SW1736 and 8505C cells, derived from ATC (29–31), were cultured in RPMI-1640 medium (Euroclone S.p.A) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.), 2 mM L-glutamine (Euroclone S.p.A) and 50 mg/ml gentamicin (Gibco; Thermo Fisher Scientific, Inc.). Cells were cultured in a humidified incubator (5% CO₂ and 95% air at 37°C). Both cell lines were validated using short tandem repeat analysis and confirmed to be mycoplasma-free. SW1736 and 8505C cells were treated with DMSO (vehicle; Sigma-Aldrich; Merck KGaA), DHT (Selleck Chemicals) or cisplatin (Cayman Chemical Company).

In order to test cell viability, the MTT assay was used. SW1736 and 8505C cells (4×10³ cells/well) were seeded in 96-well plates. The following day, cells were treated with DHT (0.5, 1, 2 or 3 µM) and cisplatin (1 µM) at different concentrations. After 24, 48 or 72 h of incubation, 4 mg/ml MTT (Sigma-Aldrich; Merck KGaA) was added to the cell medium and cells were cultured for a further 4 h in the incubator in the dark. The supernatant was removed, 100 µl/well DMSO (Sigma-Aldrich; Merck KGaA) was added, and the absorbance at 570 nm was measured. All experiments were performed as six technical repeats and cell viability is expressed as the fold-change relative to the control (DMSO-treated cells).

Cell cycle distribution was determined using flow cytometry analysis of DNA content, as previously described (32). Briefly, SW1736 and 8505C cells were treated with 1.5 µM DHT or DMSO for 48 h. The cells were collected and fixed in cold 70% ethanol, then stained with propidium iodide solution containing RNase and Triton-X100 at 4°C for 30 min. Flow cytometry analysis was performed on a FACSCalibur instrument (BD Biosciences) using ModFit LT 5.0 analysis software (Verity Software House, Inc.). A minimum of 2×10⁴ cells were analyzed for each sample. All experiments were performed in triplicate.

To assess the effects of DHT on cell death, an Annexin V and propidium iodide assay was performed using the eBioscience Annexin V-FITC Apoptosis Detection kit (cat. no. 88-8005-74; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Briefly, DHT-treated SW1736 and 8505C cells were washed with cold PBS and resuspended in 195 µl binding buffer (BB; 10 mM; HEPES/NaOH, pH 7.4; 140 mM NaCl and 2.5 mM CaCl₂). A total of 5 µl fluorescein isothiocyanate-conjugated Annexin V (Annexin V-FITC) was added and samples were incubated for 10 min at room temperature. After washing, cells were resuspended in 190 µl BB in which 10 µl propidium iodide stock solution (final concentration 1 µg/ml) were added. Flow cytometry analysis was performed on a FACSCalibur (Becton-Dickinson and Company) and analyzed using the Summit software (Beckman-Coulter). All experiments were performed in triplicate.

Total protein was extracted from SW1736 and 8505C cells, treated with 1.5 µM DHT or DMSO, using a cell scraper and lysis buffer (50 mM Tris HCl, pH 8; 120 mM NaCl; 5 mM EDTA; 1% Triton; 1% NP40; 1 mM DTT), supplemented with phenyl-methylsulphonyl fluoride and protease inhibitors. Lysates were centrifuged at 13,000 × g for 10 min at 4°C, and supernatants were quantified using a Bradford assay.

For western blot analysis, 30 µg protein was loaded per lane on a 10% SDS gel, resolved using SDS-PAGE, then transferred to nitrocellulose membranes (GE Healthcare). The membranes were blocked at room temperature for 1 h using PBS-milk (PBS; 0.1% Tween-20; 5% non-fat dry milk). The membranes were then incubated overnight at 4°C with rabbit anti-actin antibody (1:1,000; cat. no. A2066; Sigma-Aldrich; Merck KGaA), mouse monoclonal anti-MAD2 (1:500; cat. no. sc-374131; Santa Cruz Biotechnology, Inc.), mouse anti-E-cadherin (1:1,000; cat. no. 14472), mouse anti-N-cadherin (1:1,000; cat. no. 14215), rabbit anti-vimentin (1:1,000; cat. no. 5741) (all Cell Signaling Technology, Inc.) or Apoptosis Western Blot Cocktail (pro/p17-caspase-3, cleaved PARP, muscle actin) (1:400; cat. no. ab136812; Abcam). This cocktail is designed to study the induction of apoptosis in response to various stimuli and includes monoclonal antibodies specific for procaspase-3, caspase-3 and PARP. In particular, the mouse PARP antibody of this cocktail detects only the apoptosis-specific 89-kDa PARP fragment (cleaved PARP). The following day, the membranes were incubated with peroxidase-conjugated anti-rabbit or anti-mouse IgG secondary antibody (both 1:4,000; cat. nos. A6154 and A9044, respectively; both Sigma-Aldrich; Merck KGaA) for 2 h at room temperature. A UVITEC Alliance LD (Uvitec Ltd.) western blot detection system with SuperSignal Technology reagent (Thermo Fisher Scientific, Inc) and the Alliance 1D Max software (Uvitec Ltd.) was used to visualize the signals.

SW1736 and 8505C cells were treated with vehicle (DMSO) or 1.5 µM DHT for 48 h, then seeded at 1.5×10³ cells/plate. Colonies were left to form for 21 days, then washed twice with PBS and fixed with pure methanol for 15 min at 4°C. Cells were then washed twice with PBS-Triton X-0.1% (Sigma-Aldrich; Merck KGaA) and incubated at room temperature with crystal violet (Sigma-Aldrich; Merck KGaA) solution for 30 min. Crystal violet was removed and stained colonies were washed, images were captured and colonies were counted by eye.

The clonogenic ability of the SW1736 and 8505C cells after treatment with 1.5 µM DHT was evaluated using a soft agar assay. Briefly, after 48 h of treatment, cells were collected, and 1×10⁴ cells were suspended in 4 ml complete medium containing 0.25% agarose (Sigma-Aldrich), then seeded to the top of a 1% agarose complete medium layer in 6-cm plates. The colonies were counted by eye in four different fields, under a Leica DMI-600B inverted microscope (Leica Microsystems Ltd.). Data are representative of three independent experiments.

Transwell membranes coated with Matrigel™ were used to measure the ability of cells to attach to the matrix, invade into and through the matrix, and migrate towards a chemoattractant. The invasion ability of ATC cells was evaluated after treatment for 72 h with 1.5 µM DHT. For each cell line, 2.5×10⁴ live cells per condition (DHT or vehicle) were separately re-suspended in culture media containing 10% FBS and seeded in a 24-well plate Transwell coated with Matrigel, as described by Justus et al (33). RPMI-1640 medium supplemented with 25% FBS was used as the chemoattractant in the lower chamber. After 24 h, cells were fixed with 70% ethanol and stained with crystal violet solution. After staining, the top of membrane was gently scraped to remove the cells that had not migrated; the cells that had migrated through the membrane toward the chemoattractant were attached on the underside of the membrane. The number of cells on the underside were counted in six different fields of view using an inverted microscope (Leica DMI-600B). Data are representative of the average of the sum of the cells counted in all these six fields in three independent experiments.

RNA was extracted from SW1736 and 8505C cells treated with vehicle (DMSO) or DHT using a RNeasy Mini kit (Qiagen GmbH) according to the manufacturer's instructions. In total, ~1 µg RNA (RNA integrity number >7) was used as the starting material for preparation of the library using a Universal Plus mRNA-Seq kit (cat. no. 0520-A01; Tecan Group, Ltd.) according to the manufacturer's protocol. RNA samples were quantified, and the quality was assessed using an Agilent 2100 Bioanalyzer RNA assay (Agilent Technologies, Inc.) and the final libraries were checked using both a Qubit 2.0 Fluorometer (Invitrogen; Thermo Fisher Scientific, Inc.) and Agilent Bioanalyzer DNA assay. The library final concentration was 68.7 nM. Libraries (1.4 nM) were then prepared for sequencing using the single-end 75 bp mode on a NextSeq 500 (Illumina, Inc.). The Bcl2Fastq version 2.20 in the Illumina pipeline was used for processing the raw data (format conversion and de-multiplexing); adapter sequences were masked with Cutadapt version 1.11 (34) from raw fastq data and the ERNE (35) software was used to remove lower quality bases and adapters. Reads were aligned to the reference hg38 genome/transcriptome using STAR software (36). Finally, assembly and quantitation of full-length transcripts representing multiple spliced variants for each gene locus was performed using the Stringtie tool (37). Differentially expressed genes were defined as those with a log₂ fold change >1.5 or <-1.5. Raw and processed data are available on the public online repository Gene Expression Omnibus (dataset no. GSE168616).

A total of 500 ng total RNA from SW1736 and 8505, extracted as described above, was reverse transcribed to cDNA using random hexaprimers and SuperScript III (SSIII) reverse transcriptase (Thermo Fisher Scientific, Inc.). Briefly, the first strand cDNA synthesis was created in a total volume of 20 µl with 5X First-strand buffer, DTT 0.1 M, RNase OUT Recombinant RNase Inhibitor (Thermo Fisher Scientific, Inc.), dNTPs 10 mM, random primers (Thermo Fisher Scientific, Inc.) and SSIII reverse transcriptase. The reverse transcription step involved incubation at room temperature for 5 min, 50°C for 60 min, 70°C for 15 min and 45°C for 5 min. Quantitative PCR was performed using PowerUP Sybr green master mix (Thermo Fisher Scientific, Inc.) on the QuantStudio3 system (Applied Biosystems; Thermo Fisher Scientific, Inc.), as previously described (38) and following the Standard cycling mode (primer T_m <60°C): 50°C for 2 min, 95°C for 2 min and 95°C for 15 sec, 60°C for 15 sec and 72°C for 1 min for 40 cycles. The QuantStudio Design and Analysis software (Applied Biosystems; Thermo Fisher Scientific, Inc.), was used to calculate mRNA levels with the 2^−∆∆Cq method (39) and β-actin was used as reference. All experiments were performed in triplicate. Oligonucleotide primers were purchased from Sigma-Aldrich; Merck KGaA, and the sequences of the primers are listed in Table I.

Data are presented as the mean ± standard deviation. All results were analyzed using the unpaired Student's t-test or one-way ANOVA in GraphPad Prism version 6 (GraphPad Software, Inc.). After one-way ANOVA, the Dunnett's post hoc test was performed. P<0.05 was considered to indicate a statistically significant difference.

In the first set of experiments, the effects of DHT on two human ATC cell lines (SW1736 and 8505C) were evaluated. To assess the effects of DHT on cell viability using several doses of DHT, an MTT assay was performed on ATC cells treated for 24, 48 or 72 h. As shown in Fig. 1A, 2 and 3 µM DHT treatment significantly reduced the viability of both ATC cell lines at all time points. In SW1736 cells, doses <2 µM did not significantly affect viability, whereas the 8505C cells were more sensitive to DHT treatment, since they exhibited a ~50% reduction in cell viability when treated with 0.5 and 1 µM DHT after 48 h. However, the effects of DHT were transient, and cell viability recovered at the 72-h time point. Based on these data, the median ED₅₀ of 1.5 µM was selected for treatment of both cell lines in subsequent experiments.

The aforementioned experiments did not distinguish whether the reduced viability was the effect of a decrease in cell metabolism or cell number. To evaluate the effects of DHT on cell cycle progression, flow cytometry analysis was performed after 72 h of DHT treatment. As shown in Fig. 1B and C, treated SW1736 and 8505C cells exhibited a slight reduction in the proportion of cells in the G₀/G₁ phase compared with vehicle-treated cells, significative only in SW1736 cells. By contrast, DHT treatment resulted in a significant increase in the proportion of cells in the S phase from 25.8 to 32.3% and from 43.2 to 55.3% in SW1736 and 8505C cells, respectively. Moreover, 8505C treated cells showed a significant reduction in the proportion of cells in the G₂/M. The notable increase in cells with reduced DNA staining compared with G₀/G₁ cells among treated cells was hypothesized to reflect an increase in the degree of apoptosis, as previously described (40).

To evaluate whether the decrease in cell viability observed after the treatments was due to apoptotic cell death, western blot analysis of caspase-3 activation (Fig. 2A and B) and cleaved-PARP protein (Fig. 2C-F) was performed. Caspase-3 activation was evaluated after 24 and 48 h of treatment with 0.75 or 1.5 µM DHT. In both cell lines, 48-h treatment with 1.5 µM DHT resulted in a significant increase in caspase-3 activation compared with the DMSO control. These conditions also resulted in a ~60-fold increase in the expression of cleaved-PARP levels compared with the control in both cell lines (Fig. 2E and F).

To better characterize the effects of DHT on cell death, an Annexin V/propidium iodide assay was performed. As shown in Fig. 2G, in both cell lines, 1.5 µM DHT treatment resulted in an increase in the proportion of necrotic (top left area of the plot) and late apoptotic cells (top right area of the plot).

To assess the effects of DHT on markers of aggressiveness in the two ATC cells, its influence on the ability of cells to form colonies in an anchorage-dependent or independent manner was assessed using a plate colony formation assay and a soft-agar colony formation assay. As shown in Fig. 3A and B, there was a significant reduction in the number of colonies in cells treated with 1.5 µM DHT. In 8505C cells, DHT reduced the number of colonies by ~10-fold, and its effects were more potent on the SW1736 cells, in which it completely abrogated colony formation ability. To better evaluate the effects of DHT on tumor cell aggressiveness, the anchorage-independent colony formation ability of SW1736 and 8505C treated with DHT was investigated by performing a soft agar colony formation assay. Treatment with 1.5 µM DHT for 48 h resulted in a significant reduction in the number of colonies in both cell lines relative to the respective vehicle-treated cells (Fig. 3C and D).

Since traversing the basement membrane by cancer cells is an important step in the metastatic process, the effects of DHT treatment on thyroid cancer cell invasiveness was assessed using Transwell invasion assays. Treatment with 1.5 µM DHT resulted in a significant decrease in cell invasion in both ATC cell lines (Fig. 3E and F).

To evaluate the effects of DHT on gene expression, high-throughput RNA-seq analysis was performed on the SW1736 and 8505C cells treated with 1.5 µM DHT for 24 h. After filtering out low quality reads, the comparison between cells treated with DMSO or DHT showed that 1,805 and 503 genes were differentially expressed (log₂ fold change >1.5) in the DHT-treated SW1736 and 8505C cells, respectively. In SW1736 cells, 813 genes were upregulated and 992 were downregulated, whereas in 8505C, 350 genes were upregulated and 153 were downregulated in response to DHT treatment. Among these, 41 genes were commonly upregulated and 29 commonly downregulated (Fig. 4).

Gene Ontology analysis was then performed using the Panther Classification System (41). Briefly, differential expressed gene lists for each cell line (separately for up- and down-regulated genes) were loaded into PANTHER (version 16.0) and mapped to its pathway database. When the PANTHER pathway database was interrogated with the commonly dysregulated genes, no pathway has been identified as significantly deregulated. Regarding the altered genes in each cell line, the top 10 up- and downregulated pathways of each cell line are shown in Fig. 5 (and Table SI,Table SII,Table SIIITable SIV). Although the altered gene expression involved different genes, the analysis identified some commonly dysregulated pathways in both cell lines, which included ‘Wnt signaling’ (P00057), ‘apoptosis signaling’ (P00006) and ‘cadherin signaling pathway’ (P00012).

To characterize the effects of DHT on ATC cell aggressiveness at the molecular level, the expression of five universally recognized genes associated with EMT (42), a phenotypic change which endows cancer cells with increased migratory and invasive capacity, was analyzed.

The changes in mRNA levels for the EMT-related genes CDH1, CDH2, VIM, TWIST, ZEB1 and ZEB2 are shown in Fig. 6A. Treatment with 1.5 µM DHT resulted in a significant increase in the expression of CDH1, a marker of the epithelial phenotype, in treated cells compared with those treated with DMSO. Additionally, DHT treatment significantly decreased the gene expression levels of all other markers, which are associated with the mesenchymal phenotype, in both cell lines, except for CDH2, confirming that DHT modulated EMT.

Among the EMT-related genes assessed, CDH1, CDH2 and VIM are particularly interesting, as they are phenotypic markers of EMT. Thus, their protein expression levels were also assessed. VIM and CDH2 protein expression levels were significatively reduced following treatment with DHT (Fig. 6B and C). Despite the fact that CDH1 mRNA levels were increased following DHT treatment, the protein expression levels remained undetectable in the treated cells (data not shown).

Since DHT is known to inhibit the activity of the RNA-binding protein HuR (24), to verify the involvement of HuR in the effects of DHT on viability, apoptosis and tumor aggressiveness in ATC cells, MAD2, a target of HuR, was then analyzed. MAD2, which is one of the major factors involved in the spindle checkpoint, has been shown to be implicated in cancer progression (43,44) and is overexpressed in thyroid neoplasms (45). Thus, the effects of DHT on MAD2 protein expression were examined using western blotting (Fig. 7). After 48 h of treatment with 1.5 µM DHT, there was a significant decrease in MAD2 protein expression levels in both cell lines. This result demonstrated that MAD2 may be involved in anticancer effect of DHT in ATC cell lines.

Based on the results of the DHT treatment, the effects of its co-administration with cisplatin, one of the few chemotherapeutic agents used for ATC treatment, were then examined. The effect of their combined administration on the viability of SW1736 and 8505C cells was assessed following 24, 48 or 72-h treatment with cisplatin and DHT alone or combined. Since synergy is considered as the interaction between two or more drugs, which results in a total effect of the drugs, which is greater than the sum of the individual effects of each compound, drug concentrations of 1 µM for both compounds were used, as this concentration did not exert any notable effects when the compounds were administered alone.

As shown in Fig. 8, DHT and cisplatin 1 µM alone did not affect cell viability after 72-h treatment, in either cell line. However, when co-administered at the same individual concentration of 1 µM, there was a significant decrease in cell viability of around 60–70% in both the SW1736 and 8505C cells.

Thus, the effects of DHT-cisplatin combination on apoptosis and other markers of tumor aggressiveness were assessed next (Fig. 9). Western blot analysis indicated that cleaved PARP protein levels were increased 6-fold after cotreatment compared with 1 µM cisplatin alone in SW1736 cells. In 8505C cells, combined administration of DHT and cisplatin resulted in a significant increase in cleaved-PARP protein expression compared with either treatment alone. Similar to what was observed when DHT was used at an ED₅₀ dose, the lower dose of 1 µM resulted in MAD2 downregulation in the ATC cell lines. Moreover, cotreatment resulted in a greater decrease in MAD2 protein expression in the SW1736 and 8505C cells, compared with DHT or cisplatin alone.

The effects of the combination of the two compounds on the expression of EMT-related genes were then evaluated. During EMT, CDH1 expression is significantly decreased, whereas CDH2 expression is increased. Thus, CDH2/CDH1 expression ratio is widely considered a marker of EMT (46). As shown in Fig. 10, in SW1736 cells, DHT alone did not induce an increase in the CDH2/CDH1 ratio, whereas cisplatin, alone and combined with DHT, resulted in a significant reduction in the CHD2/CDH1 ratio. In SW1736, unlike what observed using the ED₅₀ dose of DHT (1.5 µM), treatment with 1 µM DHT resulted in an increase in VIM mRNA expression levels, compared with the control. When cells were cotreated with DHT and cisplatin, the mRNA expression of VIM returns to comparable levels with control, while ZEB1 decreased. In the 8505C cells, the observed effects on EMT-related genes were not as prominent and the differences were not significant.