Temozolomide increases COX-2 levels in human brain cancer cells

First, we wanted to verify the expression status of COX-2 in two GBM cell lines: T98G and U251MG. Either reverse transcriptase PCR (Figure 1A) or Western blot (WB) assay (Figure 1B) showed the COX-2 constitutive expression in T98G and the lack of COX-2 in U251MG, thus confirming previous findings [11,12,41]. Afterward, the effect of TMZ on cell proliferation was evaluated on the two GBM cell lines, exposed to increasing concentrations of drug (10–400 µM) or drug vehicle DMSO (CNTR) by IncuCyte^® Live Cell Imager instrument every 4 h for a total timeline of 3 days (72 h). The T98G cell proliferation, as expected, was significantly affected only at the highest concentration of TMZ (400 µM), thus confirming its elevated chemoresistance (Figure 1C). The U251MG cells showed a significant reduction in proliferation compared with CNTR cells at all concentrations of TMZ (Figure 1C).

With the aim of analyzing whether TMZ was able to influence the COX-2 expression in GBM cells, the T98G and U251MG cells were treated with the different concentrations of TMZ for 3 days, and the COX-2 levels were analyzed.

In T98G cells, the COX-2 mRNA levels were significantly increased after exposure with TMZ at 200 µM and enhanced dramatically at 400 µM when compared with CNTR cells (Figure 2A). In addition, WB assay revealed that the COX-2 protein levels were significantly enhanced at increasing TMZ concentrations, resulting in being significant at concentrations of TMZ ≥ 200 µM with respect to CNTR (Figure 2B). In the COX-2 negative (COX-2^–) GBM cell line, U251MG, TMZ failed to induce the COX-2 expression at all tested concentrations (Figure 2C).

Based on the obtained results on TMZ-resistant T98G regarding the up-regulation of COX-2 by TMZ, we wanted to verify whether TMZ was able to modulate the key proteins closely associated with the activity of COX-2 and strongly implicated in the progression and resistance of GBM. The effects of TMZ on the β-catenin, MGMT, and SOX-2 expression were evaluated after exposure to increasing doses of the alkylating agent (10–400 µM) for 3 days. The results of these experiments show an up-modulation of β-catenin proportional to the increasing concentrations of TMZ in T98G, resulting in being significant at TMZ ≥ 200 µM (Figure 3A). Similarly, the MGMT levels were enhanced accordingly to increasing TMZ concentrations in T98G cells, being significant at 200 and 400 µM (Figure 3B). To assess the effects of TMZ on the stemness potential of T98G cells, we also analyzed SOX-2 levels. A significant and dose-dependent increase in the SOX-2 protein expression was detected at 100 µM, 200 µM, and 400 µM with respect to CNTR (Figure 3C). On the other hand, TMZ did not affect the β-catenin expression in the U251MG cells (Figure 3D). In the MGMT negative (MGMT^–) cell line U251MG, the MGMT, as well as SOX-2 expression, was not influenced by TMZ treatment for 3 days (data not shown).

The effect of NS398 on the PGE2 levels, related to COX-2 activity, was evaluated in the supernatant of cell cultures using different concentrations (50–400 µM) for 3 days. In T98G cells, the NS398 strongly inhibited the COX-2 activity at higher concentrations (200 and 400 µM) (Figure 4A). As expected, PGE2 levels were undetectable in the culture medium of U251MG cells (data not shown). To identify the most appropriate concentration of COX-2 inhibitor to use in the next experiments, GBM cells were treated with increasing concentrations of NS398 (50–400 µM) to continuously evaluate the cell growth by IncuCyte^® Live Cell Imager instrument every 4 h for a total timeline of 3 days (72 h). The NS398 exposure influenced the proliferation rate of T98G COX-2-positive (COX-2⁺) cell line showed significant differences compared with CNTR cells at maximal concentrations, 200 and 400 µM (Figure 4B). Of note, treatment with NS398, even at the highest concentrations, had no effect on cell proliferation of U251MG (Figure 4B). Based on the obtained data, the NS398 concentration of 200 µM, being able to effectively inhibit COX-2 activity as well as proliferation rate of T98G cells, was chosen for the subsequent experiments designed to analyze the effects of combined treatment NS398+TMZ. In addition, the results obtained with U251MG, being COX-2^–cells, confirmed the specificity of the NS398 inhibitor.

To verify the effect of the COX-2 inhibitor, NS398, on TMZ-resistance, the GBM cell lines were exposed to single agents and their mixture for 3 days. Based on the above results, the chosen TMZ concentrations to keep the proliferation rate above 50% were 200 µM for T98G cells and 10 µM for U251MG cells. The biological parameters such as viability, cell cycle, and apoptosis levels were evaluated. In the T98G cell line, the single agents, NS398 and TMZ, did not induce a significant increase in dead cell percentage when compared with CNTR (Figure 5A). The TMZ exposure significantly enhanced the dead cell percentage only in U251MG, thus confirming the high chemoresistance of T98G. The co-treatment (NS398+TMZ) induced a significant increase in cell mortality in T98G, showing a synergistic effect (Figure 5A); on the contrary, the combination induced an effect similar to TMZ alone in COX-2^– U251MG, as reported in Figure 5B. Representative images from microscopic observations of the cell cultures under the different treatment conditions, displayed in the below panels of Figure 5, confirmed the reduction in cell proliferation rate.

The effect of TMZ, NS398, and their combination on cell cycle distribution by flow cytometry was analyzed on T98G and U251MG. The cell cycle analysis of T98G cells showed that NS398 treatment significantly reduced the percentage of T98G cells in G1. Meanwhile, the presence of TMZ determined not only an evident decrease in cells in G1 but also a clear increase in the S-phase. Of interest, the NS398+TMZ combination altered the cell cycle phase distribution, inducing a significant reduction in G1-phase with respect to CNTR, NS398, and TMZ, with a significant cell accumulation in G2/M-phase with respect to CNTR and TMZ (Figure 6A,C). In the U251MG cells, the treatment with TMZ caused similar effects as the combination, resulting in a reduction in the cell population in the G1 phase, with a clear increase in the G2/M phase fraction when compared with CNTR and NS398-treated cells at 3 days (Figure 6B,C).

A moderate increase in apoptotic cell percentage, detected by the sub-G1 peak, in T98G treated for 3 days with the inhibitor NS398 alone was registered. On the other hand, the exposure for 3 days to TMZ alone failed to induce significant levels of apoptosis. Of note, the addition of NS398+TMZ led to a much higher and statistically significant level of apoptosis than all other culture conditions (Figure 7A). In U251MG, NS398, which alone did not increase the sub-G1 population, when combined with TMZ, did not significantly influence the apoptotic cell percentage observed after the treatment with TMZ alone (Figure 7B). Overall, in the U251MG cell line, COX-2^–, the COX-2 inhibition by NS398 did not strengthen the antiproliferative, apoptotic, and cytostatic effect of TMZ. Otherwise, in COX-2⁺, TMZ-resistant T98G cells, the reduction in COX-2 expression and activity led to the appearance of susceptibility to TMZ, resulting in a significant cell proliferation reduction, cell cycle arrest, and increased apoptosis level, suggesting the involvement of COX-2 activity underlying the chemoresistance mechanisms.

NS398 alone significantly affected the colony generation in the COX-2⁺ cell line, T98G (Figure 8A). As expected, the number of colonies generated by T98G cells was not significantly affected by TMZ standalone (Figure 8A). Notably, the combination NS398+TMZ dramatically inhibited the clonogenic potential in T98G cells. No effect was observed in U251MG cells when exposed to NS398 compared with CNTR, while very few colonies were developed after treatment with TMZ (Figure 8B). Additionally, in the clonogenic ability, no significant difference was detected between NS398+TMZ combination and TMZ alone treatment in U251MG (Figure 8B).

The COX-2 gene expression was assessed through real-time PCR in GBM cells after a 3-day treatment with single agents or their combination, and results are presented in Figure 9A. As reported above, the TMZ exposure induced a relevant and significant COX-2 gene expression in the T98G cell line. Of note, NS398, which alone led to a downmodulation of the basal COX-2 levels, when added together with TMZ, was able to counteract the TMZ-induced COX-2 overexpression in this cell line, revealed by COX-2 mRNA and protein level assay (Figure 9A,B). PGE2 levels’ assay in T98G cell supernatants also confirmed these results, being the PGE2 production affected by treatment in an overlapping way (Figure 9C). On the other hand, COX-2, which was not expressed in U251MG cells, was not affected by any treatments (Figure 9D).

To still investigate the ability of NS398 to counteract the TMZ resistance, the MGMT mRNA and protein levels were evaluated in the T98G cell line treated for 3 days with TMZ and NS398, alone or in combination. The COX-2 inhibitor significantly down-modulated, either at the transcriptional and post-transcriptional level, the basal and TMZ-induced MGMT expression (Figure 10A,B). These results were also confirmed by immunofluorescence analysis (Figure 10C). It is known that the U251MG are MGMT^– cells [42]; however, the effect of the different treatments on MGMT expression levels was evaluated. Both the single agents and the NS398+TMZ combination did not induce MGMT protein expression in these cells (Figure 10D).

To further elucidate the mechanism underlying the NS398 ability in regulating the TMZ-resistance, we investigated the β-catenin expression involved in maintaining the stem-like state in GBM as well as in the progression and recurrence of the tumor [43]. The β-catenin protein expression in the T98G cell line exposed for 3 days to NS398 and TMZ, alone or in combination, was analyzed.

Our results show that COX-2 inhibitor, NS398, down-regulated, even though not significantly, the β-catenin expression compared with CNTR, while TMZ alone significantly up-regulated its expression (Figure 11). Of interest, the treatment with NS398+TMZ counteracted the β-catenin expression induced by TMZ, considerably reducing it even below the baseline level.

Finally, with the Wnt/β-catenin pathway being involved in GBM stem-like maintenance and considering the close link between GBM resistance and cancer stem cell presence, we wanted to analyze the effect of the COX-2 inhibitor alone or in combination with TMZ on the SOX-2 expression in T98G cells. As our group previously reported [44], we confirmed that T98G cells expressed high basal levels of SOX-2 stemness marker. The exposure to COX-2 inhibitor for 3 days slightly influenced the SOX-2 basal expression, although not significantly (Figure 12A). A huge increase in SOX-2 expression of T98G after TMZ treatment was detected, thus confirming previous data [45]. Notably, NS398 significantly counteracted the TMZ-induced SOX-2 expression (~3-fold decrease vs. TMZ alone) (Figure 12A). Representative images of immunofluorescence staining of T98G cells, showing the SOX-2 protein, localized mainly in nuclei, supported the above results (Figure 12B).

Human GBM grade IV cell lines, T98G and U251MG, were purchased from the European Collection of Authenticated Cell Cultures (ECACC) and Cell Lines Service (CLS), respectively, and manipulated according to the instructions given in the product sheet to maintain a stable genetic profile. The main characteristics of GBM cell lines are as follows: T98G cells express high levels of MGMT, p53 and PTEN are mutated, and p14ARF/p16 deleted; U251MG cells are p53 wild type, PTEN null, p14ARF/p16 deleted and MGMT null [38,61]. The basal expression of both MGMT and SOX-2 was also evaluated by Western blotting in T98G and U251MG (Supplementary Figure S1). The cell lines were grown as adherent cells in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% (v/v) of fetal calf serum (FCS), 2 mM L-glutamine, 100 U/mL penicillin, and 100 mg/mL streptomycin (complete medium) (EuroClone, West York, UK). As previously reported in Lee SY, the T98G and U251MG show a different sensitivity to TMZ: T98G cells are defined “TMZ-resistant”, showing a LC50 ranged from >250 mM to 1585 mM, and U251MG cells are considered “TMZ-sensitive”, showing a LC50 around 50 μM [38]. Cells were maintained at 37 °C in a humidified atmosphere of 95% air and 5% CO₂, and medium was totally replaced every three days.

The clinically used TMZ chemotherapy was obtained by Sigma-Aldrich (Saint Louis, MO, USA) prepared as stock solutions in sterile 100% dimethyl sulfoxide (DMSO), whose final concentration in the culture medium was 51.5 mM. This concentration did not influence cell viability and the expression of the studied proteins. Based on previous reports [62,63,64], TMZ at different final concentrations (10, 100, 200, and 400 µM) was tested by dose–response curve for 24, 48, and 72 h to select the drug concentration for the subsequent experiments. More precisely, the specific concentration of chemotherapeutic agent able to maintain the proliferation at 72 h of each cell line above 50% was chosen for the experiments designed to analyze the effects of combined treatment NS398+TMZ. The selective COX-2 inhibitor, NS398 (N-[2-(Cyclohexyloxy)-4-nitrophenyl]methanesulfonamide), was acquired from Sigma-Aldrich, and according to manufacturer instruction, the COX-2 inhibitor was stored as stock solutions in DMSO at −20 °C and diluted in cell culture medium just before use. The choice of NS398 concentration range (50–400 µM) and treatment times (up to 72 h) was first based on previous reports and in vitro studies on GBM cell lines [12,20]. Moreover, the NS398 concentration able to effectively inhibit COX-2 activity as well as proliferation rate of T98G cells was chosen for the subsequent experiments designed to analyze the effects of combined treatment NS398+TMZ. Analysis of viable and healthy cells treated with TMZ or NS398 was performed utilizing the IncuCyte^® Live Cell Imager system (Essen BioSciences, Inc., Ann Arbor, MI, USA). Briefly, cells were plated in a 96-multiwell culture plate at 1000 cells/cm² and allowed to attach overnight. Once attached, media were replaced with media containing TMZ (10, 100, 200, and 400 µM) or NS398 (50, 100, 200, and 400 µM), and NucLight Rapid Red live-cell nuclear dye (1:4000 dilution), a lentivirus-based reagent that stably integrates into the host cell line without affecting cell growth or morphology, was added. Culture plates were sited into the IncuCyte^® Live Cell imager, and images were captured using phase contrast and red (400 msec exposure) channels and were taken every 4 h in the IncuCyte ZOOM™ platform (Essen BioSciences, Inc.). Four image sets were acquired from several points of the well, using a 10× objective lens, and all the treatment conditions were run in triplicates. At the end of the experiment, the proliferation rate, based on NucLight dye internalization, was analyzed by IncuCyte^® S3 2018C software (Essen BioSciences, Inc.). The treatment with the compound vehicle alone, DMSO, was referred to as “control” (CNTR).

To evaluate the effect of NS398, TMZ, and their combination on cell viability, GBM cell lines were plated for all experiments in 25 cm² plates at 5000 cells/cm², left to adhere overnight, and then treated with the COX-2 inhibitor, NS398, TMZ, and compounds combined at selected concentrations. The time of exposure to treatments was set for 3 days (72 h). Following treatment, the cells were harvested and centrifuged for 10 min at 400× g. Cell count was assessed in a Bürker chamber by optical microscopy (Eclipse 50i, Nikon, Tokyo, Japan) using Trypan blue solution (0.04%, final concentration, EuroClone, West York, UK). Not-treated cells, referred to as CNTR, were also analyzed. Cell morphology was evaluated by an optical microscope (Eclipse 50i, Nikon), and images were acquired at 10× and 20× magnification. The cells treated with different concentrations of TMZ were also used for COX-2 mRNA and protein quantification. All reagents for cell biology and consumables were purchased from EuroClone where not otherwise specified.

To examine the basal expression of COX-2 in GBM cell lines, total RNA was isolated according to the RNeasy Mini Kit (QIAGEN, Hilden, Germany). RNA was spectrophotometrically quantified, and its quality was assessed by 1% agarose/Tris–Acetate–EDTA (TAE) gel electrophoresis. First-strand cDNA was synthesized using total RNA and the M-MLV cDNA Synthesis kit (Invitrogen, Waltham, MA, USA). Specific primers for COX-2 (forward 5′-ATTGTACCCGGACAGGATTCTATG-3′ and reverse 5′-TTTGGAGTGGGTTTCAGAAATAATT-3′) and β-actin (forward 5′-CACCATTGGCAATGAGCGGTTC-3′ and reverse 5′-AGGTCTTTGCGGATGTCCACGT-3′) were used to amplify gene fragments with the annealing temperature set at 60 °C. The PCR products were analyzed on 1.2% agarose gel and visualized by EuroSafe Nucleic Acid Staining (EuroClone, West York, UK). Densitometric analysis was performed using ImageJ software to quantify the band intensities. Data represent average values from three independent experiments.

Gene expression analysis of COX-2 and MGMT was performed in untreated (CNTR), treated NS398-, TMZ- and NS398+TMZ-T98G cells via real-time RT-PCR using a ViiA7 sequence detection system (Applied Biosystems, Foster City, CA, USA). Total RNA was extracted from cells using the RNeasy Mini Kit and quantified by spectrophotometry. An amount of 1 µg of total RNA was reverse transcribed in a final volume of 20 μL using a mixture of random primers below reported. An amount of 1 µg of each cDNA was used to perform the real-time PCR. Real-time quantitative RT-PCR analysis was carried out by SYBR Green dye detection (Thermo Fisher Scientific Waltham, MA, USA) according to the manufacturer’s instructions. Reverse and forward primers, purchased from IDT (acquired from Integrated DNA Technologies, IDT, Coralville, IA, USA), were used at a concentration of 1 µM (COX-2, MGMT, GAPDH), and their sequences were as follows: MGMT forward: 5′-GTCGTTCACCAGACAGGTGTTA-3′ and reverse 5′-ACAGGATTGCCTCTCATTGCTC-3′; COX-2 forward 5′-ATTGTACCCGGACAGGATTCTATG-3′ and reverse 5′-TTTGGAGTGGGTTTCAGAAATAATT-3′. In the T98G cells, as internal control for gene expression normalization, the mRNA level of GAPDH was measured by following primers: GAPDH forward 5′-TTGCCCTCAACGACCACTTT-3′ and reverse 5′-TGGTCCAGGGGTCTTACTCC-3′. The fold-change quantification of target genes was calculated with the 2^-∆∆Ct method [65]. The samples were run in triplicate, and the experiment was repeated twice.

Control and treated cells were harvested, washed in phosphate-buffered saline (PBS), and homogenized and lysed in ice-cold RIPA buffer (Merck KGaA, Darmstadt, Germany) containing 100 mM protease inhibitor cocktail (Sigma-Aldrich, Saint Louis, MO, USA). The protein amount was evaluated with DC Protein Assay (Bio-Rad, Hercules, CA, USA) using BSA as standard. An amount of 25 μg of protein lysates was mixed with sample buffer; samples were denatured (at 100 °C for 5 min) and run on 10% SDS polyacrylamide gel. Proteins were transferred onto 0.45 μm nitrocellulose membrane sheets (Bio-Rad) for 1 h at 4 °C at 70 V using a Mini Trans-Blot Cell apparatus (Bio-Rad). Non-specific binding sites were blocked with 5% non-fat dry milk for 1 h at room temperature and then incubated overnight at 4 °C with primary antibodies: rabbit monoclonal anti-COX-2 (Cell Signaling Technology, Danvers, MA, USA; dilution 1:1000), mouse monoclonal antibody anti-MGMT (BD Biosciences, San José, CA, USA; dilution 1:500), mouse monoclonal antibody anti-SOX2 (Origene, 9620 Medical Center Drive Suite 200 Rockville, MD, USA; dilution 1:1000), rabbit polyclonal antibody anti-β-catenin (Cell Signaling Technology, Danvers, MA, USA; dilution 1:1000), and mouse monoclonal antibody for anti-β-actin antibody (Bio-Rad, Hercules, CA, USA; dilution 1:1000). The horseradish peroxidase (HRP)-conjugated anti-mouse and anti-rabbit secondary antibodies were purchased from Sigma-Aldrich (Saint Louis, MO, USA). Immunoreactive bands were visualized by enhanced chemiluminescence (ECL, Amersham Pharmacia Biotech), according to the manufacturer’s instructions. Band relative densities were determined using a chemiluminescence documentation system ALLIANCE (UVITEC, Cambridge, UK), and values were given as relative units. All data were normalized to β-actin protein levels, used as internal control, and results are expressed as arbitrary units.

The supernatants of T98G cells were collected after treatment for 72 h with NS398, TMZ, and combination at the concentrations above indicated and then assayed for prostaglandin E2 (PGE2) levels by an enzyme-linked immunosorbent assay (ELISA) kit (Cayman Chemical Company, Ann Arbor, MI, USA) as described in the manufacturer’s instructions. The least detectable PGE2 concentration was 7.8 pg/mL. Results are presented as pg/mL of PGE2.

The effects of COX-2 inhibitor and of TMZ on the cell cycle and apoptosis were examined in T98G and U251MG cell lines exposed to agents alone and their mixture for 3 days. At the end of incubation times, adherent cells were detached using trypsin–EDTA solution (EuroClone, West York, UK), counted, and fixed by using a cooled 70% ethanol solution (Sigma-Aldrich, Saint Louis, MO, USA), in PBS, with gentle mixing at 4 °C for 30 min. Fixed cells (10⁵ cells/mL) were transferred to plastic BD tubes (BD Biosciences, San José, CA, USA), washed twice with ice-cold PBS, and resuspended with a solution containing 50 μg/mL PI, 0.1% Nonidet-P40, and RNase A (6 μg/10⁶ cell) (all reagents acquired from Sigma-Aldrich, Saint Louis, MO, USA) for 30 min in the dark at 4 °C. The percentages of cells in the G1, G2/M, and S phases (data from 10,000 events per sample) were calculated by software Modfit LT for Mac V3.0 using a FACS Calibur instrument (BD Biosciences). The assay was carried out in duplicate. Apoptotic cells were determined by their hypochromic subdiploid nuclei staining. Data from 10,000 events per sample were collected and analyzed using a FACS Calibur instrument (BD Biosciences) equipped with cell cycle analysis software (Modfit LT for Mac V3.0).

The colony formation assay, or clonogenic assay, is an in vitro cell survival assay, a gold standard technique used for the study of chemical or anti-cancer agents, evaluating the cells’ ability to generate clones following treatment and allowing the evaluation of cytotoxic and/or antiproliferative effects [66]. Briefly, T98G and U251MG were seeded in 6-well plates at a concentration of 1000 cells/well and were incubated at 37 °C overnight for attachment. The next day, cell medium was changed, cells were treated with a single dose of NS398 (200 µM), TMZ (200 µM for T98G and 10 µM for U251MG), and their combination and placed in the CO₂ incubator for 3 days. After this treatment, the cells were maintained in a fresh medium until colony formation attested on 20 and 18 days for T98G and U251MG cultures, respectively. Colonies were gently washed, fixed with cold methanol for 20 min, and stained with crystal violet 0.1% in PBS at room temperature for 10 min and air-dried. Colonies were visualized, and images were captured by optical microscopy (Eclipse 50i, Nikon, Tokyo, Japan). Colonies formed were counted using the ColonyCountJ, an add-on program, a semi-automated colony counting program to ImageJ, which provides the number of colonies [67]. The experiment set was carried out twice in duplicate, and histogram plots are the average number from three analyses.

The T98G cells were grown on coverslips in a 12-well plate (seeded at 3000 cells/cm²) and let to adhere overnight, then were treated with NS398, TMZ, and the combination for 3 days as previously reported. At the end of treatment, the coverslips were washed with PBS, fixed with 4% formaldehyde for 20 min, permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, Saint Louis, MO, USA) for 5 min, and blocked with 3% BSA (Sigma-Aldrich) for 20 min at room temperature. Cells were incubated overnight at 4 °C with mouse monoclonal antibody anti-human MGMT (BD Biosciences, San José, CA, USA; dilution 1:500). Next, the coverslips were stained using a FITC conjugated goat anti-mouse polyclonal IgG secondary antibody (Bethyl Laboratories, Inc, Montgomery, TX, USA) for 1 h at room temperature and washed. For SOX-2 immunostaining, cells were incubated overnight at 4 °C with FITC-conjugated mouse monoclonal antibody anti-SOX2 (Origene, 9620 Medical Center Drive Suite 200 Rockville, MD, USA; dilution 1:100). All samples were then incubated with TRITC labeled phalloidin (Sigma-Aldrich) for 45 min at room temperature. The coverslips were mounted with VECTASHIELD^® Antifade Mounting Medium with DAPI (Vector Laboratories, Inc., Burlingame, CA, USA) and then examined at 100× magnifications with the fluorescent microscope (Eclipse 50i, Nikon, Tokyo, Japan).

The GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA) software was used for statistical analysis. Data were assessed for normality using the Shapiro–Wilk test and found to be normally distributed (p > 0.05). Comparisons between groups were performed using one-way or two-way ANOVA followed by Tukey or Dunnett’s post hoc test, where specified. In particular, the two-way for repeated measures ANOVA was used to evaluate the effect of two within-subject factors on a continuous outcome variable simultaneously. Data are expressed as the mean ± SEM (standard error mean) of independent experiments performed in duplicate or triplicate. Results were considered significant if p < 0.05.

The datasets generated and analyzed during the current study are available from the corresponding authors upon reasonable request.