Primers and probes, TRIzol^® reagent, SuperScript III Reverse Transcriptase, SYBR qPCR mix kit was purchased from Invitrogen; Thermo Fisher Scientific, Inc. DMEM was purchased from Gibco (Thermo Fisher Scientific, Inc.).

Thyroid cancer tissues together with the adjacent non-tumor tissue were surgically removed from 8 patients who were admitted to Baoshan District Integrated Traditional Chinese and Western Medicine Hospital (Shanghai, China) between December 2016 and November 2017. The distance between adjacent non-tumor tissue and the boundary of the cancer tissue was ~1 cm. All patients were diagnosed with papillary thyroid cancer and their clinicopathological characteristics were recorded (Table I) (22,23). After the study was explained, all patients provided written informed consent. The present study was approved by the Institutional Research Board of Baoshan District Integrated Traditional Chinese and Western Medicine Hospital (Shanghai, China).

Tumor tissues and adjacent non-tumor tissues were embedded in paraffin and cut into slides (5 µm thick). Slides were fixed in 4% paraformaldehyde for 10 min at room temperature, blocked with 5% bovine serum albumin (Bio-west, Inc.) at room temperature for 30 min and incubated with primary antibodies against KLF4 (1:500; cat. no. Bs-1064R; Bioss Antibodies, Inc.) at 4°C overnight. Following overnight incubation, slides were washed with PBS and incubated in the dark with FITC-conjugated goat anti-rabbit secondary antibodies (1:1,000; cat. no. ab6717; Abcam) at room temperature for 1 h. Slides were subsequently washed with PBS (3 times; each, 30 sec). Slides were prepared with mounting media and Antifades (Invitrogen; Thermo Fisher Scientific, Inc.) and observed using a fluorescence microscope (magnification, ×100). The mean intensity, calculated by multiplying the area (size) and average density of fluorescence, was evaluated using Image-Pro Plus software (version 7; Media Cybernetics, Inc.).

The protein expression of KLF4 in tumor tissues and adjacent non-cancerous tissues was detected via western blot analysis. In addition, the protein expression of N-cadherin, MMP2, MMP9 and collagen in KTC1 cells were detected via western blotting. Tissues were digested and lysed in lysis buffer (cat. no. P0013; Beyotime Institute of Biotechnology) at 4°C with inhibitors of phosphatase and protease (cat. no. P1045; Beyotime Institute of Biotechnology). The lysis mixture was centrifuged at 4°C for 10 min at 10,000 × g and the supernatant containing cellular proteins was utilized in the following experiments. Protein concentration was determined using a BCA kit. Proteins were separated via SDS-PAGE (10% gel; 40 µg loaded per lane; 120 V). Separated proteins were then transferred to PVDF membranes (100 V for 120 min; Beyotime Institute of Biotechnology), which were subsequently blocked with 5% non-fat milk at room temperature for 1 h. Membranes were then incubated with the following primary antibodies obtained from Abcam at 4°C overnight: Anti-KLF4, anti-N-cadherin (cat. no. ab18203), anti-MMP2 (cat. no. ab97779), anti-MMP9 (cat. no. ab228402), anti-collagen (cat. no. ab138492), anti-GAPDH (cat. no. ab181602) and anti-β-actin (cat. no. ab8227; all 1:1,000). Membranes were washed with Tris-buffered saline containing Tween 20 and incubated with horseradish peroxidase-conjugated goat anti-rabbit secondary antibodies (cat. no. ab6721; 1:2,000; Abcam) at room temperature for 1 h. Membranes were incubated in enhanced chemiluminescence solution (Beyotime Institute of Biotechnology). Images were captured on film (Beyotime Institute of Biotechnology) in a dark room. Experiments were repeated three times. The western blot images were quantified in greyscale using ImageJ software (version 1.5.2; National Institutes of Health).

cDNA sequence of KLF4 (NM_004235) was synthesized and subcloned into lentivirus vector pL6.3-CMV-GFPa1-IRES-MCS (Novobio) for lentivirus production. The KLF4 recombinant lentivirus vector, pL6.3-CMV-GFPa1-IRES-KLF4, was confirmed by Sanger sequencing (24).

Packaging mix (9 µg; Novobio) and KLF4 recombinant lentiviral plasmids (3 µg) were added into Opti-Minimum Essential Medium (Opti-MEM; Thermo Fisher Scientific Inc.) and mixed. Lipofectamine™ 2000 (36 µl; Thermo Fisher Scientific Inc.) was mixed with Opti-MEM (1.5 ml) and incubated at room temperature for 5 min. The plasmid solution and diluted Lipofectamine 2000 were then mixed and incubated at room temperature for 5 min. The mixture was added into a culture dish with 293 T cells (Novobio), and cells were cultured for 48 h. Cell supernatants were then collected, centrifuged at 1,500 × g for 10 min at room temperature and filtered. The lentivirus solution was then condensed via centrifugation at 50,000 × g for 2 h at 4°C and re-suspended in DMEM. KLF4 recombinant lentivirus was derived.

The human papillary thyroid carcinoma cell line KTC1 (3×10⁵ cells/well in six-well plates) was transfected with the pL6.3-CMV-GFPa1-IRES-KLF4 and pL6.3-CMV-GFPa1-IRES-MCS (control). The transduction MOI was 30. Quantitative PCR (qPCR) was utilized to detect the efficiency of KLF4 overexpression after 48 h. There was no transfection in blank group.

The expression of E-cadherin, N-cadherin, Vimentin, MMP2, MMP9 and collagen in KTC1 cells was detected via qPCR. Total RNA was extracted using TRIzol^® reagent, according to the manufacturer's protocol. A universal cDNA synthesis kit (Invitrogen Thermo Fisher Scientific, Inc.) was utilized for reverse transcription at 42°C for 1 h. Each reaction contained 0.5 µl random hexamers primers (dN6; 0.2 µg/µl; Novobio) and 1 µl SuperScript III reverse transcriptase (200 U/µl). The specific primers used are listed in Table II. PCR was performed using a SYBR qPCR mix kit (Invitrogen; Thermo Fisher Scientific, Inc.). The PCR conditions were as follows: Pre-denaturation at 95°C for 2 min; 40 cycles of denaturation at 95°C for 10 sec annealing at 60°C for 30 sec and polymerization at 70°C for 45 sec. qPCR was performed using a CFX96 Touch™ Real-Time PCR Detection system (Bio-Rad Laboratories, Inc.). Gene expression was determined and normalized to β-actin. The primer used for rat β-actin was as follows: Forward, 5′-AGGGAAATCGTGCGTGAC-3′ and reverse, 5′-CGCTCATTGCCGATAGTG-3′. The 2^−∆∆Cq method was utilized to measure PCR results (25).

The viability of KTC1 cells was measured using a cell counting kit-8 (CCK-8; Dojindo Molecular Technologies, Inc.) cell viability assay at 24, 48 and 72 h after KTC1 cells were transfected with the aforementioned viruses. CCK-8 solution was added to each well and incubated at 37°C for 4 h. The absorbance was subsequently measured using a microplate reader at 490 nm. Relative tumor cell viability rate was calculated by dividing the reading of each group at 24, 48 and 72 h by the baseline reading at 0 h. Experiments were repeated three times.

The membrane of the upper compartment was coated with Matrigel (1 g/l; 50 µl), which was allowed to solidify via incubation at 37°C for 1 h. KTC1 cell suspension (1×10⁴ cells/ml, 200 µl) in 2% DMEM was added to the upper compartment of each Transwell insert, while 800 µl DMEM with 10% FBS (HyClone; GE Healthcare Life Sciences) was added to the lower compartment. Cells were incubated at 37°C for 24 h. Subsequently, 4% paraformaldehyde was utilized to fix cells on the microporous membrane at room temperature for 30 min. Cells on the lower side of the membrane were stained with 1% crystal violet at room temperature for 10 min and washed with PBS twice. Cells were then observed under Olympus IX50 fluorescent microscope (magnification, ×400; Olympus Corporation) and the number of cells that had transgressed through the membrane was counted. Relative tumor cell invasion was calculated by dividing the average number of cells that invaded through the membrane in the experimental groups by that in the blank group. Experiments were repeated three times.

A confluent monolayer of KTC1 cells was used in the scratch migration assay. A marker pen was used to draw a straight line at the back of plate. Pippet tips were utilized to draw scratch lines vertical to the straight line on the second day. PBS was used to wash cells three times and DMEM without serum was added. Images were taken under Olympus IX50 (magnification, ×400) and cultured in an incubator with 5% CO₂ at 37°C. Images were then taken at 24 h and migration distances were calculated under the same field using Image-Pro Plus software. Migration distance at 24 h relative to 0 h was recorded for both negative control and overexpression groups. Then relative migration distances were normalized to the negative control group.

Experiments were repeated three times. Statistical data was analyzed using GraphPad Prism software (version 5.0; GraphPad Software Inc.). The results are presented as the mean ± standard deviation. Differences among more than three groups were compared by a one-way analysis of variance followed by the Bonferroni post-hoc test. Differences between two groups were compared using Student's unpaired t-test. P<0.05 was considered to indicate a statistically significant difference.

The expression of KLF4 in thyroid tumor tissue and paracarcinoma tissue was detected by IHC. In the microscopic image of KLF4 staining (Fig. 1), positive cell nucleus was stained by diaminobenzidine (DAB) and appeared as brown. Negative cell nucleus was stained by hematoxylin and appeared as blue. Compared with adjacent non-cancerous tissue, the protein expression of KLF4 was significantly lower in thyroid tumor tissue (P<0.001; Fig. 1). In addition, the protein expression of KLF4 in tumor and adjacent non-cancerous tissue was detected via western blot analysis. The protein expression of KLF4 was markedly lower in thyroid tumor tissue when compared with adjacent non-cancerous tissues (P<0.05; Fig. 2).

KTC1 cells were transfected with viruses carrying KLF4 overexpression vectors. The relative expression of KLF4 was detected via qPCR. The expression of KLF4 in KTC1 cells was significantly increased compared with the blank or negative control groups (P<0.001; Fig. 3).

The mRNA levels of E-cadherin, N-cadherin, Vimentin, MMP2, MMP9 and collagen in KTC1 cells were detected via qPCR. The protein expression of N-cadherin, MMP2, MMP9 and collagen in KTC1 cells were confirmed by western blotting. Among all the genes screened, the mRNA expression of N-cadherin, MMP2, MMP9 and collagen were significantly decreased in the KLF4 overexpression group when compared with the blank or negative control groups (P<0.05 for N-cadherin and MMP2; P<0.01 for MMP9; P<0.001 for collagen; Fig. 4). No significant differences were observed in the levels of E-cadherin and Vimentin (Fig. 4). In addition, the protein expression of N-cadherin, MMP2, MMP9 and collagen were significantly decreased in the KLF4 overexpression group when compared with blank or negative control group (P<0.01 for N-cadherin; P<0.001 for MMP-2, MMP-9 and collagen; Fig. 5).

The viability of KTC1 cells was detected via a CCK-8 assay at 24, 48 and 72 h. The viability of KTC1 cells was markedly decreased in the KLF4 overexpression group at 24, 48 and 72 h when compared with the blank or negative control group (P<0.01 for 24 h; P<0.001 for 48 and 72 h; Fig. 6).

Cell invasion was investigated using a transwell invasion assay. Compared with the negative control group, the invasion of KTC1 cells in the KLF4 overexpression group were significantly decreased (P<0.01; Fig. 7).

Cell migration ability was examined via a scratch migration assay at 0 and 24 h. Compared with the negative control group, the migration of KTC1 cells in the KLF4 overexpression group was significantly decreased at 24 h (P<0.01; Fig. 8).