Keywords

1 Introduction

Glioblastoma or glioblastoma multiforme (GBM), originating in the glial cells of the brain, is classified by WHO as a grade IV astrocytoma and has an extremely poor prognosis with a 5-year survival rate of 5% after diagnosis [1]. Conventional therapy using temozolomide [2,3,4] along with radiotherapy or surgery fails to prevent disease relapse, primarily due to the presence of residual cancer stem cells (CSCs) [5]. Constitutive activation of several signaling cascades in CSCs, such as the mitogen-activated protein kinase or MAPK pathway [6, 7], promotes chemoresistance [8], tumor heterogeneity and disease progression [9].

Upregulation of growth factor receptors that belongs to the receptor tyrosine kinase (RTK) family in GBM leads to the downstream activation of the MAPK/ERK pathway [10]. In addition, high phosphorylated ERK levels correlate with higher proliferation and invasion capacity of GBM cells. Inhibition of the MEK/ERK1/2 pathway increases ECM adhesion and reduces the migration ability of the cells [11]. The FDA-approved pan-RTK inhibitor sorafenib is a drug that is orally administered [12] to patients in the advanced stages of hepatocellular carcinoma (HCC) and renal cell cancer [13]. It inhibits proliferation and angiogenesis in HCC, often used in combination with other therapeutic agents [14,15,16] and also inhibits the migratory and invasive abilities of breast cancer cells in vitro [17]. Sorafenib targets the cell-surface RTKs (including VEGFR, PDGFR-β, c-KIT and Flt-3) as well as intracellular serine/threonine kinases (like Raf-1 and wild-type or mutated B-RAF) [18, 19]. Ulixertinib (BVD523), a reversible, small molecule inhibitor, has a high selectivity for ERK1/2. BVD523 inhibits the proliferation of cancer cells, upregulates apoptosis-related genes, and attenuates ABCB1 and ABCG2-mediated chemoresistance [20, 21], making it a suitable candidate for treating drug-resistant cancers. Preclinical trials with BVD523, in combination with or without other chemotherapeutic agents, have shown promising outcomes [22,23,24]. Another ERK inhibitor, BIX02188, targets ERK5 by inhibiting the kinase activity of MEK5 [25]. Although not very well characterized, BIX02188 induces apoptosis in FLT3-mutated cancers such as AML [26].

In the current study, the GBM cell line U87MG was treated with sorafenib, BVD523 and BIX02188 their effect on proliferation, survival, migration and self-renewal was determined. We found that pan-RTK inhibition significantly downregulated the proliferation and migration of glioblastoma cells compared to the selective inhibition of ERK1/2, whereas ERK5 inhibition alone did not produce any significant effect.

2 Materials and Methods

2.1 Differential Gene Expression and Pathway Analysis

The clinical data of glioblastoma patients and their respective mRNA expression profiles were obtained from The Cancer Genome Atlas (TCGA) database (https://cancergenome.nih.gov/). TCGA-Glioblastoma multiforme (GBM) dataset contains 169 GBM samples and five normal samples. R statistical software (version 4.1.3; https://www.r-project.org/) and Bioconductor packages (http://www.bioconductor.org/) were used to process and normalize the raw data. The Limma package was utilized to identify the differential gene expression between the normal and tumor tissues. Pathway enrichment analysis for differentially expressed genes was determined using gProfiler.

2.2 Reagents and Cell Lines

High glucose DMEM, propidium iodide and protease inhibitor cocktail were obtained from Sigma-Aldrich. Fetal bovine serum, trypsin, phosphatase inhibitor cocktail, primary antibodies (anti-human) against phosphoERK1/2, beta-catenin, BCL2, GAPDH and HRP conjugated secondary antibodies (anti-mouse and anti-rabbit) were procured from ThermoFisher Scientific. Fluorescent dye conjugated antibody against CD24 was purchased from ThermoFisher Scientific. BVD523 and BIX02188 were purchased from Selleck Chemicals LLC. Sorafenib was purchased from Natco Pharma Ltd. Glioblastoma cell line U87MG was purchased from NCCS, Pune, and plastic wares for cell culture were purchased from Eppendorf.

2.3 Colony Assay

Colony assay was carried out as previously described [27]. 100 cells/well were seeded in a 6-well plate and treated with the inhibitors. During pre-treatment, inhibitors were added for the initial 48 h, after which fresh media was added to induce the formation of the colonies. In some cases, treatment was extended until the completion of the experiment. The resulting colonies were counted microscopically after staining them with 0.1% crystal violet solution.

2.4 Protein Isolation and Immunoblotting

Cells were seeded, allowed to attach for 24 h, and treated with sorafenib or BVD523 for 48 h. As described previously, cells were lysed using RIPA buffer at the end of treatment period to collect the total protein [17, 27]. ~25 µg protein was loaded per well in a 10% polyacrylamide gel and transferred to a nitrocellulose membrane through semi-dry blotting. The membrane was incubated with the desired antibodies to detect the protein expression level.

2.5 Wound Healing Migration Assay

A previously published protocol was followed for the wound healing migration assay [17, 27]. 10,000 cells/cm2 were added to each well of a 12-well plate and incubated until they formed a monolayer, following which they were serum-starved for 12 h. A scratch was made using a micro tip, and media containing the respective inhibitors was added to each well. Microscopic images were taken at regular intervals and analyzed further to determine the migration speed.

2.6 Spheroid Assay

Glioblastoma cells were seeded at a density of 0.5–2 × 104 cells/mL in a low-attachment 96-well plate for the formation of spheroids. The spheroids were imaged periodically, and their area was determined using ImageJ software.

2.7 Proliferation Assay

5000 cells/cm2 were seeded in each well of a 24-well plate and allowed to attach. Cells were then treated with different concentrations of BVD523 for 48 h. The cells were trypsinized, counted, stained with propidium iodide (PI) and analyzed using a flow cytometer.

2.8 Surface Marker Analysis

Cell surface expression of CD24 was determined using a flow cytometer. Cells were trypsinized and stained with fluorescence conjugated antibody against CD24 for 30 min in the dark at 4˚C and analyzed by flow cytometry.

2.9 Cell Cycle Analysis

A previously published protocol was followed to perform cell cycle analysis [28]. Following treatment with the respective inhibitors, the cells were fixed and permeabilized with 70% ice-cold ethanol, followed by RNase A treatment. DNA was stained using PI and analyzed with a flow cytometer.

2.10 Data Analysis

Images from colony and spheroid formation assays were analyzed using the ImageJ software. The migration of cells in the wound healing assay was analyzed using the TScratch software, and flow cytometric data were analyzed with FCS Express5. Quantification of protein bands in the Western blots was done using the ImageLab software (Bio-Rad) and normalized to GAPDH expression levels.

3 Results

3.1 Identification of Differentially Expressed Genes and Pathways

The genes that are significantly differentially expressed between normal and glioblastoma tumor samples are shown in the heatmap (Fig. 1). The pathways linked to the differentially expressed genes are shown in the table (Table 1). MAPK signaling pathway was found to be dysregulated in the TCGA-GBM cohorts, along with the Wnt/β-catenin pathway. So, we studied the role of the MAPK pathway in vitro in glioblastoma cells using MAPK pathway inhibitors.

Fig. 1
A heatmap of traces of the genetic compositions in different genotypes. A color key is provided on the left side of the heatmap, which measures the row z score from negative 6 to positive 6.

Heatmap representing differentially expressed genes between normal and glioblastoma patients

Full size image
Table 1 Pathways associated with differentially expressed genes in glioblastoma indicate dysregulation of the MAPK pathway
Full size table

3.2 RTK Inhibition Decreases the Self-renewal, Proliferation and Migration of Glioblastoma Cells

Sorafenib was reported to inhibit the growth of glioblastoma cells and selectively inhibit the tumor-initiating population in primary cells [29, 30]. We previously reported that sorafenib (10 μM) significantly inhibits the proliferative, migratory and invasive abilities of breast cancer cells and modifies their intracellular signaling pathways [17]. Sorafenib treatment inhibited the tumor-initiating ability in GBM, where the sorafenib-treated U87MG cells failed to form colonies (Fig. 2a). Furthermore, upon treatment with sorafenib, the GBM cells formed smaller spheroids, indicating that sorafenib inhibits cell proliferation and self-renewal ability (Figs. 2b, 4b). CD24 is a marker of stem cells in several cancers, including glioma [31,32,33,34] and we found a significant decrease in the CD24 + population after sorafenib treatment (Fig. 2c–d), suggesting that sorafenib treatment reduces the CSC population in GBM.

Fig. 2
An assay output of CON and SORA for 48 hours and 7 days. The assay outputs of SORA and CON were measured in 4, 7, 10, and 14 days. A graph of cell count versus C D 24 expression. Both plots have 2 peaks. A bar graph of C D 24 expression versus CON and SORA. The plot of CON is greater than SORA.

a Colony formation assay was performed with control untreated U87MG cells (CON) or cells treated with sorafenib (SORA, 10 μM) for 48 h (48H) or 7 days (7D) b Spheroid formation assay was done without (CON) or with sorafenib treatment, and the spheroids were imaged at the intervals of 4, 7, 10 and 14 days (4, 7, 10, 14D). The black line indicates the scale bar (200 μm) in all the microscopic images c Flow cytometry analysis was done to determine the expression levels of CD24 on the surface of U87MG cells in untreated (CON), and sorafenib (SORA) treated cells d Represents the normalized Mean Fluorescence Intensity (MFI) of CD24 in U87MG cells. Values are mean ± SE, n = 3, *p < 0.05 e Western blotting was performed to determine the expression of phosphorylated ERK1/2 (pERK1/2), BCL2 and β-catenin after 48 h of SORA treatment, CON represents the untreated control cells

Full size image

Protein expression analysis showed that sorafenib induces the accumulation of phosphoERK1/2 and downregulates the expression of β-CATENIN, implying that cell growth and proliferation were hindered. Further, a reduction in BCL2 expression was observed, which suggests that sorafenib redirected the cells towards apoptosis (Fig. 2e). Thus, we conclude that sorafenib decreases the self-renewal of glioblastoma cells and alters the MAPK/ERK signaling cascade, which is essential for proper cellular functioning.

3.3 ERK1/2 Inhibition Decreased the Viability of Glioblastoma Cells in a Concentration-Dependent Manner

To further understand the effect of MAPK inhibition in GBM, we treated U87MG with the ERK1/2 selective inhibitor BVD523 (BVD) and determined its effect on proliferation and migration. The viability of U87MG glioblastoma cells decreased with increasing BVD concentration, with an IC50 of 10 μM (Fig. 3a, b). This reduction in cell survival was accompanied by a dramatic increase in the accumulation of phosphoERK1/2 and downregulation of β-CATENIN protein levels (Fig. 3c). Based on these results, 10 μM of BVD was used for further experiments.

Fig. 3
2 line graphs of live cells versus B V D and cell count multiplied by 10 to the power 4 versus B V D. Both the plots follow a negative trend. Western blot analysis of the electrophoresis plate with pERK 1 by 2, B CATENIN, and GAP D H. A line graph of pERK expression versus B V D follows a positive trend.

a Live cell percentage post-treatment with the indicated concentrations of BVD523 (BVD) for 48 h was determined by staining with propidium iodide (PI) and analyzed using flow cytometer. b The cell count was determined after treatment with BVD for 48 h c Immunoblotting analysis was performed to determine the phosphoERK1/2 (pERK1/2) and β-CATENIN expression level after treatment with the indicated concentrations of BVD for 48 h. d The pERK1/2 levels shown in c were quantified and normalized against their respective GAPDH levels

Full size image
Fig. 4
A microscope image of the output of CON and B V D measured in 4, 7, 10, and 14 days. A line graph of spheroid area versus days. The plots follow a positive trend. A graph of cell count with sharp peaks. A graph of cell count versus C D 24 expression with 2 peaks. A bar graph of C D 24 expression.

a U87MG cells were allowed to form 3D spheroids in the absence (CON) or presence of BVD523 (10 μM, BVD). The black line shows the scale bar of 200 μm b The graph represents the spheroid size of U87MG cells upon treatment with sorafenib (SORA) or BVD523 (BVD) for the indicated time. c The cell cycle profile of untreated (CON) and BVD-treated U87MG cells was determined by PI staining through flow cytometric analysis. The histogram shows the cell cycle profile. d The cell surface expression of CD24 in control (CON) or BVD (10 μM, 48 h) treated U87MG cells was determined by flow cytometry analysis. The black and red lines indicate the isotype control and CD24 stained sample respectively. e Graph represents the CD24 expression levels in control (CON) and BVD-treated U87MG cells. Values are mean ± SE, n = 2–3, *p < 0.05, **p < 0.005

Full size image

3.4 ERK1/2 Inhibition Affects the Proliferative, Migratory and Self-renewal Ability of Glioblastoma Cells

Spheroid formation assay was performed to find the effect of BVD on the survival and proliferation of U87MG cells. BVD treatment significantly inhibited spheroid growth, and the inhibitory effect was sustained until 14 days (Fig. 4a, b). Cell cycle analysis showed that BVD treatment induces G0/G1 arrest in U87MG cells (Fig. 4c) in agreement with the reduced proliferation. However, we noticed a significant increase in the expression levels of CD24 after BVD treatment (Fig. 4d, e). BVD treatment, however, lowered the migration ability (Fig. 5a, b) and significantly decreased the colony formation ability (Fig. 5c, d). The colony formation ability was completely abrogated when treatment with BVD was extended beyond 72 h, suggesting that prolonged BVD treatment inhibits the self-renewability of U87MG cells. The effect of BVD on self-renewal was found to be irreversible, where the cells did not regain their colony formation ability when BVD was removed after 72 h of treatment.

Fig. 5
A bar graph of migration versus sample. The maximum bar height is present for BIX. The analysis outputs of the samples in 0 and 28 hours. The analysis outputs of BIX and B V D in untreated, 48 hours, and contd. A bar graph of the colony area with bar heights of BIX greater than B V D.

a, b Migration of U87MG cells was checked by wound healing assay after treatment with sorafenib (SORA, 10 μM), BVD523 (BVD, 10 μM) and BIX02188 (BIX, 2 μM). c The ability of U87MG cells to self-renew was determined by colony formation assay after treatment with BVD or BIX for 48 h (48H) or continuously (CONT) throughout the duration of the assay. UNTR represents the untreated control d The graph represents the colony area represented in c. Values are mean ± SE, n = 2–3, *p < 0.05. The scale bar (200 µm) was indicated by black lines in all the microscopic images

Full size image

3.5 Effect of ERK5 Inhibition on Glioblastoma Migration and Self-renewal Ability

Next, we studied the effect of MEK5 inhibition on the self-renewal and migration of U87MG cells. MEK5, an upstream activator of ERK5, was inhibited by treatment with BIX02188. MEK5 inhibition did not affect the migration ability of U87MG cells, where the migration speed after treatment with BIX02188 (BIX) was similar to that observed with the control cells (Fig. 5a, b). The migration ability was significantly inhibited, however, after sorafenib treatment. Nevertheless, we observed a moderate reduction in the colony formation ability with BIX treatment, and the prolonged treatment with BIX did not further inhibit the ability of U87MG cells to form colonies (Fig. 5c, d).

4 Discussion

Survival of a stem cell-like population in glioblastoma tumors after conventional treatment with temozolomide poses a significant challenge in preventing disease relapse [5]. This highlights the need for alternative therapeutic strategies to target self-renewing and chemoresistant cells in cancer. In this study, we found that MAPK pathway genes were differentially expressed in TCGA-GBM cohorts. So, we initially examined the effects of the broad-spectrum RTK inhibitor sorafenib on the proliferation, migration and self-renewal of glioblastoma cells. As reported by previous studies in glioblastoma [29], breast cancer [17] and hepatocellular carcinoma [35], sorafenib treatment inhibited colony formation in vitro, indicating that sorafenib abrogates the self-renewal ability. This was accompanied by a reduction in the cell surface expression of CD24, a CSC marker [31,32,33,34]. Decreased migration in 2D culture and smaller spheroid size in 3D culture indicate that sorafenib treatment inhibits the metastatic and proliferation ability of glioblastoma cells.

Furthermore, the decreased survival and proliferation ability of GBM cells after sorafenib treatment was indicated by reduced BCL2 and beta-catenin levels, respectively. Interestingly, pERK1/2 levels were upregulated, in contrast to reduced pERK1/2 levels reported in hepatocellular and breast carcinomas [17, 36]. The mechanism of this upregulation in glioblastoma is not well understood but may be caused by the multikinase inhibitory activity of sorafenib on ERK1/2 rather than on its upstream activators, leading to altered feedback loops within the RAS/RAF/ERK pathway.

It is reported that elevated expression of pERK1/2 is associated with glioblastoma progression. In order to understand the mechanism, ERK1/2 inhibitor BVD523 was utilized in this study. BVD effectively reduced cell viability and proliferation in 2D and 3D cultures [37,38,39] and induced cell cycle arrest. BVD also inhibited the colony formation and migration ability of U87MG cells. BVD treatment altered two intrinsic signaling pathways, the RAS/RAF/ERK pathway and the Wnt/β-catenin pathway. β-catenin downregulation leads to reduced proliferation [40, 41]. However, a concentration-dependent increase of pERK1/2 levels after BVD treatment indicates its aberrant kinase activity rather than its phosphorylation status, thereby preventing activation of downstream targets [20]. Although sorafenib and BVD523 showed a similar effect in terms of abrogating tumor initiation, proliferation and migration, the effects observed were more pronounced with the former, implying that the activity of sorafenib is mediated by ERK1/2 inhibition in conjunction with inhibition of other tyrosine kinase receptors. Selective inhibition of MEK5 kinase activity, thereby ERK5 inhibition with BIX02188, moderately decreased the self-renewal ability of U87MG but did not alter the cell migration. Nevertheless, it cannot be concluded that ERK5 inhibition does not affect tumor proliferation or metastasis since the effect might become pronounced when used in combination with one or more inhibitors.

Thus, sorafenib and BVD523 effectively reduced tumor initiation, proliferation, survival and metastasis by altering the ERK signaling cascade, whereas BIX02188, when used alone, was not effective against the glioblastoma cells. However, combinatorial therapy with drugs and inhibitors that target alternate cellular mechanisms is therapeutically more beneficial than single pathway inhibition.