 2 years after diagnosis (3).
GB displays the most malignant and aggressive phenotype of all gliomas, characterized by high cellular heterogeneity, with both differentiated and non-differentiated cancer cells and areas of necrotizing tissue, resistance to apoptosis, neoangiogenesis, vascular thrombosis and rapid proliferation (1,4). One hallmark of cancer that can be targeted as a means to achieve an overall decrease of all tumor cell functions is the altered energy metabolism. Metabolic profiles of cancer cells have long been described as abnormal, and GB cells are no exception (5,6). GB cells use glucose-derived carbon to produce lactate, even in the presence of normal oxygen concentrations (aerobic glycolysis) and fully functioning mitochondria, rather than to produce pyruvate to fuel the tricarboxylic acid (TCA) cycle and to generate ATP through mitochondrial oxidative phosphorylation (OXPHOS) (7). Aerobic glycolysis favors the acidification of the extracellular environment, which contributes to the degradation of the extracellular matrix and facilitates tumor cell migration and invasion into the adjacent healthy tissues (8,9). Furthermore, aerobic glycolysis supports cell survival and growth by providing substrate for synthesis of macromolecules, while conserving cell redox status (10,11).
Due to the heterogeneity of tumor cells and complexity of tumoral processes, a therapeutic approach based on a single target can be less than efficient. In fact, targeting more than one molecule or process can result in a therapeutic benefit larger than the mere sum of the individual effects, due to the interconnection of cellular pathways. In this regard, microRNAs (miRNAs) are extremely versatile regulatory elements, each miRNA having hundreds of different possible molecular targets in each cell type. MiRNAs are extremely important in the regulation of cellular gene expression and play a pivotal role in the most fundamental cellular processes, including cell differentiation, metabolism, proliferation and cell death in a wide range of organisms, including humans (12,13).
Expression levels of miR-200c in human GB tissue (a) and U87 and DBTRG cells (b). mRNA expression levels of miR-200c targets IDH1, PDHA1 and TIGAR in U87 (black bars) and DBTRG (gray bars) cells (c), with respect to normal human astrocytes (NHA). Fold change in miR-200c levels was estimated with respect to SNORD44 reference miRNA levels by qRT-PCR. The Pfaffl method was used to determine the expression levels of miRNA target genes, as compared with non-transfected cells, normalized to the housekeeping gene HPRT1. Data are presented as mean ± SD of at least three independent experiments. Pairwise data comparisons were performed between miR-200c or mRNA expression levels determined in U87 and DBTRG cells and those in NHA (**P < 0.01; ***P < 0.001; ns, not significant).
In cancer cells, dysregulated miRNAs have been shown to be involved in tumor initiation, progression and formation of metastases (14). GB tumors have previously been characterized in terms of their dysregulated miRNAs (13), and miRNA-based therapeutic approaches have been explored over the years and successfully applied in pre-clinical models for several human malignancies, GB being one of the most promising (12,15,16). Taking advantage of the potential of miRNA modulation to induce alterations in multiple hallmarks of cancer, the present work aims at combining therapeutic modulation of miRNAs implicated in GB pathogenesis with tumor cell exposure to a chemotherapeutic drug used as standard treatment in primary GB, in order to maximize the outcome of the anti-tumor therapy. Due to the role of the metabolic reprogramming in conferring drug resistance in GB, as well as in promoting tumor cell proliferation and migration ability, modulation of cell metabolic activity appears as a promising strategy to address GB treatment. In this work, miR-200c mimics were used to transfect two human GB cell lines, one described as ‘likely glioblastoma’ by the american type culture collection (ATCC) (U87) and extensively employed as a cellular model in GB studies and the other from a GB recurrent after chemo- and radiotherapy (DBTRG). MiR-200c was selected on the basis of our recent studies, which have shown that this miRNA is strongly downregulated in these two cell lines (17), similarly to what was previously described by others (18). In addition, miR-200c was predicted to potentially regulate messenger RNA (mRNA) expression of enzymes involved in metabolic activity, which were identified as predicted targets according to miRWalk 2.0 multiple predictors algorithm, including MicroT4, miRanda, RNA22 and RNAhybrid (http://zmf.umm.uni-heidelberg.de/apps/zmf/mirwalk2/). Although its role in cell metabolism has not been established, among the predicted targets of miR-200c, we identified hexokinase II (HKII), TP53-induced glycolysis and apoptosis regulator (TIGAR), pyruvate dehydrogenase E1, subunit α1 (PDHA1) and isocitrate dehydrogenase 1 (IDH1) (Supplementary Material, Fig. S1). HKII converts the uptaken glucose to glucose-6-phosphate (G6P), which then either participates in glycolysis with subsequent production of ATP or contributes to macromolecular biosynthesis of nucleotides, amino acids and fatty acids through the pentose phosphate pathway (PPP)-mediated production of ribose 5-phosphate, erythrose 4-phosphate and reducing equivalents nicotinamide adenine dinucleotide phosphate (NADPH), respectively (19). TIGAR, otherwise known as fructose-2,6-bisphosphatase, acts as an inhibitor of HKII activity, by dephosphorylating the glycolytic activator fructose-2,6-biphosphate (19,20). PDHA1 is one of the subunits of the pyruvate dehydrogenase complex, which decarboxylates pyruvate to produce acetyl-CoA that fuels the TCA cycle (21). IDH1 is the cytoplasmic IDH isoform, which converts isocitrate to α-ketoglutarate providing cellular NADPH.
Results
Analysis of expression of miR-200c and its target mRNAs in human GB tissue and cell lines
The expression levels of miR-200c, a miRNA previously shown to be implicated in the malignant phenotype of GB cells (22) and predicted by bioinformatics data bases to target mRNAs that encode proteins involved in cellular energy metabolism, was quantified in human GB tissue samples and U87 and DBTRG cells and compared with normal human astrocytes (NHA). MiR-200c was found to be significantly downregulated in all of the analyzed human GB samples (22/22) (Fig. 1a) as well as in both human GB cell lines (Fig. 1b), as compared with NHA.
The expression profile of genes whose mRNAs were predicted to be targeted by miR-200c, and that are known to be related with energy metabolism in GB, was assessed in U87 and DBTRG cells in comparison with NHA. Figure 1c shows the mRNA expression levels of IDH1 and PDHA1, genes associated with mitochondrial respiration, and TIGAR, which acts as a regulator of both glycolysis and the PPP in cancer cells. Although IDH1 gene expression was not altered in both types of GB cells, the mRNA levels of PDHA1 and TIGAR were found to be significantly upregulated in U87 cells, with respect to NHA. On the other hand, in DBTRG cells, PDHA1 was significantly downregulated, whereas TIGAR expression levels did not significantly change, with respect to NHA (Fig. 1c).
Expression levels of miR-200c (a, c) and of mRNA for PDHA1 and TIGAR (b, d) and of protein for TIGAR and HKII (e–h) in U87 (a, b, e, g) and DBTRG cells (c, d, f, h) after transfection with miR-ctrl (black bars) or miR-200c mimics (gray bars). Fold change in miRNA levels were estimated with respect to SNORD44 reference miRNA levels by qRT-PCR. The Pfaffl method was used to determine the expression levels of miRNA target genes, as compared with non-transfected cells, normalized to the housekeeping gene HPRT1. Protein levels were determined by western blot in protein extracts obtained from cell lysis in buffer containing protease inhibitors, normalized to β-actin levels. Protein levels are presented as fold change with respect to non-transfected control cells (NTC). Representative western blot images are depicted in g and h. Data are presented as mean ± SD of at least three independent experiments (nd, not detected). Pairwise data comparisons were performed between mRNA expression levels determined in U87 or DBTRG cells transfected with miRNA mimics and those in non-transfected control cells, and between protein expression levels determined in U87 or DBTRG cells transfected with miRNA-200c mimics and those in non-transfected control cells (*P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant).
Unfortunately, we were unable to design a suitable primer set for HKII, which encodes the enzyme HKII responsible for the first obligatory step of glycolysis, the production of G6P from glucose, whose mRNA is also a predicted direct target of miR-200c. Due to the lack of introns in HKII, all of the attempted primers failed to amplify only the mRNA, so that the results obtained were not reliable and, therefore, were not included in Figure 1c.
Analysis of protein levels of miR-200c targets after miRNA overexpression in GB cells
To determine the functional consequences of miR-200c overexpression, U87 and DBTRG cells were transfected with miRCURYLNA miR-200c mimics, resulting in increased miR-200c expression levels, which were not detected in non-transfected cells (NTC) or cells transfected with a non-targeting negative control miRNA (miR-ctrl) (Fig. 2a and c). In U87 cells, upregulation of miR-200c did not result in any decrease of the mRNA levels of its predicted targets TIGAR and PDHA1 (Fig. 2b). On the contrary, in DBTRG cells transfected with miR-200c mimics, the mRNA expression levels of both TIGAR and PDHA1 were substantially decreased, when compared with cells transfected with miR-ctrl (Fig. 2d). Regarding protein expression, miR-200c overexpression resulted in a significant increase of TIGAR levels in U87 cells (Fig. 2e and g). This unexpected result could be due to a compensatory mechanism activated in these cells in order to overcome miR-200c effects. In contrast, in DBTRG cells, TIGAR protein expression followed the same tendency observed for its mRNA levels, being strongly downregulated as a consequence of miR-200c overexpression (Fig. 2f and h). Western blot analysis of total HKII showed that, in U87 cells, miR-200c overexpression did not alter HKII expression (Fig. 2e and g), whereas in DBTRG cells, miR-200c overexpression resulted in a significant decrease of the levels of this protein (Fig. 2f and h).
Effect of miR-200c modulation on GB cell metabolism
To evaluate the effects of miR-200c overexpression on GB cell metabolism, cellular oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured in U87 and DBTRG cells, 48 h after miRNA modulation, using a XF24 Extracellular Flux Analyzer. In U87 cells, no alterations in oxygen consumption parameters were detected (Fig. 3a and d). Regarding extracellular acidification parameters, taken as proportional to the glycolytic activity, miR-200c overexpression significantly decreased glycolytic flux in resting U87 cells cultured in a medium containing glucose (Fig. 3b and e). In these cells, ATP production was strongly increased after miR-200c overexpression, both in normal conditions and upon treatment with oligomycin (inhibitor of ATP synthase) and 2-Deoxy-D-glucose (2-DG) (a glucose analog with a 2-hydroxy group replaced with hydrogen, which prevents glycolysis progression by inhibiting G6P production). The increased ATP levels observed in 2-DG-treated cells could be explained by the use of the glutaminolysis pathway, at the expense of glycolysis, in order to produce TCA cycle intermediates to fuel OXPHOS. Consistently, records of OCR after 2-DG addition registered an increase of OCR (Fig. 3d). Since oligomycin-induced ECAR stimulation in cells overexpressing miR-200c was not significantly different from that detected in control cells (non-treated or transfected with miR-ctrl), the ATP content increase in miR-200c-treated cells could not be assigned to a higher glycolytic capacity as compared with that of control cells. On the other hand, although oligomycin should stop ATP synthesis by blocking the F0 subunit of the proton channel ATP synthase, the concentration of that inhibitor, which was selected as the highest that did not exert significant toxicity (loss of cell viability <20%), might not have been high enough to fully inhibit the ATP synthase complexes and completely ablate energy production. In these conditions, one can assume that, upon miR-200c overexpression, the complexes that were not inhibited by oligomycin, i.e. those remaining active, were, most likely, stimulated to increase the ATP production rate (Fig. 3c).
Effect of overexpression of miR-200c (gray bars) or miR-ctrl (black bars) on U87 (a–e) and DBTRG (f–j) cell oxygen consumption rate (OCR) (a, d, f, i), extracellular acidification rate (ECAR) (b, e, g, j) and ATP production (c, h), as compared with non-transfected control cells (white bars, NTC). OCR and ECAR profiles, monitored using a Seahorse Bioscience XF24 Extracellular Flux Analyzer, are presented in d, i and e, j, respectively. Data were normalized to total protein content of each well. Metabolic parameters were calculated as described in the Materials and Methods section. ATP production was determined from cell lysates in a VICTOR Multilabel Plate Reader (Perkin Elmer) using a luciferase/luciferin luminescence assay and an ATP standard curve. ATP production was normalized to total protein content of each well. Data are presented as mean ± SD from three independent experiments. Pairwise data comparisons were performed for each parameter between cells transfected with miRNA-200c mimics and the respective non-transfected control cells (*P < 0.05; **P < 0.01; ***P < 0.001) or cells transfected with miR-ctrl (#P < 0.05; ##P < 0.01; ###P < 0.001).
In DBTRG cells, overexpression of miR-200c resulted in a statistically significant decrease in mitochondrial respiration in cells in basal conditions (prior to oligomycin addition), as compared both with NTC and cells transfected with miR-ctrl, which was accompanied by an increase in non-mitochondrial respiration (Fig. 3f). Maximum respiratory capacity, measured after carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) addition, decreased with miR-200c overexpression, indicating inhibition of the respiratory chain (Fig. 3f and i). Regarding ECAR measurements, only a small decrease in non-glycolytic acidification, albeit statistically significant, was observed in DBTRG cells overexpressing miR-200c, with respect to control conditions (NTC or miR-ctrl-transfected cell) (Fig. 3g and j). In these cells, ATP production increased with oligomycin treatment in the control conditions, indicating that the oligomycin-free ATP synthase complexes were stimulated to produce more ATP in order to overcome the energy depletion induced by this inhibitor. However, as opposed to what was observed in U87 cells, in DBTRG cells overexpressing miR-200c and treated with oligomycin, ATP production was severely impaired, probably due to inhibition of the respiratory chain by miR-200c overexpression, as indicated by the decreased maximal respiratory capacity (Fig. 3f and i). A similar effect on ATP content was found in DBTRG cells treated with 2-DG (Fig. 3h), consistent with the observation that this inhibitor did not induce a significant change in OCR in miR-200c overexpressing cells.
GB cell dependence on glucose/glutamine supply
DBTRG cells revealed to be much more metabolically active than U87 cells, in basal conditions (prior to oligomycin addition), when grown in complete medium containing glucose and L-glutamine, as assessed by the determination of OCR and ECAR normalized for the total protein amount (Fig. 4a and b). Further investigation regarding the dependency of U87 and DBTRG cells on the supply of glucose and glutamine for their growth showed that both cell lines thrived when both nutrients were present in the media, although being completely dependent on glutamine to proliferate (Fig. 4c and e). In contrast, growth curves of cells cultured in the presence of glutamine and absence of glucose showed patterns similar to those of cells grown in complete medium up until day 5 (U87) and day 4 (DBTRG). Only after these time points, the growth curve changed to indicate a decrease in cell density, suggesting a process of cell death. This apparent death event could be caused by depletion of glutamine from the medium after this period of time, since the media were not replaced by fresh media during the course of these experiments. Results of cell density determined 48 h after overexpression of miR-200c in cells grown in complete medium or in medium lacking glucose or glutamine showed that, in the absence of glutamine, miR-200c was able to reduce the viability of both U87 and DBTRG cells (Fig. 4d and f). This supports the hypothesis that miR-200c acted over the glycolytic and respiratory pathways, which are replaced by glutaminolysis, when glutamine is available, thus generating cellular energy to fulfill the energetic and biosynthetic needs of the tumor cells.
Metabolic activity of U87 and DBTRG cells in terms of oxygen consumption rate (OCR) (a) and extracellular acidification rate (ECAR) (b), effect of glucose and glutamine deprivation on U87 (c) and DBTRG (e) cell growth and effect of miR-200c overexpression on the density of U87 (d) and DBTRG (f) cells grown in complete medium and in medium lacking glucose or glutamine. Metabolic activity (a and b) was monitored in a Seahorse Bioscience XF24 Extracellular Flux Analyzer. OCR and ECAR values correspond to the baseline oxygen consumption and extracellular acidification of cells in resting state. Cell growth (c and e) was assessed by SRB assay over time (black circles: cells grown in complete medium; open squares: cells grown in medium with glucose but lacking glutamine; gray triangles: cells grown in medium with glutamine but lacking glucose). Cell density (d and f) was determined by SRB assay 48 h after transfection in cells grown in the same media and transfected with miR-200c mimics (gray bars) or with a miR-ctrl (black bars). Data are presented as mean ± SD of three independent experiments. Pairwise data comparisons were performed for cell density between cells transfected with miR-200c mimics and with miR-ctrl (**P < 0.01; ***P < 0.001).
Effect of miR-200c overexpression on GB cell mobility
In order to determine the outcome of miR-200c overexpression in GB cell mobility, U87 and DBTRG cells were transfected with miR-ctrl or miR-200c mimics and their mobility rate and traveled distance were assessed using μ-Slide Chemotaxis, ibiTreated chambers (Ibidi, Germany). Cell mobility was evaluated over a 14 h period with photographs taken every 5 min, which were used to determine the pathway described by each cell. Overexpression of miR-200c was unable to alter the mobility capacity of U87 cells (Fig. 5a and b). However, DBTRG cells transfected with miR-200c mimics showed significantly impaired mobility, both regarding traveled distance (displacement of center of mass) (Fig. 5c) and mobility rate (Fig. 5d), as compared with cells transfected with miR-ctrl and non-transfected control cells.
Effect of miR-200c overexpression on the mobility of U87 (a, b) and DBTRG (c, d) cells. Results of cell mobility rate (a, c) and displacement of the center of mass (b, d) represent the mean ± SD obtained from two independent experiments, in which at least 20 cells were followed per experiment. Pairwise data comparisons were performed between cells overexpressing miR-200c and non-transfected control cells (NTC) or cells transfected with miR-ctrl, and between non-transfected control cells (NTC) and cells transfected with miR-ctrl (*P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant).
Effect of miR-200c overexpression on GB cell sensitization to TMZ
Although a decrease in the proliferation of GB cells overexpressing miR-200c was observed only when they were grown in medium lacking glutamine, we hypothesized that miR-200c overexpression could result in a potential anti-tumor effect, albeit small, even in cells cultured in the presence of glutamine, which might be enhanced by cell exposure to other stressors, namely TMZ, the standard first-line chemotherapeutic agent for GB. In this regard, studies addressing the effect of miR-200c overexpression on cell cycle progression were performed by analyzing the DNA content of U87 and DBTRG cells transfected with miR-200c mimics, as compared with those NTC or transfected with miR-ctrl, through flow cytometry. As shown in Figure 6, ~60% of U87 cells and 70% of DBTRG cells were in G0/G1 phase under control conditions (NTC or transfected with miR-ctrl). The remaining percentage of U87 cells was distributed evenly between S and G2/M phases (Fig. 6a), whereas the remaining DBTRG cells were mostly in the S phase (25%), with only a residual fraction of the cells in the G2/M phase (Fig. 6b). The quantitative DNA analysis revealed no significant alterations in U87 or DBTRG cells transfected with the miR-ctrl, as compared with non-transfected control cells (Fig. 6a and b). Overexpression of miR-200c resulted in a decrease of the percentage of U87 and DBTRG cells in G0/G1 phase, which was accompanied by a small increase of the percentage of U87 cells in G2/M phase, although not statistically significant (Fig. 6a), and a significant increase of the percentage of DBTRG cells in S phase (Fig. 6b). Incubation of U87 and DBTRG cells with TMZ (400 μM) resulted in cell cycle arrest in G2/M and S phases accompanied by a decrease of cells in G0/G1 phase (Fig. 6a and b). Importantly, the effect of TMZ on cell cycle arrest in the S phase was significantly enhanced by previous miR-200c overexpression in both types of GB cells (Fig. 6a and b), although being less expressive in U87 cells co-treated with miR-200c mimics and TMZ (Fig. 6a).
Effect of miR-200c overexpression and/or incubation with 400 μM TMZ on cell cycle progression (a, b), cell death (c, d) and invasion ability (e, f) of U87 (a, c, e) and DBTRG (b, d, f) cells. DNA content was evaluated by flow cytometry and the percentage of cells in each cell cycle phase was determined using the ModFit software (G0/G1, black; S, dark gray; G2/M, light gray). The percentage of cells labeled with annexin-FITC and/or PI was determined by flow cytometry and the percentage of viable cells (annexin-negative/PI-negative), apoptotic cells (annexin-positive/PI-negative), dead cells (annexin-positive/PI-positive) and cell debris (annexin-negative/PI-positive) was calculated from the cells in each quadrant of a FITC/PI dot plot using CellQuest software. The invaded area was calculated from images of the spheroids, taken 1 and 4 days after embedding in a collagen matrix in a Axio Observer Z1 widefield microscope. The results represent the mean ± SD obtained from three independent experiments. For each cell cycle phase and quadrant of cells labeled with annexin-FITC/PI, pairwise comparisons were performed between cells transfected with miR-200c and miR-ctrl (*P < 0.05; **P < 0.01; ***P < 0.001), between cells transfected with miR-ctrl and the respective non-transfected control (not significant), and between cells transfected with miR-ctrl and incubated with TMZ and cells transfected with miR-200c and incubated with TMZ (#P < 0.05; ##P < 0.01; ###P < 0.001). For each time point of cell invasion, pairwise comparisons were performed between cells transfected with miR-200c and cells transfected with miR-200c and subsequently treated with TMZ (**P < 0.01), and between cells treated with TMZ and cells transfected with miR-200c and subsequently treated with TMZ (##P < 0.01).
To address the possibility that the combined treatment encompassing miR-200c overexpression and TMZ incubation might induce cell death, flow cytometry analysis of cells labeled with annexin-fluorescein isothiocyanate (FITC) and/or propidium iodide (PI) was performed. As observed in Figure 6, the majority of untreated cells were viable (90% of U87 cells and 80% of DBTRG cells) (Fig. 6c and d). Single treatment, either involving miR-200c overexpression or TMZ incubation, did not significantly alter annexin or PI labeling of U87 and DBTRG cells. However, the combined treatment of miR-200c overexpression plus TMZ incubation resulted in a significant decrease of the percentage of unlabeled cells and increase of the population labeled both with annexin and PI. These findings indicate that, besides cell cycle arrest, coupling miR-200c overexpression with TMZ treatment induced GB cell death, apparently by an apoptosis-independent mechanism (Fig. 6c and d).
Remarkably, in addition to inducing cell cycle arrest and cell death, miR-200c overexpression plus TMZ incubation was able to significantly reduce DBTRG cell ability to invade a polymeric matrix (Fig. 6f). As shown, DBTRG cells grown as spheroids, transfected with miR-200c mimics and subsequently incubated with TMZ, invaded a significantly smaller area after 1 or 4 days of treatment, than untreated DBTRG spheroids or DBTRG spheroids transfected with miR-200c mimics (or with miR-ctrl) alone or treated only with TMZ (Fig. 6f). Such effect, however, was not observed in U87 cells (Fig. 6e).
Discussion
The present work outlines the impact of miR-200c overexpression on GB cell energy metabolism, contributing to decrease GB cell viability in glutamine-deprivation conditions, and promoting cell cycle arrest in TMZ-exposed GB cells. Furthermore, in GB cells unable to overcompensate the metabolic impairment induced by miR-200c overexpression (a process dependent on TIGAR expression), cell mobility was also compromised and cell invasion was severely affected by the combined therapy, encompassing miRNA modulation and TMZ treatment.
MiR-200c was previously reported to be downregulated in cerebrospinal fluid samples of patients diagnosed with GB (22), which is consistent with our results, both in the 22 human GB tumor samples analyzed (Fig. 1a) and in the two human GB cell lines studied (Fig. 1b). Although the predicted targets of miR-200c include a series of metabolic enzymes (HKII, IDH1, TIGAR, PDHA1), to the best of our knowledge, its role in cell metabolism has not been established yet. Our results indicate that miR-200c overexpression in DBTRG cells was able to downregulate both PDHA1 and TIGAR (Fig. 2d), which are involved in OXPHOS and glycolysis regulation, respectively. PDHA1 acts as a linker between glycolysis and the TCA cycle, by converting pyruvate produced from glycolysis into acetyl-CoA to fuel the TCA towards the generation of energy and macromolecular intermediates (23–25). In fact, PDHA1 was described as the major provider of carbon for the TCA in GB (26). In cancer cells, the TCA may be decoupled from OXPHOS to provide carbon intermediates for the anabolic reactions necessary for the production of building blocks in order to support cell growth rather than ATP synthesis (27). The decrease of PDHA1 levels in DBTRG cells may have disturbed this process, thus activating a compensatory mechanism relying on the use of reductive glutamine metabolism to generate acetyl-CoA for fatty acid production (7). This compensatory mechanism may be on the basis of the observed reduction in the viability of DBTRG cells overexpressing miR-200c and cultured in medium lacking glutamine (Fig. 4f). On the other hand, the glycolysis inhibitor and PPP promoter TIGAR was shown to upregulate mitochondrial respiration, improving the energy production from glucose, and to promote the PPP at the expense of glycolysis, thus protecting cells from reactive oxygen species (ROS)-induced DNA damage and preventing tumor development (28). Thus, the strong TIGAR downregulation as a consequence of miR-200c overexpression in DBTRG cells might explain the impaired mitochondrial respiration, as indicated by the decreased ATP-coupled respiration and inhibition of respiratory chain. The decreased supply of metabolic intermediates to the PPP reduces the ability of DBTRG cells to produce glutathione and, consequently, to cope with oxidative stress (20,28). Besides the possible sensitivity that TIGAR downregulation may confer to other DNA damage-inducing agents, such as TMZ (Fig. 6b and f), the decrease of metabolic intermediates available to the PPP also inhibits nucleotide synthesis (29), which prevents the correct duplication of nuclear material during the S phase of the cell cycle, and may explain S phase cell cycle arrest induced by miR-200c overexpression and enhanced by TMZ incubation (Fig. 6b).
Contrary to what was observed in DBTRG cells, in U87 cells, PDHA1 and TIGAR mRNA levels remained unchanged and TIGAR protein levels were strongly increased after transfection with miR-200c mimics (Fig. 2b and e), which could be a response of these cells to miR-200c overexpression. Consequently, the effect of miR-200c overexpression on the energy metabolism was very mild in U7 cells (Fig. 3a and b), resulting in a decrease of the glycolytic flux (Fig. 3b), with a strong increase in ATP production by miR-200c-overexpressing cells (Fig. 3c), which is consistent with the described effect of TIGAR in increasing mitochondrial respiration (28). Although in agreement with the glycolytic flux results, the increase in TIGAR protein levels cannot be attributed to a direct miR-200c target of TIGAR mRNA. In fact, the mRNA levels of TIGAR were not affected by miR-200c overexpression, which suggests that TIGAR is not a preferential target of miR-200c in this cell type. However, regulators of TIGAR through p53, not explored in this work but which are other miR-200c targets, could be involved in such effect. One hypothesis is that miR-200c may be targeting the hypoxia inducible factor 1 subunit α inhibitor, a predicted target of this miRNA. This would result in a decrease of the inhibition of hypoxia-inducible factor 1-alpha (HIF-1α), which would lead to the increased expression of junction-mediating and regulatory protein (JMY). JMY is a p53 cofactor, thus, its increased availability could result in increased p53 activity with induction of TIGAR (30). Another more direct effect could be the one mediated by MDM4 regulator of p53 (MDM4), which is also a predicted target of miR-200c and has a p53-binding domain that inhibits p53. Therefore, by miR-200c-mediated inhibition of MDM4, p53 would present increased activity, which could lead to increased TIGAR levels (31).
DBTRG cells were found to be much more metabolically active than U87 cells, as assessed both in terms of OCR and ECAR (Fig. 4a and b). One possible explanation for the apparent resistance of U87 cells to miR-200c overexpression regarding potential alterations in cell metabolism is that some metabolic hallmarks of U87 cells may reflect the long-term culture adaptation of this cell line to optimal culturing conditions rather than malignancy (32). In fact, the higher sensitivity to energy metabolism-targeting therapeutic strategies of DBTRG cells as compared with U87 cells may be a consequence of the higher-grade malignancy of DBTRG cells, which would be expected for a cell line obtained from a recurrent GB after local brain irradiation and multidrug chemotherapy. Chemoresistance was previously associated with elevated glycolysis, contributing to the activation of DNA repair mechanisms, namely non-homologous end joining and homologous recombination (33). Although U87 and DBTRG cells presented similar growth rates in normal medium (containing both glucose and glutamine), in the absence of glucose, U87 cells took longer than DBTRG cells to consume all the glutamine present in the medium, to give rise to a progressive loss of cell mass, probably reflecting dying cells (Fig. 4c and e). However, in medium containing glucose but lacking glutamine, both cell lines were unable to proliferate, indicating a strong dependence on the glutaminolysis pathway for thriving. Only in the absence of glutamine, it was possible to observe a decrease of cell viability as a consequence of miR-200c overexpression, in both cell lines (Fig. 4d and f), which is consistent with GB cell dependency on glutamine and reductive IDH1-mediated flux for TCA supply towards the production of building blocks for cell growth (7).
MiR-200c was previously suggested as a regulator of migration and invasion processes in various types of cancer (34–36). In fact, our results show that miR-200c overexpression was able to decrease DBTRG cell mobility (Fig. 5c and d). Similarly, in GB cells grown as spheroids, which better mimic the tridimensional nature and the cell–cell interactions and microenvironment of GB solid tumors (37), the combination of transfection with miR-200c mimics and treatment with TMZ reduced DBTRG cell invasion ability to a higher extent than any of the treatments per se (Fig. 6f). The observed effect of miR-200c on the mobility of DBTRG cells could be attributed to miR-200c modulation of E-cadherin levels, and consequent repression of epithelial–mesenchymal transition (EMT), the driving force of tumorigenesis, through regulation of zinc-finger E-box-binding homeobox (ZEB) family of transcription factors (38,39). In this regard, ZEB1 and ZEB2, the two major members of the ZEB family, are direct targets of the miR-200 family of miRNAs and strong suppressors of E-cadherin expression (35). In addition, moesin (MSN), a protein belonging to the ERM family, which acts as bridge between membrane proteins and actin cytoskeleton and is also a direct target of miR-200c, has been suggested as responsible for miR-200-c-mediated decrease of GB cell growth and invasion (18). Impairment of EMT, a process in which a polarized epithelial cell acquires cell motility and resistance to genotoxic agents, translating into a more aggressive tumor phenotype (40) may explain the sensitization of DBTRG cells to TMZ, in terms of cell cycle arrest, cell death and invasion ability (Fig. 6b, d and f). However, alterations of cell mobility and invasion abilities were not observed in U87 cells transfected with miR-200c mimics and incubated with TMZ (Figs 5a and b and6e ). The higher baseline expression levels of TIGAR in U87 cells, as compared with NHA, and the absence of a statistical difference between TIGAR levels in DBTRG cells and NHA (Fig. 1c), support the protective role mediated by TIGAR in U87 cells (28). Figure 7 depicts a possible mechanism for miR-200c effect on DBTRG cell proliferation and migration/invasion.
Schematic diagram showing the molecular and cellular consequences of miR-200c overexpression on malignant glioma cells. MiR-200c targeting of TIGAR results in inhibition of the PPP, thus reducing the production of aromatic amino acids through erythrose-4-phosphate, and of nucleic acids, via ribose-5-phosphate. In addition, PPP inhibition also prevents the recycling of glutathione due to the decreased availability of NADPH necessary for this process. In parallel, PDHA1 downregulation reduces mitochondrial respiration, which may be a consequence of reductive glutamine metabolism, fueling the TCA cycle to run towards the production of acetyl-CoA for fatty acid synthesis. In the absence of glutamine, fatty acid production becomes impaired, which contributes to decreased cell growth. On the other hand, miR-200c targeting of ZEB leads to upregulation of E-cadherin, thus reducing tumor aggressiveness, namely cell mobility and resistance to the genotoxic agent TMZ, characteristic features of cells undergoing EMT.
Conclusion
The extremely complex nature of GB makes this type of tumor a specially challenging therapeutic target. However, the increasing amount of knowledge gathered over the years regarding the specific hallmarks displayed by malignant brain tumors has helped unravel new promising molecular targets in anticancer therapies, namely towards GB. The metabolic profile of GB appears as a critical knot linking a number of major cancer phenotypes, including fast and uncontrolled cell proliferation and high rates of cell migration and invasion. Thus, targeting key enzymes of the metabolic pathways can have an important impact on several properties of cancer cells and contribute to an efficient treatment strategy. In this regard, modulation of miRNAs, which are able to interfere with several cellular pathways, including the metabolic railway node, through their multiple molecular targets, in combination with currently used chemotherapeutic drugs may constitute a multidimensional and effective therapy against GB. Overall, our observations indicate that the overexpression of miR-200c, a miRNA found to be downregulated in GB and predicted to target mRNAs of crucial enzymes involved in cellular metabolism, resulted in metabolic alterations and significant impairment of the migration capacity of the recurrent GB DBTRG cells. Importantly, sequential treatment with miRNA mimics and TMZ, the first-line drug for GB, repressed the invasion capacity of DBTRG cells when grown as spheroids. Furthermore, cell cycle analysis of DBTRG cells pointed to an arrest in S phase induced by miR-200c, which was reinforced upon addition of TMZ. Thus, overexpression of miR-200c emerges as a promising approach to improve the therapeutic effect of TMZ in GB; hence, deserving to be explored towards an effective anti-tumor therapy.
Materials and Methods
Cell lines and culturing conditions
The U-87 MG (U87) human glioma cell line, kindly provided by Dr Peter Canoll (Columbia University, New York, NY), was maintained in dulbecco's modified eagle medium, high glucose (DMEM-HG) (Sigma, D5648), supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS) (Gibco, Paisley, Scotland), 100 U/ml penicillin (Sigma-Aldrich), 100 μg/ml streptomycin (Sigma-Aldrich), 10 mmol/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) and 12 mmol/L sodium bicarbonate. The DBTRG-05MG (DBTRG) human recurrent GB cell line, established from a 59 years Caucasian female patient with GB treated with local brain irradiation and multidrug chemotherapy and kindly provided by Dr Massimiliano Salerno (Siena Biotech, Italy), was maintained in RPMI-1640 (Sigma-Aldrich, R4130), supplemented with 10% heat-inactivated FBS (Gibco), 100 U/ml penicillin (Sigma-Aldrich), 100 μg/ml streptomycin (Sigma-Aldrich) and 12 mmol/L sodium bicarbonate. NHA, kindly provided by Dr Anne Régnier-Vigouroux (Johannes Gutenberg Universität-Mainz), were maintained in DMEM–HG (Sigma-Aldrich, D5648), supplemented with 2% heat-inactivated FBS, 100 U/ml penicillin, 100 μg/ml streptomycin and 10 mmol/L sodium bicarbonate. NHA medium was freshly supplemented with 1% N2 (Sigma-Aldrich) and 1% non-essential amino acids (Sigma-Aldrich). The cells were cultured at 37°C under a humidified atmosphere containing 5% of CO2 and grown adherent, being detached upon addition of enzyme-free dissociation buffer.
Patient tumor samples
A total of 22 GB samples from the Tumor Bank of the Centro Hospitalar e Universitário de Coimbra were evaluated regarding the levels of miR-200c. Total RNA, including miRNAs, was isolated from frozen tumor samples (ca. 10 mg) with the AllPrep DNA/RNA/miRNA Universal Kit (Qiagen) according to manufacturers’ instructions for purification of total RNA from tissues.
Lipoplex preparation
MiRIDIAN hsa-miR-200c-3p (miR-200c) mimics and the non-targeting miRNA control (miR-ctrl) (Dharmacon) (Table 1) were formulated into lipoplexes prepared from delivery liposomal system (DLS) liposomes, as previously described (41). DLS/miRNA lipoplexes were prepared freshly for every experiment by mixing each miRNA mimic with the appropriate volume of DLS suspension, previously diluted in OPTIMEM, to achieve the ratio of 95 μg DLS/10 μg miRNA and the final miRNA concentration of 50 nM per well. The complexes were incubated for 30 min at room temperature prior to addition to the cells.
MiRNA mimics sequences
| Sequence (5′–3′) | |
|---|---|
| Mimic control (miR-ctrl) | UUCUCCGAACGUGUCACGUdTdT |
| hsa-miR-200c-3p mimic (miR-200c) | UAAUACUGCCGGGUAAUGAUGGA |
| Sequence (5′–3′) | |
|---|---|
| Mimic control (miR-ctrl) | UUCUCCGAACGUGUCACGUdTdT |
| hsa-miR-200c-3p mimic (miR-200c) | UAAUACUGCCGGGUAAUGAUGGA |
MiRNA mimics sequences
| Sequence (5′–3′) | |
|---|---|
| Mimic control (miR-ctrl) | UUCUCCGAACGUGUCACGUdTdT |
| hsa-miR-200c-3p mimic (miR-200c) | UAAUACUGCCGGGUAAUGAUGGA |
| Sequence (5′–3′) | |
|---|---|
| Mimic control (miR-ctrl) | UUCUCCGAACGUGUCACGUdTdT |
| hsa-miR-200c-3p mimic (miR-200c) | UAAUACUGCCGGGUAAUGAUGGA |
Cell transfection
For quantitative real-time polymerase chain reaction (qRT-PCR) and western blot analysis, cell mobility and flow cytometry experiments, cells were plated onto 12-well plates (Costar) at a density of 6 × 104 cells/well. For cellular bioenergetics analysis, U87 and DBTRG cells were seeded at a density of 6 × 103 and 5 × 103 cells/well, respectively, onto XF24 Cell Culture Microplates (Seahorse Bioscience). For cell density analysis, U87 and DBTRG cells were seeded at a density of 6 × 103 and 5 × 103 cells/well, respectively, onto 96-well plates (Costar), and the plate border wells were filled with sterile water. One day after plating, complete medium was replaced with OPTIMEM (Gibco) and freshly prepared lipoplexes were added to the cells at a final concentration of 50 nM per well. After 4 h of incubation in a humidified CO2 incubator at 37°C, OPTIMEM medium was replaced with complete DMEM (U87) or RPMI (DBTRG), or DMEM or RPMI lacking glucose or glutamine (analysis of cell density in the absence of glucose or glutamine supply).
For spheroid transfection, cells were plated onto Corning Costar Ultra-Low Attachment (ULA) 96-well plates with round bottom. Briefly, 3 × 103 U87 cells in 100 μl of complete DMEM medium and 3 × 103 DBTRG cells in 100 μl of complete RPMI medium were plated in each well, and the ULA plates were centrifuged for 2 min at 200g in a bench centrifuge. After the cells have grown into spheroids (3 days for U87 and 4 days for DBTRG cells) at 37°C under a humidified atmosphere containing 5% of CO2, half the medium was removed from each well and the spheroids were transfected with miRNA mimics formulated into complexes with the commercially available reagent Lipofectamine RNAiMax, at a Lipofectamine RNAimax/miRNA mimic ratio of 1.5 μl Lipofectamine RNAimax/30 pmol miRNA, according to supplier’s instructions. After incubation at room temperature for 20 min, the complexes were added to the spheroids to achieve a miRNA concentration of 50 nM in each well. After 4 h at 37°C, the spheroids were collected for invasion assays.
Drug incubation
TMZ (Temodar, Merck) was purchased from Selleckchem. TMZ was added to the cells, to achieve the final concentration of 400 μM, 4 h after transfection, the time point at which the transfection medium was replaced with fresh medium. Cells were incubated with TMZ for 44 h in a humidified CO2 incubator at 37°C, prior to further analysis.
Total RNA extraction and complementary DNA synthesis
The miRCURY Isolation Kit (Exiqon) was used for total RNA, including miRNAs, extraction from cells, according to manufacturer’s instructions for cultured cells. RNA was quantified using the NanoDrop 2000 Spectrophotometer (Thermo Scientific). Synthesis of complementary DNA (cDNA) for miRNA quantification was performed from 10 ng of total RNA in a 10 μl reaction using the Universal cDNA Synthesis Kit (Exiqon), according to the suppliers’ instructions. The obtained cDNA was then diluted 40 times with RNase-free water and stored at −20°C. For mRNA quantification, cDNA synthesis was performed using 500 ng of total RNA in a 10 μl reaction, with the NZY First-Strand cDNA Synthesis Kit (NZYtech, Lisbon, Portugal), according to the protocol: 10 min at 25°C, 30 min at 50°C and 5 min at 85°C, following digestion of the RNA template in cDNA:RNA hybrids with NZY RNase H (Escherichia coli) for 20 min at 37°C. The produced cDNA was diluted 20 times with RNase-free water and stored at −20°C.
qRT-PCR
Quantitative PCR was performed in a StepOnePlus Thermocycler (Applied Biosystems) using 96-well microtiter plates. The miRCURRY LNA TM Universal RT miRNA PCR system (Exiqon) was employed for miRNA quantification, using SYBR Green Master Mix (Exiqon), with primers from Exiqon, including the reference gene small nucleolar RNA, C/D box 44 (SNORD44). For miRNA quantification, qPCR was performed according to the supplier’s protocol, with reactions performed in duplicate. For mRNA quantification, the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) was used. The primers for the tested genes were designed using the bioinformatics primer designing tool Primer-BLAST and purchased from Invitrogen. The reference gene hypoxanthine phosphoribosyltransferase 1 (HPRT1) was used as internal normalizer (Table 2). The reaction conditions consisted of 1.5 min at 95°C, followed by 45 cycles of 10 s at 95°C, 30 s at 55–60°C and 30 s at 72°C. Each reaction was performed in duplicate and a final concentration of 1 μM for each primer pair was used. The No Template Control and the No Reverse Transcriptase Control were assessed, for each primer set, in all experiments. The threshold value for threshold cycle determination was defined as 10 000 and the baselines adjusted for each sample. Fold changes of miRNAs and mRNA levels were determined according to the Pfaffl method, with respect to the house-keeping genes SNORD44 (miRNAs) and HPRT1 (mRNA), taking into consideration primer amplification efficiency (80–120%), according to the formula: E = 10–1/S, where S is the slope of the obtained standard curve.
Primer sequences used for qPCR analysis and respective amplicon sizes
| Gene | Primer sequence (5′–3′) | Amplicon | |
|---|---|---|---|
| HPRT1 | Forward | CTGCGTAACTCCATCTGA | 103 bp |
| Reverse | ACCGTAATTGGCATCGT | ||
| IDH1 | Forward | ATGGTGACGTGCAGTCGG | 76 bp |
| Reverse | GGACAAACCAGCACGCT | ||
| PDHA1 | Forward | TGGAGTCAGTTACCGTACACGAG | 107 bp |
| Reverse | CCACACTGGCAAGATTGCTG | ||
| TIGAR | Forward | CTGACTGAAACTCGCTAAGG | 105 bp |
| Reverse | CAGAACTAGCAGAGGAGAGA |
| Gene | Primer sequence (5′–3′) | Amplicon | |
|---|---|---|---|
| HPRT1 | Forward | CTGCGTAACTCCATCTGA | 103 bp |
| Reverse | ACCGTAATTGGCATCGT | ||
| IDH1 | Forward | ATGGTGACGTGCAGTCGG | 76 bp |
| Reverse | GGACAAACCAGCACGCT | ||
| PDHA1 | Forward | TGGAGTCAGTTACCGTACACGAG | 107 bp |
| Reverse | CCACACTGGCAAGATTGCTG | ||
| TIGAR | Forward | CTGACTGAAACTCGCTAAGG | 105 bp |
| Reverse | CAGAACTAGCAGAGGAGAGA |
Primer sequences used for qPCR analysis and respective amplicon sizes
| Gene | Primer sequence (5′–3′) | Amplicon | |
|---|---|---|---|
| HPRT1 | Forward | CTGCGTAACTCCATCTGA | 103 bp |
| Reverse | ACCGTAATTGGCATCGT | ||
| IDH1 | Forward | ATGGTGACGTGCAGTCGG | 76 bp |
| Reverse | GGACAAACCAGCACGCT | ||
| PDHA1 | Forward | TGGAGTCAGTTACCGTACACGAG | 107 bp |
| Reverse | CCACACTGGCAAGATTGCTG | ||
| TIGAR | Forward | CTGACTGAAACTCGCTAAGG | 105 bp |
| Reverse | CAGAACTAGCAGAGGAGAGA |
| Gene | Primer sequence (5′–3′) | Amplicon | |
|---|---|---|---|
| HPRT1 | Forward | CTGCGTAACTCCATCTGA | 103 bp |
| Reverse | ACCGTAATTGGCATCGT | ||
| IDH1 | Forward | ATGGTGACGTGCAGTCGG | 76 bp |
| Reverse | GGACAAACCAGCACGCT | ||
| PDHA1 | Forward | TGGAGTCAGTTACCGTACACGAG | 107 bp |
| Reverse | CCACACTGGCAAGATTGCTG | ||
| TIGAR | Forward | CTGACTGAAACTCGCTAAGG | 105 bp |
| Reverse | CAGAACTAGCAGAGGAGAGA |
Western blot analysis
Total protein extracts were obtained from U87 and DBTRG cells using lysis buffer (50 mm NaCl, 50 mm ethylenediaminetetraacetic acid (EDTA) and 1% Triton X-100) supplemented with protease inhibitor cocktail (Sigma), 10 μg/ml 1,4-dithiothreitol (DTT) and 1 mM phenylmethylsulfonyl fluoride (PMSF). Extracts were subjected to three freeze–thaw cycles, and protein was quantified with the Bio-Rad DC protein assay (Bio-Rad). For each sample, 60 μg (U87) or 20 μg (DBTRG) of total protein was resuspended in loading buffer (20% glycerol, 10% sodium dodecyl sulfate (SDS) and 0.1% bromophenol blue), denaturated for 5 min at 95°C and loaded onto a 10% polyacrylamide gel. After electrophoresis, the proteins were blotted onto PVDF membranes for 2 h at 1 A. After blocking for 1 h in 5% non-fat milk, membranes were incubated overnight at 4°C with the appropriate primary antibody (anti-HKII 1:1000, anti-TIGAR 1:1000, anti-LC3B 1:1000, Cell Signaling), and, subsequently, with the appropriate horseradish peroxidase (HRP) secondary antibody (1:10 000, anti-rabbit, Invitrogen) for 2 h at room temperature. Results were normalized to β-actin (anti-β-actin primary antibody, 1:10 000, Sigma) and with the appropriate secondary antibody. Membranes were revealed by incubation with enhanced chemiluminescence substrate for 5 min and images taken in ChemiDoc (BioRad) were analyzed with Image-Lab software (BioRad).
Cellular bioenergetics analysis
OCR and ECAR were measured in U87 and DBTRG cells using a XF24 Extracellular Flux Analyzer, as previously described (17). Cells were transfected according to the protocol described above for 96-well plates. Briefly, the plate was calibrated according to manufacturer’s instructions previous to the experiment. Solutions of oligomycin (1 μM), FCCP (0.3 μM), rotenone (1 μM) and 2-DG (1.2 M) were prepared in XF Base Medium supplemented with 25 mM glucose (Sigma) and 4 mM L-glutamine (Sigma) (U87 cells) or with 2 mM L-glutamine (Sigma) (DBTRG cells). Bioenergetics parameters were calculated from OCR measurements (Fig. 8a): mitochondrial respiration was taken as the difference between OCR of resting cells and OCR after rotenone injection; ATP-coupled respiration was considered the difference between OCR of resting cells and OCR after oligomycin injection; proton leak was evaluated as the difference between OCR after oligomycin injection and rotenone injection; maximum respiratory capacity and non-mitochondrial respiration were determined as the average of three measurements after FCCP and rotenone injection, respectively. Bioenergetic parameters determined from ECAR measurements (Fig. 8b) were glycolytic flux, determined as the difference between ECAR of resting cells and after 2-DG injection; glycolytic capacity, calculated as the difference between ECAR measurements after oligomycin stimulation and after 2-DG inhibition; glycolytic reserve determined as the difference between ECAR after oligomycin injection and ECAR of resting cells; and non-glycolytic acidification, determined after 2-DG injection.
Graphical representation of the bioenergetics parameters extrapolated from OCR (a) and ECAR (b) records (see text for details) obtained upon sequential injection of oligomycin, FCCP, 2-DG and rotenone in the XFe Seahorse Bioanalyzer.
OCR and ECAR were normalized to total protein levels in each well, quantified using the BioRad DC Protein Assay (BioRad). Hundred microliter lysis buffer (50 mM NaCl, 50 mM EDTA and 1% Triton X-100) were added to each well of XF24 Cell Culture Microplate (Seahorse Bioscience), lysates were freeze–thawed three times and transferred to microcentrifuge tubes and centrifuged for 5 min at 14 000g. Protein content was determined in the supernatants according to the manufacturer’s instructions, with reference to a bovine serum albumin (BSA) standard curve, the absorbance being measured at 750 nm in a microplate spectrophotometer (SpectraMax Plus 384, Molecular Devices).
ATP quantification
ATP was quantified in a VICTOR Multilabel Plate Reader (Perkin Elmer), using the software Wallac 1420, for each condition of cellular energetics analysis. Supernatants from cell lysates were loaded onto a white-walled 96-well plate and 20 μl of 1:1000 luciferase solution (Sigma) and 20 μl of 400 μM luciferin (Sigma), both in lysis buffer, pH 8.0, were added to each well. Luminescence data were normalized using a standard curve for ATP obtained from nine ATP serially diluted solutions at concentrations ranging from 10−12 to 10−4 M, and a blank, without ATP.
Cell mobility and trajectory analysis
Cell mobility assays were performed using the μ-Slide Chemotaxis (Ibidi), as previously described (17). Briefly, 72 h after transfection, cells were detached from 12-well plates, resuspended and replated (10 μl of cell suspension containing 1.8 × 104 cells) in the central compartment of a μ-Slide Chemotaxis (Ibidi) plate. Cells were allowed to adhere to the slide for 2 h in an incubator at 37°C and 5% CO2, and, afterwards, complete cell medium was applied to each side compartment. Cell migration was recorded for 14 h, with photographs taken every 5 min, under a Carl Zeiss Axio Observer Z1 microscope with an incubator for temperature, humidity and CO2 control. The trajectory of each cell was determined and analyzed using the Image J software (v. 1.48, Wayne Rasband, National Institutes of Health, USA) with the Cell Tracking plugin and the Chemotaxis and Migration tool from Ibidi. For each experimental condition, images were taken in four different locations and a minimum of 20 cells were tracked. Experiments were independently repeated twice.
Cell invasion analysis
For cell invasion determination, 48-well plates were coated with 200 μl per well of a collagen matrix prepared by gently mixing PureCol bovine collagen type I solution (3 mg/ml; Advanced BioMatrix) with minimum essential medium (10×) and NaOH 0.1 M (5×) to achieve a final collagen concentration of 2.2 mg/ml. Plates were briefly incubated at 37°C to allow initiation of collagen polymerization before spheroid embedding. Four hours after transfection with DLS/miRNA complexes, each spheroid was collected with a p200 large-bore pipette tip and placed in the center of each well of a 48-well plate containing the collagen matrix. After complete collagen polymerization, 200 μl of complete DMEM (U87) or roswell park memorial institute (RPMI) (DBTRG) medium, without or with TMZ (at a final concentration of 400 μM per well) were added to each well on top of collagen. Cell invasion was monitored over the course of 4 days by digital photography at day 1 and day 4 using the 4×/0.10 objective of a Carl Zeiss Axio Observer Z1 microscope equipped with a CCD digital camera (AxiocamHRm). Invasion areas were quantified using Image J software (v. 1.48, Wayne Rasband).
Cell density evaluation
For the assessment of the effect of miR-200c overexpression on the density of cells grown in complete media or media lacking glucose or glutamine, cells plated in 96-well plates were transfected as described above, and 4 h after transfection, OPTIMEM medium was replaced with complete media or media lacking glucose or glutamine. Forty-eight hours after transfection, cells were fixed and the SRB assay was performed as described above.
Cell cycle analysis
The distribution of the cell population through the cell cycle phases was determined by quantifying cell DNA content using PI for DNA labeling, followed by flow cytometry analysis. Forty-four hours following drug treatment, cells plated in 12-well plates were detached and resuspended in ice-cold phosphate buffered saline (PBS). Cells were washed twice in ice-cold PBS by centrifugation at 600g at 4°C for 5 min, the supernatant was removed and the cells were fixed by slowly adding 70% ethanol under vortex followed by incubation at 4°C for 3 h. Cells were washed with 2% BSA in PBS and centrifuged at 600g at 4°C for 5 min. The resulting pellet was resuspended in PI (10 μg/ml)/RNase A (0.2 mg/ml) solution (Immunostep, Salamanca, Spain) and incubated for 15 min at room temperature in the dark. Cells were analyzed in a Becton Dickinson FACSCalibur Flow Cytometer (BD, Biosciences). Cells were appropriately gated by forward/side scattering and pulse width to discriminate viable and dead cells and to exclude doublets. Data were collected from at least 10 000 single cell events, stored in CellQuest software and analyzed using ModFit LT 3.0.
Cell death determination
Cell death was evaluated by flow cytometry with FITC-labeled annexin V (Immunostep) and PI (Sigma) double staining. Briefly, at 48 h after transfection and 44 h after TMZ incubation, U87 and DBTRG cells and supernatants were collected and washed by centrifugation in binding buffer (140 mM NaCl, 2.5 mM CaCl2, 10 mM HEPES/NaOH, pH 7.4) at 300g, 4°C, for 5 min. The cells were resuspended in 100 μl of binding buffer containing annexin V-FITC and PI at the final concentrations of 1 and 0.5 μg/ml, respectively. After 15 min incubation at room temperature in the dark, each sample was diluted 1:3 in binding buffer. The cells were immediately analyzed in a Becton Dickinson FACSCalibur flow cytometer, as previously described (42). A total of 10 000 events were collected for each sample. Analysis was performed using CellQuest software.
Statistical analysis
All data are presented as mean ± standard deviation of at least three independent experiments performed in duplicate, unless stated otherwise. One-way analysis of variance combined with the Tukey post hoc test was used for multiple comparisons (unless stated otherwise) and considered significant when P < 0.05. Statistical differences are presented at probability levels of P < 0.05, P < 0.01 and P < 0.001. Calculations were performed using standard statistical software (Prism 6, GraphPad, San Diego, CA, USA).
Conflict of Interest
None declared.
Funding
The European Regional Development Fund, through the Centro 2020 Regional Operational Programme (under project CENTRO-01-0145-FEDER-000008) (BrainHealth 2020) and through the COMPETE 2020—Operational Programme for Competitiveness and Internationalisation and Portuguese national funds via FCT—Fundação para a Ciência e a Tecnologia [under projects POCI-01-0145-FEDER-016390 (CANCELSTEM) and UIDB/04539/2020].
