Cell lines BT-549 (breast ductal carcinoma; HTB-122™), TOV-21G (ovarian adenocarcinoma; CRL-11730™), RKO-AS45-1 (colorectal carcinoma; CRL-2579™) and WI-26 VA4 (lung fibroblast used as a control; CCL-95.1™) were purchased from the American Type Culture Collection. The BT-549 cell line was cultured in RPMI-1640 medium (Sigma-Aldrich; Merck KGaA) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) inactivated at 56°C for 30 min, 1% 0.2 M L-glutamine and 10 mg/ml bovine insulin (Sigma-Aldrich; Merck KGaA). The TOV-21G cell line was grown in high-glucose Dulbecco's modified Eagle's medium (Sigma-Aldrich; Merck KGaA) supplemented with 15% FBS and 1% 0.2 M L-glutamine, and the RKO-AS45-1 and WI-26 VA4 cell lines were cultured in RPMI-1640 medium supplemented with 10% FBS and 1% 0.2 M L-glutamine. Cells were incubated at 37°C in a humidified atmosphere enriched with 5% CO₂. The experiments were carried out obeying a certain passage number (between passages 4 and 6).

Everolimus (molecular weight, 958.224 g/mol) was suspended in dimethyl sulfoxide (DMSO) to achieve a concentration of 1 mM; for all assays, the final concentration of 100 nM diluted in culture medium was used.

Total RNA was extracted from tumor cells in culture before and after treatment with everolimus at 100 nM for 24 h and from the normal cell line WI-26 VA4 without treatment using the TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. The quantification of the total RNA extracted from cells was performed using a Nanovue™ Plus Spectrophotometer. The integrity of the extracted total RNA was evaluated by 1% agarose gel electrophoresis. A total of 2 µg total RNA was treated with the RNAse-Free DNase set (Qiagen), and complementary DNA (cDNA) synthesis was performed using the M-MLV Reverse Transcriptase^® kit (Sigma-Aldrich; Merck KGaA), Oligo dT (Eurofins Scientific) and dNTPs (Promega Corporation) according to the manufacturer's instructions.

Transcript analysis of ZEB1, TWIST1 and TGFB1 was performed by qPCR using PowerUp™ SYBR^® Green Master Mix (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. The primers used are shown in Table I. The TATA-binding protein (TBP) gene was used as the endogenous housekeeping gene. The reactions were subjected to specific amplification cycles (40 cycles of denaturation at 95°C for 15 sec and annealing/elongation at 60°C for 1 min) and the dissociation curve according to the manufacturer's instructions. The expression of lncRNA HOTAIR was evaluated using TaqMan^® Non-Coding RNA assay (Applied Biosystems; Thermo Fisher Scientific, Inc.). TaqMan^® Universal Master mix (Thermo Fisher Scientific, Inc.) was used according to the manufacturer's instructions, and TBP and 18S genes were used as endogenous controls, with the results normalized to the average of the endogenous controls. This assay was performed once in duplicate. Total RNA and cDNA-free samples were used as controls in each assay. Cycling and data collection were performed on a StepOne™ Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). Gene expression was quantified using the relative 2^−ΔΔCq quantification method (13). In order to determine the relative changes in the mRNA expression levels of TGFB1, TWIST1 and ZEB1 in the cells treated with everolimus compared with untreated samples, ΔCt values from three normalization controls were used: i) Untreated samples of each cell line; ii) untreated BT-549 samples; and iii) untreated WI-26 VA4 samples.

The cell lines (BT-549, 5×10⁴ cells/well; TOV-21G and RKO-AS45-1, 2.5×10⁴ cells/well) were seeded on Lab-Tek slides (Thermo Fisher Scientific, Inc.) and incubated overnight at 37°C in 5% CO₂. After treatment with 100 nM everolimus for 24 h, BT-549, TOV-21G and RKO-AS45-1 were fixed with 4% formaldehyde for 1 h, permeabilized with 0.1% Triton X-100 for 10 min and labeled with DAPI, Phalloidin and LysoTracker^® Red DND-99 probes (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions, all at room temperature. The labelling was assessed using an Axiovert 200M fluorescence microscope (Carl Zeiss AG) with AxioVision Rel. 4.8.2 software (Carl Zeiss AG). The fluorophores were read according to each emission spectrum (DAPI, 461 nm; Phalloidin, 518 nm; and LysoTracker, 590 nm). To evaluate whether everolimus affected the cytoskeleton of the cells, the nuclei of 25 cells labeled with DAPI and Phalloidin were measured (in the same slide). Using ImageJ version 1.50i software (National Institutes of Health), a region of interest was drawn manually around each LysoTracker-labeled cell, and the corrected total cell fluorescence (CTCF) was calculated by subtracting the fluorescence intensity of the selected cell area multiplied by the mean bottom well fluorescence (background).

The cell lines (BT-549, 5×10⁴ cells/well; TOV-21G and RKO-AS45-1, 2.5×10⁴ cells/well) were seeded on Lab-Tek slides (Thermo Fisher Scientific, Inc.) and incubated overnight at 37°C in 5% CO₂. Following 24-h treatment with 100 nM everolimus, the cells were fixed with 4% formaldehyde solution for 1 h, permeabilized with 0.1% Triton X-100 for 10 min and labeled for 45 min with anti-E-cadherin (1:50; cat. no. 610182; BD Biosciences), anti-P-cadherin (1:50; cat. no. 610228; BD Biosciences) and anti-fibronectin antibodies (1:50; cat. no. A21316; Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturers' instructions, all at room temperature. After incubation, the wells were washed with 1X PBS three times, and the wells incubated with the primary antibodies against E/P-cadherin were incubated with a FITC-conjugated anti-mouse IgG secondary antibody (1:128; cat. no. F5262; Sigma-Aldrich; Merck KGaA), whereas the wells incubated with anti-fibronectin were incubated with an Alexa 594-conjugated anti-chicken IgG secondary antibody (1:500; cat. no. A32759; Invitrogen; Thermo Fisher Scientific, Inc.) for 2 h, all at room temperature. After incubation, the wells were washed three times with 1X PBS, and the nuclei were labeled with DAPI. The markers were analyzed under the Axiovert 200M fluorescence microscope using AxioVision Rel. 4.8.2 software. The 25-cell CTCF in the same slide was also calculated using ImageJ software as aforementioned. The fluorescence intensity of the nuclei was subtracted from the integrated density obtained within the outline of each cell.

Cell viability in the presence of 100 nM everolimus was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide method (MTT) as previously described by Mitchell et al (14). Untreated cells and those treated with hydrogen peroxide were used as negative and positive controls, respectively. The absorbance measurement was performed using a spectrophotometer (SpectraMax^® M5e; Molecular Devices, LLC) at 550 nm. The viability of the treated cells was calculated by dividing the absorbance value of the treated cells by the mean of the untreated control multiplied by 100.

The clonogenic assay was performed as previously described by Franken et al (15). RKO-AS45-1 (500 cells) and TOV-21G (1,500 cells) cell lines were treated with 100 nM everolimus and seeded in 60-mm culture dishes in triplicate. The cells were incubated at 37°C for 7 days. Colonies with ≥50 cells were counted. The plating efficiency (PE) was determined by dividing the mean number of colonies of the controls by the number of plated cells. The surviving fraction (SF) was calculated by dividing the mean of colonies formed after treatment by the number of plated cells and multiplying the result by the PE. The result was multiplied by 100%. The colonies were analyzed with DAPI and Phalloidin probes as aforementioned.

RKO-AS45-1 cells were treated in individual wells with either: i) 1 µg their own total RNA; ii) 700 ng/ml TGF-α; or iii) 100 nM everolimus, and seeded in RPMI-1640 medium supplemented with 10% FBS, 100 U penicillin and 100 µg streptomycin in a Transwell^® system (Corning, Inc.) with a 3-µm polystyrene membrane and a growth area of 1.12 cm². Cell resistance was measured using the Millicell^® ERS 2 apparatus (EMD Millipore) at 0, 3, 24 and 48 h. To obtain the correct resistance in Ohms, the resistance value of the wells with medium was subtracted from the resistance value of the wells with cells, and the correct resistance value was obtained by multiplying this with the membrane area. This experiment was carried out in biological triplicates.

The STRING database (version 10.5) was used to evaluate PPIs (16). Classification of the mutations in EMT genes evaluated in the present study was performed using the somatic mutation database COSMIC v85 (https://cancer.sanger.ac.uk/cosmic). The molecular classification of RKO, the parental cell line of RKO-AS45-1, was used due to the absence of the data for RKO-AS45-1 cells in the database.

Statistical analyses were performed with SPSS 20 software (IBM Corp.). Differences in gene expression and cell resistance between groups were analyzed using one-way ANOVA with the least significant difference post hoc test. The results of the microscopy assays were analyzed using the non-parametric Mann Whitney test. The results of the cell viability and clonogenic assays were evaluated using the Kruskal Wallis test followed by the least significant difference test. P<0.05 was considered to indicate a statistically significant difference.

The results of the qPCR analysis demonstrated that TGFB1 appeared to be upregulated in the RKO-AS45-1 and BT-549 cell lines treated with everolimus independently compared with all three types of normalization control, whereas in the TOV-21G cell line, TGFB1 appeared to be downregulated (Fig. 1A). However, these results were not statistically significant. Significant differences between TWIST1 RQs normalized to the three controls were observed in the RKO-AS45-1 and BT-549 cell lines following everolimus treatment (Fig. 1B). TWIST1 was downregulated in RKO-AS45-1 compared with untreated BT-549 and WI-26 VA4 cells, but upregulated compared with untreated RKO-AS45-1 cells. In BT-549 cells, TWIST1 was upregulated compared with all three normalization controls. In TOV-21G cells, TWIST1 appeared to be downregulated compared with the control, and no statistical differences were observed between the analyzed groups. The analysis of ZEB1 RQ (Fig. 1C) demonstrated that the expression of its mRNA appeared to be downregulated in the three cell lines compared with that in untreated BT-549 cells, but upregulated in RKO-AS45-1 and TOV-21G cell lines compared with the respective untreated cell lines and WI-26 VA4. However, only the differences observed in the expression of ZEB1 in TOV-21G were statistically significant.

Only the BT-549 triple negative breast cancer cell line demonstrated amplified lncRNA HOTAIR following treatment with everolimus (Fig. 1D).

In untreated BT-549 cells, extensive cytoplasmic projections were observed before exposure to everolimus (white arrow; Fig. 2A). Analyses of RKO-AS45-1 (Fig. 2C) and TOV-21G (Fig. 2E) nuclei did not reveal any significant changes in cell size. However, after treatment with everolimus, the RKO-AS45-1 cells exhibited cytoplasmic extensions (white circle) similar to those of untreated BT-549 cells. BT-549 cells were negative for the epithelial markers E/P-cadherin (Fig. 2B), but positive for the mesenchymal marker fibronectin (Fig. 3). RKO-AS45-1 and TOV-21G cells were positive for E/P-cadherin (Fig. 2D and F, respectively), but most of the RKO-AS45-1 and TOV-21G cells were negative for fibronectin (Fig. 3).

Following treatment with everolimus, a loss of BT-549 cell cytoskeleton integrity was observed, as demonstrated by the CTCF analysis of cell nuclei measured (Fig. 4A). No statistical differences in lysosome labeling were observed between the untreated and everolimus-treated groups of BT-549 cells. Analysis of the lysosomal profile revealed a significant decrease in the CTCF of RKO-AS45-1 cells and a significant increase in the CTCF of TOV-21G cells treated with everolimus compared with untreated cells (Fig. 4A). After treatment with everolimus, no significant changes were observed for E/P-cadherin or fibronectin CTCF in the BT-549 cell line (Fig. 4B-D). Treatment with everolimus also did not affect the expression of E-cadherin in RKO-AS45-1 and TOV-21G cells. However, a decrease in P-cadherin expression and an increase in fibronectin expression were observed in the RKO-AS45-1 and TOV-21G cell lines treated with everolimus compared with the untreated control cells (Fig. 4B-D).

The MTT assay was performed to evaluate the survival of BT-549, RKO-AS45-1 and TOV-21G cells following treatment with everolimus to determine whether the 100 nM concentration of the drug may lead to an increase in cell proliferation (Fig. 5A). The viability of RKO-AS45-1 and TOV-21G cells treated with everolimus was higher compared with that of the positive and negative controls; however, the differences between the everolimus-treated and untreated groups were not significant. By contrast, the viability of BT-549 cells after treatment with everolimus was significantly lower compared with that of the untreated controls.

RKO-AS45-1 cells treated with 100 nM everolimus were observed to form colonies that were farther distanced from each other compared with those formed by untreated cells (Fig. 5B). A high SF was observed, especially for TOV-21G cells, but the difference between the treated and untreated cells was not statistically significant (Fig. 5C).

According to the aforementioned results, RKO-AS45-1 cells exhibit more EMT characteristics after treatment with 100 nM everolimus. Therefore, this cell line was selected for the TEER assay. The TEER of the RKO-AS45-1 monolayer started to decrease at 48 h post-treatment with TGF-α and everolimus, indicating changes in the integrity of the cell monolayer. However, no significant differences were observed among the treatments (Fig. 5D).

The analysis in the STRING database was performed to explore the PPIs of the EMT targets evaluated in the present study (Fig. 6). No mutations were identified in the EMT genes in BT-549 and TOV-21G cells (Table II). For the RKO cell line, missense mutations were identified in the TGFB1 and P-cadherin genes, and nonsense mutations were identified in ZEB1.