Fifty pairs of papillary thyroid cancer samples consisting of 50 PTCs and 50 matched normal thyroid tissues (1 cm away from the edge of the tumor tissue) were obtained between 2012 and 2015 from patients who underwent thyroid cancer surgery at the Shanghai Cancer Center in Fudan University. Tissue specimens were frozen in liquid nitrogen immediately after surgical resection and stored at −80°C. Final histological classification was based on paraffin-embedded sections. The use of clinical materials for research purposes in the present study was approved by the Institutional Research Ethics Committee of Fudan University prior to the obtained written consent of patients.

The KMT5A gene expression and clinical data from 455 papillary thyroid carcinomas in the TCGA database are available from the website www.cbioportal.org/. The relationship between KMT5A and the clinical parameters was analyzed using the TCGA data (Table I).

Human thyroid follicular epithelial cell line Nthy-ori 3-1 and human papillary thyroid cancer cell lines K1 and TPC-1 were grown in RPMI-1640 media (Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Invitrogen, Carlsbad, CA, USA) and 1% penicillin-streptomycin (10,000 U/ml penicillin and 10 mg/ml streptomycin in 0.9% NaCl; Sigma-Aldrich, St. Louis, MO, USA) at 37°C with 5% CO₂. We constructed a lentiviral system to knock down KMT5A: HEK-293T cells were co-transfected with pLKO.1-KMT5A 1 and 2 (Biotend, Shanghai, China) using Lipofectamine 2000 (Invitrogen, Shanghai, China). Forty-eight hours after transfection, the virus-containing medium collected from HEK-293T cells was used to infect PTC cells K1 and TPC-1 to establish KMT5A-shRNA cells, followed by 2 µg/ml puromycin selection (Sigma-Aldrich). Human thyroid follicular epithelial cell line Nthy-ori 3-1 was obtained from Sigma-Aldrich. Human papillary thyroid cancer cell lines K1 and TPC-1 were purchased from the University of Colorado Cancer Center Cell Bank.

Total RNA of tissue or cell lines was extracted using TRIzol reagent (Invitrogen). Total RNA (1 µg) was used for cDNA synthesis employing PrimeScript™ RT Reagent kit (Takara Bio, Inc., Shiga, Japan). RT-PCR was performed in triplicate with SYBR Premix Ex Taq™ kit (Takara Bio) according to the manufacturer's instructions. The primers for the target genes were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) (Table II). The thermal cycling parameters were: 95°C for 30 sec, 95°C for 5 sec and 60°C for 30 sec. β-actin was used as an internal control. The threshold cycle (Ct) values were analyzed using the comparative Ct (2^−ΔΔCt) method. The level of target genes was calculated by normalizing to the endogenous reference and relative to a control (16).

Immunohistochemical staining (IHC) was carried out according to the manufacturer's protocol. Briefly, formalin-fixed and paraffin-embedded tissue sections were deparaffinized in xylene and hydrated through descending concentrations of ethanol before being placed in a blocking solution to inhibit endogenous peroxidase activity. The slides were incubated with primary antibodies (rabbit anti-human KMT5A; dilution, 1:200; Cell Signaling Technology, Shanghai, China) at 4°C overnight. HRP-conjugated rabbit secondary antibody (dilution, 1:100; Cell Signaling Technology) was added for 60 min at room temperature, followed by DAB development (DAB Substrate Chromogen system; Dako, Agilent Technologies, Shanghai, China) following standard staining protocol. The slides were fixed and images obtained with the Olympus IX71 inverted microscope using the DP2-BSW Olympus image acquisition software system. The results were confirmed by two experienced pathologists who were blinded to the clinicopathological data of the patients. The staining results were scored based on the percentage of stained tumor cell nuclei (0, no staining; 1, ≤10%; 2, >10–50%; and 3, >50%).

The cells were lysed in RIPA buffer and boiled at 100°C for 10 min to obtain the protein lysates. Approximately 20 µg protein lysates was extracted from each sample and separated by 10% SDS-PAGE. After being blocked in 5% non-fat milk at room temperature for 1 h, the protein concerned was probed using the primary antibodies: KMT5A (dilution, 1:1,000), p53 (dilution, 1:1,000; Santa Cruz Biotechnology, Shanghai, China), β-actin (dilution, 1:1,000; Cell Signaling Technology) at 4°C overnight, and then incubated with goat anti-rabbit IgG or goat anti-mouse IgG (dilution, 1:5,000; Cell Signaling Technology) at room temperature for 1 h and detected with enhanced chemiluminescence reagents (Thermo Fisher Scientific, Shanghai, China). The bands were visualized using 1-step™ NBT/BCIP reagents (Thermo Fisher Scientific, Rockford, IL, USA) and detected by the AlphaImager (Alpha Innotech, San Leandro, CA, USA).

Cell viability was determined using Cell Counting Kit-8 (CCK-8) assay according to the manufacturer's protocol. Briefly, the cells were seeded into 96-well plates at 4×10³ cells/well. An aliquot of 10 µl CCK-8 solution was added to each well and the plate was incubated for 3 h at 37°C. At the indicated time-points, the absorbance at 450 nm was assessed using a spectrophotometer. For each group, data from five wells were pooled. Each experiment was performed in triplicate.

The apoptotic cells were assayed by flow cytometric analysis using the Annexin V-FITC/propidium iodide (PI) staining kit (BD Biosciences, Shanghai, China). The suspended cells were stained with 5 µl Annexin V-FITC and 10 µl PI at room temperature for 15 min. Flow cytometry was then used to immediately analyze the cells. The apoptotic cells were marked based on the Annexin V expression (Annexin V⁺/PI⁻ and Annexin V⁺/PI⁺), quantified and compared to the controls from each group. The experiment was performed in triplicate.

The migration of cancer cells was assayed using 6.5-mm diameter chambers with 8-µm pore filters (Transwell, 24-well cell culture; BD Biosciences), while the invasion of cancer cells was assayed using chambers precoated with 20 µg Matrigel (BD Biosciences). Both cell lines were suspended at 3×10⁴ cells/ml in serum-free media and then 0.1 ml cell suspension was added to the upper chamber. Subsequently 0.6 ml complete medium was added to the lower chamber. The chambers were incubated for 24 h at 37°C with 5% CO₂. After incubation, the filters were fixed and stained. The upper surface of the filters was scraped with cotton swabs to remove non-migrating cells. The experiments were repeated in triplicate wells and the number of migrating cells in five high-power fields per filter was counted microscopically at ×200 magnification.

In order to synchronize the cell cycle, the cells were harvested by trypsinization, washed in ice-cold phosphate-buffered saline (PBS) and fixed in 70% ice-cold ethanol in PBS. After being resuspended in cold PBS, RNAase (2 µg/ml; BD Biosciences) was added and the cells were incubated at 37°C for 30 min, followed by incubation in 400 µl PI (10 µg/ml; BD Biosciences) for 40 min at room temperature. The DNA content was analyzed by flow cytometry and the cell population in the G0, G1, S, G2 and M phase was analyzed. All experiments were performed in triplicate.

Malondialdehyde (MDA) in the cells was determined using a lipid peroxidation MDA kit purchased from Beyotime Institute of Biotechnology (Shanghai, China) according to the manufacturer's instructions. Lysates from 2×10⁶ cultured cells were collected with a mixture of ProteoJET Mammalian Cell Lysis Reagent (Fermentas, Shanghai, China). Afterwards the cells were homogenized in lysis solution on ice and then centrifuged at 12,000 × g for 10 min. TBA reagent was added into each well containing 200 µl standard and 200 µl sample and incubated at 95°C for 60 min. Subsequently the mix was cooled at room temperature in an ice bath for 10 min. The activity of SOD was determined using a spectrophotometer at Ex/Em = 532/553 nm for fluorometric assay.

The reactive oxygen species (ROS) level in the K1 and TPC-1 cells after the knockdown of KMT5A was assessed by flow cytometry using Reactive Oxygen Species Assay kit (Beyotime Institute of Biotechnology) following the manufacturer's instructions. After being stained as described in the kit, ROS generation of cells was determined by flow cytometry (FC500; Beckman Coulter, Inc., Brea, CA, USA).

Statistical analyses were performed using GraphPad Prism 5.1 (GraphPad Software, Inc., La Jolla, CA, USA). Student's t-test and one-way ANOVA were performed to analyze the data. P<0.05 was considered to indicate a statistically significant difference. The data are expressed as mean ± SEM. All the experiments were performed at least three times.

To evaluate KMT5A expression in PTC tissue, qRT-PCR and western blotting, as well as IHC were performed in PTC and matched normal thyroid tissue specimens. The results revealed that KMT5A was transcriptionally upregulated in PTC samples (P<0.05; Fig. 1A). Western blotting (Fig. 1B) and IHC (P<0.001; Fig. 1C and D) analysis confirmed elevated KMT5A protein expression in PTC and low KMT5A protein expression in normal thyroid tissue specimens. To investigate whether KMT5A expression in PTC cell lines recapitulated its expression patterns observed in PTC patient tissues, human normal thyroid epithelial NTHY-ori 3-1 cells and PTC K1 and TPC-1 cells were tested. KMT5A was found to be overexpressed in K1 (P<0.05) and TPC-1 cells (P<0.05) compared to NTHY-ori 3-1 cells, both transcriptionally (Fig. 1E) and at protein level (Fig. 1F), which was consistent with the results in the PTC tissue samples.

To illuminate the efficacy of KMT5A on tumor cell proliferation and apoptosis, the expression of KMT5A was knocked down using targeted lentiviruses in K1 and TPC-1 cell lines. Transcriptional and protein knockdown was achieved using two separate lentiviral construct shRNA1 and shRNA2 as compared with shRNA-control-infected cells in both K1 (shRNA1, P<0.05; shRNA2, P<0.05) and TPC-1 (shRNA1, P<0.05; shRNA2, P<0.01) cells (Fig. 2A and B). Targeted knockdown of KMT5A resulted in attenuated proliferation in both K1 (shRNA1, P<0.001; shRNA2, P<0.001; Fig. 2C) and TPC-1 (shRNA1, P<0.001; shRNA2, P<0.001; Fig. 2D) cells after 48 h. Moreover, a significant increase in cell apoptosis measured by flow cytometry of PI/Annexin V-stained cells was observed in both K1 (shRNA1, P<0.001; shRNA2, P<0.001; Fig. 2E) and TPC-1 (shRNA1, P<0.05; shRNA2, P<0.01; Fig. 2F) cells using both shRNA1 and shRNA2 constructs.

Next we evaluated the protumorigenic role of KMT5A in the migration and invasion of PTC cells. The results revealed that both TPC-1 (shRNA1, P<0.01; shRNA2, P<0.05) and K1 (shRNA1, P<0.01; shRNA2, P<0.01) cells transfected with KMT5A shRNA1 or shRNA2 displayed significantly decreased cell migration capacity than that of the shRNA-control vector (Fig. 3A and B). As shown in Fig. 3C and D, the invasion capacity was examined by Transwell assay and the invasive rate was markedly decreased in the KMT5A shRNA1- or shRNA2-transfected TPC-1 (shRNA1, P<0.05; shRNA2, P<0.01) and K1 cells (shRNA1, P<0.05; shRNA2, P<0.05) than in the shRNA-control transfected cells.

Several lines of evidence have revealed that KMT5A exerts crucial biological functions in modulating cell cycle and DNA damage response (17). We further verified the effect of KMT5A on the cell cycle of PTC cells by flow cytometry assay. As shown in Fig. 4, both K1 (Fig. 4A and B) (shRNA1, P<0.05; shRNA2, P<0.05) and TPC-1 (Fig. 4C and D) (shRNA1, P<0.05; shRNA2, P<0.05) cells transfected with KMT5A shRNA1 or shRNA2 had a significantly lower percentage of cells in the G1/S phase compared with the control group. The data revealed that the downregulation of KMT5A resulted in the cell cycle arrest of the PTC cells in the G1/S phase.

As a lysine methyltransferase enzyme, KMT5A regulates posttranslational modification of p53 at lysine 382 and cell cycle arrest functions of p53 (18). Our results revealed that KMT5A silencing led to the downregulation of p53 in the K1 and TPC-1 cells (Fig. 5A), which is consistant with previous studies. Recently, p53 has been reported to regulate metabolic pathways of breast cancer via sterol-regulatory element-binding factor (SREBP) transcription factors, a key modulator gene of the lipid/cholesterol metabolism (19–22). To investigate the impact of KMT5A silencing on the lipid metabolism, we examined the expression of SREBP1 in KMT5A knockdown K1 and TPC-1 cells. As shown in Fig. 5B and C, the inhibition of KMT5A reduced SREBP1 in K1 and TPC-1 when compared with control transfected cells. Several lipid metabolism-related genes are controlled by SREBP1, such as stearoyl-CoA desaturase (SCD), fatty acid synthase (FASN) and acetyl-CoA carboxylase (ACC) (23). Subsequently, we revealed that KMT5A inhibition also decreased SCD, FASN and ACC transcriptional expression in the PTC cell lines (Fig. 5B and C).

Lipid peroxidation is the degradation of lipids that occurs as a result of oxidative damage. Polyunsaturated lipids are susceptible to oxidative attack, typically by ROS, resulting in a well-defined chain reaction with the production of end products such as MDA and ROS level. To further identify the effect of KMT5A on lipid metabolism, we assessed the lipid peroxidation level using MDA assay kit and the ROS generation level using Reactive Oxygen Species assay kit. We demonstrated that silencing of KMT5A markedly decreased both the level of MDA and ROS in the K1 and TPC-1 cells compared with the shRNA-control-transfected cells (Fig. 5D-F).

To investigate the relationship between KMT5A and prognosis of PTC patients, we analyzed the KMT5A expression and clinical data in the TCGA database. Our results revealed that elevated KMT5A was significantly correlated with extrathyroidal extension (P=0.0001), lymph node metastasis (P=0.0057) and advanced pathological stage (P=0.0001) in PTC tissues. However, KMT5A expression had no correlation with other clinicopathological paremeters, such as age (P=0.769), sex (P=0.259) or tumor size (P=0.29625) (Table I). These data reveal that high KMT5A expression is associated with poor outcome in PTC patients.