The thyroid cancer cell lines KTC-1, BCPAP, 8505C, and KHM-5M cells were obtained from the Cell Culture Collection of the Chinese Academy of Sciences (Shanghai, China). The sources of the TPC-1 (a thyroid cancer cell line) and Nthy-ori 3-1 cells (an immortalized thyroid epithelial cell line) were as previously described (20). The KTC-1, BCPAP, and KHM-5M cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 (Gibco) with 10% fetal bovine serum (FBS; Gibco). The 8505C, TPC-1, and Nthy-ori 3-1 cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco) with 10% FBS. The cells were routinely cultured at 37°C in an atmosphere containing 5% CO₂.

SS was purchased from Sigma (Cat#: S9133). It was dissolved in sterile water in a biosafety cabinet to a concentration of 50 mg/mL for storage at -20°C. A working solution was prepared using sterile water before use. N-acetylcysteine (NAC) was also purchased from Sigma (Cat#: A7250). The NAC ampoule was fully dissolved in sterile water in a 37°C water bath (to accelerate the dissolution) and then stored at -20°C. The working concentration of NAC used for cell culture was 2 mM, and the pH was adjusted to 7.0 before use.

Cells were seeded in 96-well flat-bottom plates at a density of 1000 cells per well. After 24 h, the cells were exposed to different concentrations of SS with or without NAC. Cell viability was evaluated using Cell Counting Kit-8 (CCK8) assays (Dojindo Laboratories). After incubation with a medium containing CCK8 solution for 2 h, the absorbance was assessed using a Multimode Microplate Reader (Thermo Fisher Scientific). Half maximal inhibitory concentration (IC50) values were calculated using GraphPad 5.0.

Cells were seeded in 6-well plates and treated with different concentrations of SS for 2 weeks. Colonies were then fixed with 4% paraformaldehyde for 15 min and stained with 5% crystal violet for 15 min. Photos of the colonies were taken after washing with double-distilled water.

Cells were seeded in 6-well plates and treated with different concentrations of SS for 24 h. The cells were then harvested and fixed in 70% ethanol at -20°C overnight. Thereafter, they were incubated with RNase A and stained with propidium iodide (PI) at 37°C for 1 h. The cells were evaluated using a flow cytometer (Thermo Fisher Scientific). The distribution of cells according to the cell cycle phases was then analyzed using FlowJo 10.6.2.

Cells were seeded in 6-well plates and treated with different concentrations of SS for 24 h. Cells in the medium and adherent cells were all collected and stained using an Annexin V- fluorescein isothiocyanate (FITC)/PI Apoptosis Detection Kit (AD10; Dojindo Laboratories) according to the manufacturer’s protocol. Apoptotic cells were then analyzed using flow cytometry.

Cells were seeded in 96-well plates (black plates, clear bottom with lid) and treated with different concentrations of SS for 24 h. The cells were then stained with Caspase-3/7 Green ReadyProbes Reagent (R37111; Thermo Fisher Scientific) and Hoechst-33342 (Beyotime Institute of Biotechnology) for 30 min, and the fluorescent signals were detected using FITC and 4’,6-diamidino-2-phenylindole (DAPI) filters.

Cells were harvested and lysed using lysis buffer for 15 min at 4°C. The protein concentration was assessed using a bicinchoninic acid (BCA) assay. Western blotting was carried out as described previously (21). The primary antibodies were as follows: phospho-AKT S437 (4060T; Cell Signaling Technology), total AKT (4691T; Cell Signaling Technology), phospho-mTOR S2448 (ab109268; Abcam), total mTOR (ab32028; Abcam), phospho-ERK1/2 T202/Y204 (#4370; Cell Signaling Technology), total ERK1/2 (#4695; Cell Signaling Technology), and β-actin (ab8226; Abcam).

Cells were seeded in 96-well clear-bottom black plates and treated with 10 µM SS or phosphate-buffered saline (PBS). For some cells, NAC (2 mM) was added to eliminate the cellular ROS. After 6 h of treatment, the cells were incubated for 1 h in a medium containing 10 µM dichlorofluorescin diacetate (DCFH-DA) at 37°C. The cells were then washed twice and re-suspended in a growth medium for detection. The fluorescent signals were detected by microscope.

After treating thyroid cancer cells with SS or PBS (control), total RNA extraction, cDNA synthesis, and qRT-PCR were performed as described previously (21). The expression of β-actin was used as the reference. The sequences of the primers are presented in Supplementary Table 1 .

Total RNA was extracted using TRIzol reagent and purified using oligo(dT)-attached magnetic beads. The purified RNA was fragmented into small pieces, and the first-strand cDNA was generated using random priming, followed by second-strand cDNA synthesis. After end repair, the cDNA was amplified by PCR, and the products were validated using an Agilent Technologies instrument. The final library was generated after heat denaturation and circularization of the PCR products. Thereafter, single-end 50-base reads were generated on a BGIseq500 platform (BGI-Shenzhen, China). The sequencing data were filtered using SOAPnuke, stored in FASTQ format, and mapped to HISAT2 (v2.0.4). Expression levels were calculated using RSEM (v1.2.12). The heatmap was constructed using pheatmap (v1.0.8). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed using Phyper. There were two biological replicates for each experiment. A q value ≤0.05 based on the Bonferroni correction was considered statistically significant.

EIF4EBP3-targeting siRNA and negative control siRNA were purchased from YouBio (Changsha, China). Thyroid cancer cells were seeded in 6-well plates, cultured for 24 h until they reached 40–50% confluence, and then transfected with EIF4EBP3-targeting siRNA or negative control siRNA using a Lipo3000 Transfection Kit (Invitrogen) according to the manufacturer’s instructions. After 48 h of transfection, the cells were harvested and used for further experiments. The sequence of the negative control siRNA was: 5’- UUCUCCGAACGUGUCACGUTT-3’. The sequence of the EIF4EBP3-targeting siRNA was: 5’- CGCACAAUUUGAAAUGGAC-3’.

Animal studies were carried out according to the Guidelines of the Animal Investigation Committee and approved by the Ethical Committee of The First Affiliated Hospital of Sun Yat-sen University. KHM-5M cells (5×10⁶) or BCPAP cells (5×10⁶) were injected subcutaneously into the right flank of 4-week-old BALB/c nude mice. The mice with xenografts were then given 2, 4, or 8 mg/kg/day SS (200 µL/day; Cat#: S5261; Sigma) or saline through oral administration. For the SS+NAC group, pH-corrected NAC (30 mM) was added to the drinking water. Bodyweight and tumor size (using digital calipers) were assessed every 3 days, and the tumor volumes were calculated according to the formula: length × width² × 0.5. All mice were sacrificed at the end of treatment. Tumor tissues and major organs were collected for immunohistochemical and hematoxylin and eosin (HE) staining.

Tumor tissues and major organs were fixed with 4% paraformaldehyde, dehydrated, embedded in paraffin, and cut into 5-µm-thick sections. The tissue sections were dewaxed (by heating at 60°C for 2 h), washed in xylene, and then rehydrated using a graded ethanol series. After antigen retrieval using microwave heating and blocking with 20% goat serum, the tissue sections were incubated with primary antibody (anti-Ki67 antibody; ab15580; Abcam) overnight at 4°C. Subsequently, immunodetection was carried out using a 3,3’-diaminobenzidine (DAB) peroxidase substrate kit (K5007; Dako) according to the manufacturer’s protocol. For HE staining, the slides were stained with hematoxylin and eosin sequentially after dewaxing and rehydration.

Tissue sections were dewaxed and rehydrated as mentioned above, and then incubated with proteinase K for 30 min at 37°C. After washing with PBS, the tissue sections were incubated with TUNEL reaction solution for 60 min at 37°C in the dark and then dyed with DAPI and fixed in fluorescent mounting media.

Student’s t-test and one-way ANOVA analysis were used to determine statistical significance. Statistical analysis was performed using GraphPad Prism 5.0 (GraphPad Software, Inc., La Jolla, CA, USA). A p-value of <0.05 was considered statistically significant.

To investigate the effect of SS on thyroid cancer, we first evaluated the cell viability by treating six thyroid cell lines with SS. As shown in Figure 1A , the viability of the five thyroid cancer cell lines treated with SS were all significantly reduced compared to the control group in a time- and dose-dependent manner. However, selenite had no significant inhibitory effect on the viability of the immortalized thyroid epithelial cell line Nthy-ori 3-1. The IC50 of selenite was higher in the thyroid cancer cells than in the Nthy-ori 3-1 cells ( Figure 1B ). The selenite dose-dependent inhibition of thyroid cancer cells was also demonstrated by the colony formation assays ( Figures 1C and S1A ). Notably, there was no significant difference in the anticancer effect of selenite between the BRAF wildtype (WT) tumor cells (TPC-1, KTC-1, and KHM-5M) and the BRAF^V600E-positive tumor cells (BCPAP and 8505c). Taken together, these data showed that thyroid cancer cells were more sensitive to SS compared to human thyroid epithelial cells.

To investigate whether the modulation of the cell cycle contributes to the anti-proliferation effect of SS, we analyzed the cell cycle distribution in thyroid cancer cells by flow cytometry. Compared to the control group, the percentage of cells at the G0/G1 phase was increased in the SS group, while cells in other phases were decreased ( Figure 2A ). These results implied that SS effectively induced cell cycle arrest at the G0/G1 phase in a dose-dependent manner.

Typical morphological features of apoptosis were observed in thyroid cancer cells during prolonged selenite treatment ( Figure S1B ), suggesting that SS also induced apoptosis of thyroid cancer cells. Next, we performed Annexin V-FITC/PI staining assays followed by flow cytometry, and caspase activity detection, to confirm this. As shown in Figures 2B, C , the proportion of Annexin V-positive cells and the activity of caspase 3/7 were significantly increased after SS treatment. Moreover, the anti-apoptotic protein Bcl-2 was decreased after SS treatment ( Figure S1C ). Taken together, the results show that SS was capable of impeding cell cycle progression and triggering apoptosis in thyroid cancer cells.

We next conducted a comprehensive RNA-seq analysis of early transcriptomic changes in KHM-5M cells to identify the underlying mechanism of SS. The principal component analysis clearly separated the SS-treated cells from the control cells. SS affected the expression of 4564 genes, of which 176 were altered by ≥2 fold ( Figure 3A ). We performed a KEGG pathway enrichment analysis of the differentially expressed genes to determine the critical pathways underlying the anticancer effect of SS. The top 10 enriched pathways, including the oxidative phosphorylation and mTOR signaling pathways, are shown in Figures 3B, C . Specifically, 105 genes were enriched in the oxidative phosphorylation pathway ( Figure 3D ). The expression of the top 10 enriched genes was validated by qRT-PCR in KHM-5M and three other thyroid cancer cells ( Figures 3E and S2A ). Among the genes enriched in the oxidative phosphorylation pathway, most encode components of the mitochondrial electron transport chain (mETC) complex, which is an important source of ROS. These results suggest that SS upregulated intracellular ROS levels by promoting mitochondrial oxidative phosphorylation.

As the transcriptomic analysis suggested that oxidative phosphorylation is involved in the effects of SS, we detected the intracellular ROS levels in thyroid cancer cells using DCFH-DA. As shown in Figure 4A , SS treatment for 6 h significantly increased the cellular ROS levels, which were restored by the ROS inhibitor NAC (2 mM) ( Figure 4A ). Moreover, the SS-induced reduction in cell viability was restored by NAC ( Figure 4B ), while the SS-induced cell cycle arrest and apoptosis were also reversed by NAC ( Figures 4C, D ).

To confirm that the role of ROS regarding the effects of SS, we also performed RNA-seq in SS+NAC treated cells. There was distinct segregation between the control and SS groups but not between the control and SS+NAC groups ( Figure S2B ). The differentially expressed genes between the SS and SS+NAC groups were enriched in the oxidative phosphorylation pathway ( Figure S2C ). These results strongly indicated that ROS induction played a key role in mediating the effect of SS in thyroid cancer cells.

The AKT/mTOR and MAPK pathways are major pathways in the development of thyroid cancer. Therefore, we investigated the effect of SS on these two pathways in thyroid cancer cells. As shown in Figure 5A , SS treatment significantly decreased the levels of phosphorylated and total AKT in a dose-dependent manner, and the phosphorylation levels of mTOR were correspondingly reduced. In contrast, there were no changes in the protein levels of key molecules in the MAPK/ERK pathway, including phosphorylated and total ERK. Importantly, the inhibitory effect of SS on the AKT/mTOR pathway was reversed by NAC treatment ( Figure 5B ).

4E-BP3, encoded by the EIF4EBP3 gene, is an important molecule that mediates the effects of mTOR inhibitors (22). Our RNA-seq results showed that SS significantly upregulated EIF4EBP3 mRNA in KHM-5M cells, which was confirmed by qRT-PCR in KHM-5M and BCPAP cells ( Figure 5C ). Based on these results, we speculate that 4E-BP3 may be involved in the anticancer effects of SS. To confirm our hypothesis, we transiently knocked down 4E-BP3 using siRNA and assessed the cell behaviors. Although there was no significant difference in cell proliferation between WT and 4E-BP3-knockdown cells under normal conditions ( Figure 5D ), the latter cells had higher viability and clonogenicity after SS treatment ( Figures 5E, F ), and a significantly lower proportion of G0/G1 cells after SS treatment ( Figure 5G ). Taken together, our findings indicate that the effects of SS on thyroid cancer cells were mediated by the AKT/mTOR/4E-BP3 axis in a ROS-dependent manner.

To investigate the anticancer effect of SS in vivo, thyroid cancer cells (KHM-5M or BCPAP) were injected into the right flank of BALB/c nude mice. Mice with xenografts were treated with 2, 4, or 8 mg/kg SS or saline (control) daily via oral administration. As shown in Figure 6A , xenograft tumors treated with SS exhibited slower growth than tumors in the control group, and the effect was dose-dependent. NAC treatment reduced the inhibitory effect of SS on the xenograft tumors ( Figures 6B, C ). Next, we evaluated the effect of SS on cell proliferation and apoptosis in the xenograft tumors using Ki67 staining and TUNEL assays. The results showed that the proportion of Ki67-positive cells in tumor tissues was significantly lower in the SS group than the control group, while the percentage of TUNEL-positive cells was significantly higher. The effects of SS on cell proliferation and apoptosis in xenograft tumors were reversed by NAC ( Figures 6D, E ). There were no significant differences in body weight between the experimental and control groups ( Figure S3A ). The toxicity of SS on the major organs of mice was evaluated by HE staining, with no obvious damage being observed between the SS and control groups ( Figures S3B, C ). These findings indicated that SS treatment inhibited thyroid cancer via ROS induction in vivo.