2.1 IR induces RET/PTC1 rearrangement in thyroid follicular epithelial cells in a dose-dependent manner

It could be detected when RET/PTC1 rearrangement occurs in one cell and is mixed in 10⁶ cells (Figure S2A,B). RET/PTC1 rearrangement was not observed in either 0.2 or 0.5 Gy after 3 days of exposure to IR (Figure S2C). Three RET/PTC1 rearrangement-positive wells were detected in the 0.2 Gy group, six RET/PTC1 rearrangement-positive wells were detected in the 0.5 Gy group at the sixth day (Figure S2C), and 6 days was determined as the collection time in the following experiments.

To further identify the association between the dose of IR and the number of RET/PTC1 rearrangement-positive wells in normal human thyroid epithelial cells, N-thy-ori-3-1 cells were exposed to five doses (2/1/0.5/0.2/0.1/0.05 Gy), and detected the RET/PTC1 rearrangement at sixth day after IR (Figure 1A and Table 1). The result showed that the RET/PTC1 rearrangement increased after IR in a dose-dependent manner (Figure 1B,C). Therefore, IR-induced RET/PTC1 rearrangement might be a potential mechanism of IR-induced PTC.

2.2 RET/PTC1 rearrangement promotes PTC tumorigenesis and progression in vitro and in vivo

We further assessed function of RET/PTC1 rearrangement in PTC and discovered that over-expression of RET/PTC1 in the N-thy-ori-3-1 cells could activate AKT, ERK, and STAT3, all of which have been shown to be oncogenesis key signals that can be activated by RET/PTC1 (Figures 2A and S3A). In contrast, the inhibition of RET/PTC1 in the TPC-1 cells could inactivate these oncogenesis signals (Figure 2A and S3B). In vitro, over-expression of RET/PTC1 in N-thy-ori 3-1 cells markedly increased the cell proliferation, as determined via cell counting assays (CCK-8) and plate colony formation assays (Figure 2B,C). However, the CCK-8 was used to identify that RET/PTC1 deletion could reduce the cell proliferation (Figure 2B). Concordantly, RET/PTC1 depletion also decreased the number of cell colonies, as determined by plate colony formation assays (Figure 2C). Furthermore, we verified the oncogenicity of RET/PTC1 in vivo. We observed that over-expression of RET/PTC1 in N-thy-ori 3-1 cells leads to a significant increase in the growth of subcutaneous xenograft tumors in nude mice (Figure 2D), including the sizes and weights of tumors (Figure 2E,F). In addition, knockdown of RET/PTC1 in TPC-1 cells leads to a significant reduction in the growth of subcutaneous xenograft tumors in nude mice (Figure 2H,I). These results indicated that RET/PTC1 drives the formation and progression of PTC in vitro and in vivo.

2.3 NHEJ repair contributes to RET/PTC1 rearrangement induced by IR in thyroid follicular epithelial cells

To determine the underlying mechanism of RET/PTC1 rearrangement, we found that γ-H2AX and pCHK1 were highly expressed at 0.5 h exposed with 0.5 Gy IR, indicating that DNA damage reached its maximum after 0.5 h of IR (Figure 3A). The NHEJ repair pathway inhibitor SCR7 and the homologous recombination (HR) repair pathway inhibitor RI-1 were employed to target DNA ligase IV and Rad51, respectively, which are key proteins in both pathways. A more increase in γ-H2AX expression was detected after 0.5 Gy of IR combining use of SCR7 and RI-1 than IR, and pCHK1 showed a same increase after the use of NHEJ inhibitors (Figure 3B). NHEJ and HR repair efficiency experiments showed that RI-1 decreased the HR repair efficiency and increased the NHEJ repair efficiency (Figure 3C). Conversely, SCR7 decreased the NHEJ repair efficiency and increased the HR repair efficiency (Figure 3D). These findings indicate that NHEJ inhibition promoted HR repair and that HR increased that's of NHEJ (Figure 3C,D). Next, we investigated whether the DNA damage repair efficiency can regulate the rate of RET/PTC1 arrangement. The results showed that the number of RET/PTC1 rearrangement-positive wells significantly increased after HR inhibition and decreased after NHEJ inhibition (Figure 3E,F and Table 2). In addition to canonical NHEJ and HR, the MMEJ is also an important DNA damage repair mechanism. To ascertain the role of MMEJ on RET/PTC1 rearrangement induced by IR, the MMEJ inhibitor Olaparib was employed to treat N-thy-ori-3-1 cells. Compared with NC group, the NHEJ and HR repair efficiency showed no significant alterations after treated with Olaparib (Figure S4A). Furthermore, the positive rate of IR-induced RET/PTC1 rearrangement remained unmodified upon Olaparib treatment (Figure S4B,C). Additionally, the combining use of SCR7 and RI-1 could also reduce the formation of RET/PTC1 induced by IR, and this reduction extend is consistent with the use of SCR7 (Figure S4B,C and Table S1). These findings indicted that the NHEJ not HR or MMEJ is the cause of IR-induced RET/PTC1 rearrangement.

To further elucidate the involvement of NHEJ in RET/PTC1 rearrangement induced by IR. Studies have shown that the cell cycle phase is associated with the selective of DNA damage repair mechanisms: cells in the G1 phase predominantly utilize the NHEJ pathway, whereas cells in the G2 phase are more inclined to employ the HR pathway.^{24, 25} Therefore, the N-thy-ori-3-1 cells were treated with thymidine to achieve cell synchronization, and the G1 and G2 phase cells were separately exposed to IR (Figure S5A–C). Results indicated a significant increase in RET/PTC1 rearrangement in G1 phase cells compared to the control IR group (Figure 3G,H). In contrast, a notable decrease in RET/PTC1 rearrangement was observed in G2 phase cells relative to the IR group (Figure 3G,H and Table 3). These findings indicate that the induction of RET/PTC1 rearrangement by IR is specific to the G1 phase, emphasizing the role of NHEJ in this process.

2.4 DNA-PKcs-mediated NHEJ contributes to IR-induced RET/PTC1 rearrangement in thyroid follicular epithelial cells

In view of the pivotal role of DNA-PKcs in response to IR and NHEJ, we found that the pDNA-PKcs was highly expressed at 0.5 Gy after 0.5 h (Figure 3A). Inhibition of DNA-PKcs with siRNA-DNA-PKcs (siDPK) and kinase inhibitor Nu7441 could significantly promoted the IR-induced pCHK1 expression increase (Figure 4A) and decrease the NHEJ repair efficiency (Figure 4A–C). To investigate the role of DNA-PKcs in IR-induced RET/PTC1 rearrangement generation, N-thy-ori-3-1 cells were subjected to simultaneous DNA-PKcs inhibition and 0.5 Gy IR treatment using siDPK or Nu7441. The results showed a significant decrease in the number of RET/PTC1 rearrangement-positive wells following DNA-PKcs inhibition (Figure 4D,E and Table 4), suggesting that DNA-PKcs-mediated NHEJ was critical for the occurrence of RET/PTC1 rearrangements induced by IR. Taken together, these results demonstrated that DNA-PKcs-mediated NHEJ plays a crucial role in the occurrence of IR-induced RET/PTC1 rearrangement.

4.1 Cell culture

The human normal thyroid epithelial cell line (N-thy-ori-3-1) and thyroid papillary carcinoma cell line (TPC-1: CVCL_6298), purchased from the Pricella Life Science & Technology, were grown in Roswell Park Memorial Institute 1640 (RPMI 1640 Sigma, Inc.), supplemented with 1% antibiotics/antimycotics (Invitrogen) and 10% fetal bovine serum (FBS) (ExCell Bio). The HLF cell line (Human Lung Fibroblasts) purchased from Pricella Life Science & Technology, was cultured in Dulbecco's Modified Eagle Medium (DMEM, Sigma, Inc.) supplemented with 10% FBS (ExCell Bio) and 1% antibiotics/antimycotics (Invitrogen).

4.2 Cell irradiation

Trypsin digestion of well-grown N-thy-ori-3-1 cells was carried out and counted to ensure that the number of cells per well in a six-well plate was 2 × 10⁴ cells. After that, the cells were passaged into six-well plates for 16 h and then exposed with a single dose of IR from a γ ⁶⁰Co source at a dose rate of 60 cGy/min. Cobalt-60 (⁶⁰Co), an artificial radionuclide that emits high-energy electrons, is a commonly acceptable source of gamma-ray radiation. Cells were cultured in a humidified thermostatic incubator at 37°C containing 5% CO₂, according to the experimental design, following grown for 3−6 days and harvested.

4.3 Gradient dilution experiment

To assess the sensitivity of the nested PCR for the detection of RET/PTC1, and to determine the number of cells to be inoculated and the limits of the assay for each subsequent independent experiment. A gradient dilution experiment was performed by mixing 10⁶, 10⁵, 10⁴,10³, 10², 10¹, and 10⁰ TPC-1 cells with 10⁶ thyroid follicular epithelial cells to test the accuracy of the experiment. The sensitivity of the assay was determined to be one positive cell mixed in 10⁷ during this study. To ensure the accuracy of the experiments, 10⁶ cells were used, and irradiation was performed after inoculating them in six-well plates.

4.4 Detection of RET/PTC1 rearrangements

Total RNA was isolated by TRIzol from cells (Ambion, Thermo Fisher Scientific) and eluted in 20 µL of RNase/DNase-free buffer (Biomed, RA114-02). A Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific) was employed to determine RNA concentration and quality. RNA reverse transcription was performed with Rever Tra Ace qPCR RT Master Mix and gDNA Removal Kit (Large Edition). The reverse transcription system was carried by Applied Biosystems PCR (Ambion, Thermo Fisher Scientific). The first and the nested PCRs were carried out with 35 cycles in Applied Biosystems PCR (Applied Biosystems). The first-round PCR products were pooled, diluted 1:5, and then, 1 µL was used for one nested PCR. The PCR amplification products were analyzed with 2%−3% agarose (Beyotime, ST004L) gel electrophoresis and ethidium bromide staining (Thermo Fisher Scientific, 15585011). The specificity of the amplification product with predicted size was confirmed by Sanger sequencing. TPC-1 cells were used as a positive control and HLF cells were used as a negative control.^{40, 66} The primers used are listed in Table S2.^{40, 67}

4.5 Antibodies and chemicals

Anti-GAPDH (Santa Cruz, sc-32233), anti-H2AX (Santa Cruz, sc-517336), anti-γH2AX (Santa Cruz, sc-517348), anti-CHK1 (Santa Cruz, sc-84084), anti-pCHK1 Ser345 (CST, 2348S), anti-DNA-PKcs (Invitrogen; Thermo Fisher Scientific, Inc.; MA5-13238), anti-pDNA-PKcs S2056 (Abcam; cat. no. ab18192), anti-STAT3 (CST, 9139S), anti-p-STAT3-Tyr705 (CST, 9145S), anti-AKT (CST, 9272S), anti-pAKT-Ser473 (CST, 4060S), anti-ERK1/2 (CST, 9102S), and anti-p-ERK1/2-Thr202/Tyr204 (CST, 9102S). All antibodies were diluted at 1:1000. The chemical Rad51 inhibitor RI-1 (cat. no. S8077), DNA ligase IV inhibitor SCR7 (cat. no. S7742), DNA-PK inhibitor Nu7441 (cat. no. S2638), Parp1 inhibitor Olaparib (cat. no. S1060), and thymidine (cat. no. S4803) were purchased from Selleck Chemicals.

4.6 Western blotting procedures

At 4°C, cells were washed with phosphate buffered saline(PBS), collected into EP tubes by centrifuging at 3000 rpm for 3 min, lysed on ice for 30 min with RIPA (Beyotime, P0013B), and centrifuged at 12,000 rpm for 10 min. The protein concentration was measured by Nanodrop 2000c spectrophotometer and 40 µg was loaded per lane on 6%−12% SDS–PAGE gels according to molecular weight of proteins. Proteins were transferred to nitrocellulose membranes and blocked with 5% bovine serum albumin(BSA) for 1 h at room temperature. Membranes were incubated overnight at 4°C with primary antibodies. After washing, membranes were incubated with secondary antibodies (1:4000) at room temperature for 1 h. After washing, SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific) was uniformly added to the membrane. The bands were visualized using an Image Quant LAS 500 and the Image Quant LAS 500 1.1.0 software (GE Healthcare Life Sciences).

4.7 Transfection

The siRNA and plasmid used in the present study were purchased from Gene Pharma (Gene Pharma); the sequences were as follows: siRET/PTC1.1—5′-AACTATCAAACGTGTCC-3′; siRET/PTC1.2—5′-AACATCTCGGGACAAGC-3′; siRET/PTC1.3—5′-TATCTTCAGCACCTTGTTC-3′; siDPK1—5′-GGGCGCTAATCGTACTGAA-3′; and siDPK2—5′-GGAAGAAGGUGAAGGAGAATT-3′. All transfections were conducted using lipofectamine 2000 (Invitrogen, 11668019) according to the manufacturer's instructions.

4.8 NHEJ assay

NHEJ-GFP plasmid was digested with HindIII enzyme overnight at 37°C and recovered using AxyPrep DNA Gel Extraction Kit (Axygen; Corning, Inc.) according to the manufacturer's instructions. N-thy-ori-3-1 cells (3 × 10⁵) were pretreated with SCR7 (50 µM) or RI-1 (30 µM) for 1 h. The cells were transfected with 0.25 µg of pCherry and 1 µg of the digested NHEJ-GFP plasmid and mixed with 2.5 µL of Lipofectamine 2000. Following 6 h, the culture medium of the transfected cells was replaced with medium containing SCR7 (50 µM) or RI-1 (30 µM) and further cultured for 2 days. The detailed procedure has been described.⁶⁸ The cells were trypsinized (0.25%) and resuspended in PBS. The cellular fluorescence was measured by flow cytometry analysis (NovoCyte; ACEA Biosciences, Inc.) and the NovoExpress 1.3.0 software (ACEA Biosciences, Inc.).

4.9 HR assay

N-thy-ori-3-1 cells (3 × 10⁵) were pretreated with SCR7 (50 µM) or RI-1 (30 µM) for 1 h. Then, they were transfected with a single copy of a DR-GFP, I-SceI expression plasmid and with a pCherry plasmid used as a transfection efficiency control. Following 6 h, the culture medium of the transfected cells was replaced with medium containing SCR7 (50 µM) or RI-1 (30 µM), after which the cells were further cultured for 2 days and subjected to flow cytometry analysis (NovoCyte; ACEA Biosciences, Inc.) and the NovoExpress 1.3.0 software (ACEA Biosciences, Inc.), and then, the GFP-positive cell population was measured. The mean values were obtained from three independent experiments. Little variation was observed among the three independent experiments.⁶⁸

4.10 Real-time PCR assay

Total cellular RNA was extracted by TRIzol and the reverse transcription procedure was carried out according to the product instructions. The primers were designed and synthesized by Tingke Biotech. β-Actin was used for normalization as a control.

4.11 Animal experiments

BALB/c-nu mice (4 weeks old) were purchased from SPF (Beijing) Biotechnology Co., Ltd., and experiment was conducted after 1 week of stabilization. TPC-1 (siRET/PTC1) and N-thy-ori-3-1 (OE RET/PTC1) cells were digested down separately with trypsin, washed twice with PBS, and 0.2 mL cell suspension was taken at a concentration of 3 × 10⁷/mL and injected into the right subcutaneous side of nude mice. The animals were housed in ventilated microisolators with free access to sterile food and water. Twenty-one days after injection, BALB/c-nu mice were sacrificed, and tumors were observed, collected, weighted, and measured.

4.12 Cell proliferation assay

For CCK-8 assay, cells were diluted with complete medium to a concentration of 15,000/mL and added to 96-well plates for culture at 100 µL per well. Absorbance values were measured at 12/24/48/72 h after cell inoculation. For the colony formation assay, cells were diluted with complete medium to a concentration of 1000/mL and added to six-well plates for culture. After 10 days of culture, the cell colonies were fixed with 4% paraformaldehyde (Beyotime, P0099) and stained with Giemsa (Beyotime, C0133).

4.13 Cell cycle synchronization using thymidine double block

The cells were initially cultured to approximately 50%‒60% confluence in complete growth medium. The first block was achieved by incubating cells with 2 mM thymidine for 16 h. Following this, cells were released from the block by washing three times with PBS and incubating them in fresh thymidine-free medium for 8−10 h to allow cell cycle progression. A second thymidine block was then applied by re-treating the cells with 2 mM thymidine for an additional 16 h. After the second block, cells were washed thoroughly with PBS to remove residual thymidine and incubated in fresh medium. Synchronization efficiency was assessed by flow cytometry.

4.14 Statistical analysis

Statistical analyses were conducted with GraphPad Prism (version 8.0) software. The data were derived from at least three independent experiments and are presented as mean ± SEM. p < 0.05 was considered statistically significant.