The high expression of MTH1 and NUDT5 predict a poor survival and are associated with malignancy of esophageal squamous cell carcinoma

Quantification of the MTH1 and NUDT5 expression by immunohistochemistry

Two commercial tissue microarrays (TMAs) consisting of 99 ESCC patients were purchased from Shanghai Outdo Biotech (ODCT; Shanghai, China). TMAs consisting of 81 paired tumor and adjacent normal specimens and 18 ESCC tumor tissues singly. The experience was approved by the Ethics Review committee of Shanghai Outdo Biotech (ODCT; Shanghai, China). The ethical permit number was YB M-05-02. We obtained clinical data of 94 patients who possessed complete overall survival information. The ESCC diagnosis was confirmed by a post-operative pathological examination. The patients involved in the study had undergone surgery from January 2006 to October 2008, follow-up had been conducted until July 2015. Median survival time of these patients was 16 months (range: 2–107 months).

Immunohistochemistry experiments were performed as previously described (Li et al., 2017). Antibodies against MTH1(ab187531) and NUDT5(ab129172) were purchased from Abcam (Cambridge, MA, USA). Normal goat serum (5%) was incubated with the negative control section as a negative control. Visualization was carried out with a Nikon Eclipse 80i microscope (Nikon, Tokyo, Japan).

Immunohistochemistry staining was evaluated by taking into account both the percentage of positive cells and the intensity of staining (Li et al., 2017). On the basis of the percentage of immunoreactive cells, the intensity category was scored from 0 to 4, as follows: 0 (negative), 1 (1–25%), 2 (26–50%), 3 (51–75%) and 4 (76–100%). Staining intensity of the specimen was graded from 0 to 3, as follows: 0 (negative), 1 (slightly stained), 2 (moderately stained) or 3 (heavily stained). The final score was then determined as the proportion of positive cells multiplied by the intensity category. Based on the final score, the expression of MTH1 and NUDT5 was classified as either low immunoreactivity (0–6) or high immunoreactivity (7–12).

Cell lines and cell culture

Six types of human esophageal squamous cell lines were obtained from Beijing Beina Chuanglian Co., Ltd., (BNBIO; Beijing, China), these cell lines were KYSE30, KYSE50, KYSE70, KYSE140, KYSE450, KYSE520 and checked in the NCBI databases. The human normal fibroblast cell line WI-38 and human lung fibroblast IMR-90 cells were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA). KYSE30 and KYSE450 cells were cultured in EMBM media with 20% FBS. WI-38, IMR-90, KYSE50, KYSE70, KYSE140 and KYSE520 cells were cultured in DMEM media with 10% FBS. All cell lines were cultured in a humidified incubator set to 5% CO2.

RNA interference and transfection

The nontargeting control short hairpin RNA CON077 (shRNA) (5′-TTCTCCGAACGTGT CACGT-3′), MTH1 shRNA sequences (5′-CGACGACAGCTACTGGTTT-3′) and NUDT5 shRNA sequences (5′-ATGGATCCTACTGGTAAA-3′) were inserted into the lentiviral vector GV248-puro cut by AgeI/EcoRI sites. Lentivirus LV-MTH1-RNAi, LV-NUDT5-RNAi or LV-CON077-RNAi containing MTH1, NUDT5 or control shRNA, respectively, was then produced by transient transfection in 293T cells.

For transfection, KYSE50 or KYSE70 cells were seeded into 24-well plates at 60–70% confluence. After 24 h, the cells were cultured with virus particles LV-MTH1-RNAi, LV-NUDT5-RNAi or LV-CON077-RNAi in the presence of Polybrene^® (4 μg/mL) for 24–48 h. We then used puromycin (one µg/ml final concentration) to select positive clones after transfection. The stable colonies were amplified after 14 days’ selection on puromycin.

Western blotting

Western blotting experiments were performed as previously described (Li et al., 2017). Commercial antibodies used in this study were as follows: antibodies against MTH1 and NUDT5 were purchased from Abcam (Cambridge, MA, USA); mouse anti-E-cadherin, anti-N-cadherin and anti-Vimentin antibodies were from BD Biosciences (San Jose, CA, USA); rabbit anti-AKT, anti-p-AKT, anti-MEK, anti-p-MEK, anti-ERK and anti-p-ERK were from CST (Danvers, MA, USA); anti-MMP-7, anti-p16, anti-p21 and anti-p27 were from ABclonal (Boston, MA, USA); and anti-Tubulin mouse antibody was from ZSGB-BIO (Beijing, China). The experiment was repeated three times.

The cell counting kit-8 proliferation assay

KYSE cells were plated at 3 × 10³ cells per well into 96-well plates and then incubated at 37 °C in a CO₂ incubator for 0, 24, 48 and 72 h sequentially. The increase in cells was calculated by the CCK-8 assay according to the manufacturer’s protocol. Subsequently, the reagent (10 μl) (Beyotime, Jiangsu, China) was added to particular well. The cells were then incubated at 37 °C for 1 h in the humidified incubator. Optical density (450 nm) was monitored. Three wells were assessed per group, with the results expressed as the means ± standard deviation (SD). Each experiment was repeated three times.

Cell cycle assay

Cells were seeded in 6-well plates and then maintained with 5% CO₂ for 36–48 h at 37 °C. Harvested the cells and subjected to washing with phosphate-buffered saline (PBS) and then fixed at 4 °C overnight with 70% ethanol. Removed the ethanol and suspended the cells with PBS. The cells were finally resuspended in a solution of PI (50 μl/ml) and RNase1 (250 μg/ml) for 30 min in a dark environment at room temperature. The cell cycle was measured using a flow cytometer at 488 nm. The cell-cycle phase was detected with the BD FACSD via a software program to analyze the percentages of cells in G0/G1, S and G2/M phases based on the DNA content.

Cell migration and invasion assay

Cell migration was performed using 6.5-mm transwell permeable supports with 8.0-µm pore polycarbonate membranes inserts. Cell invasion was performed with 6.5-mm transwell inserts with 8.0-µm pore polycarbonate membranes which were pre-coated with Matrigel (Costar, Corning Incorporated, NY, USA) for invasion. The cells were suspended with serum-free DMEM, and 8 × 10⁴ cells were seeded into the upper chambers in 200 μl of DMEM with 1% FBS, while DMEM containing 20% FBS (700 μl) was added to the lower chamber. After culturing for 48 h, the cells on the upper surface of the membrane were removed, and those on the bottom of the membrane were fixed with 10% formalin for 30 min and then stained with 0.1% crystal violet for 4 h. Cells from 10 random microscopic fields/insert were counted in triplicate at ×100 magnification. Calculated the mean number of migrated or invaded cells.

Statistical analyses

All statistical analyses were performed using the SPSS software program (SPSS Inc., Chicago, IL, USA). The means were compared between two groups using Student’s t-test. Pearson’ χ² or Fisher’s exact tests was used to calculate the correlation between MTH1 or NUDT5 and clinicopathological features of patients with ESCC. OS was calculated by the Kaplan–Meier analysis and compared between two groups with the log-rank test. Effects of ESCC clinicopathological variables and the expression of MTH1 or NUDT5 protein on the survival and prognosis of ESCC patients were assessed with the Cox proportional hazards regressions model. P values of less than 0.05 were considered to indicate statistical significance.

The protein expression of MTH1 and NUDT5 in human ESCC tissues: the high expression of MTH1 and NUDT5 predicted a poor prognosis

Immunohistochemical analyses of TMAs verified that MTH1 and NUDT5 proteins were highly expressed in tumor tissues (Figs. 1A–1H). The clinical details are listed in Table 1. MTH1 and NUDT5 were poorly detected in adjacent normal tissues, as revealed by immunochemistry (Figs. 1A–1J). Tumor tissues varied in staining intensity (weak or strong) (Figs. 1A–1H). Representative images of immunohistochemical staining of MTH1 and NUDT5 proteins are shown in Figs. 1A–1J. On the basis of intensity category and immunostaining intensity, among the 94 ESCC specimens, 43 exhibited relatively high levels of MTH1 and 58 showed relatively high NUDT5 expression (Figs. 1K and 1L), but no high expression was found in the adjacent tissues.

We then determined the association between the protein expression of MTH1 and NUDT5 and the clinicopathological characteristics of 94 ESCC specimens (Table 1). Significant positive correlations were identified between the expression of MTH1 and the AJCC and T stages (p = 0.040 and p = 0.048, respectively), as well as between the expression of NUDT5 and the AJCC stage and lymph node invasion (p = 0.001 and p = 0.002, respectively) (Pearson’ χ² or Fisher’s exact tests, p < 0.05). No other statistically significant associations between the protein expression and established prognostic factors (N stage, cell differentiation, vascular invasion, gender and age) were observed (Table 1).

The association between the MTH1 and NUDT5 protein expression and the OS rate of ESCC patients was evaluated using a Kaplan–Meier survival analysis with the log-rank test. As shown in Figs. 1M and 1N, the patients with low MTH1 (p = 0.0311) and NUDT5 (p = 0.0014) levels had longer OS values than those with high levels. The results of a univariate Cox regression analysis indicated that the following characteristics were significantly associated with a reduced OS: AJCC stage (p < 0.001), T stage (p = 0.011), N stage (p = 0.002), MTH1 expression (p = 0.014) and NUDT5 expression (p = 0.003) (Tables 2 and 3). Furthermore, a multivariate Cox regression analysis demonstrated that the high expression of NUDT5 independently predicted a low OS in patients with ESCC (hazard ratio (HR) 1.751; 95% confidence interval (CI) [1.056–2.903]; p = 0.030) (Table 3).

The MTH1 and NUDT5 protein expression in human ESCC cell lines: silencing MTH1 and NUDT5 inhibited the proliferation of ESCC cell lines

To assess the expression of MTH1 and NUDT5 as a whole, we detected the levels of these two proteins by performing Western blotting in six human ESCC cell lines (KYSE30, KYSE50, KYSE70, KYSE140, KYSE450 and KYSE520), the human normal diploid fibroblast line WI38 and the normal human fibroblast model of senescence IMR90 (Fig. 2A). The data indicated that the normalized proteins level of MTH1 and NUDT5 in ESCC cell lines were significantly upregulated than those in the control cell lines WI38 and IMR90 (Student’s t-test, p < 0.05) (Figs. 2B and 2C). All of these data showed that proteins level of MTH1 and NUDT5 were significantly upregulated in the ESCC cell lines.

Protein levels in human ESCC tissues and cell lines were significantly upregulated, thus, we further explored the potential effects of the MTH1 and NUDT5 protein on the ESCC cell viability. We first established cell models with MTH1 or NUDT5 depletion using two ESCC cell lines (KYSE50 and KYSE70) transfection with either LV-MTH1-RNAi, LV-NUDT5-RNAi and control CON077-(hU6–MCS–Ubiquitin–EGFP–IRES–puromycin) and verified the changes in the expression by Western blotting. As shown in Figs. 2D and 2E, the expression of MTH1 and NUDT5 was significantly reduced. MTH1 and NUDT5 depletion significantly attenuated the cellular proliferation of KYSE50 and KYSE70 cells (Student’s t-test, p < 0.05) according to the CCK-8 proliferation assay (Figs. 2F and 2G). These data suggested that MTH1 and NUDT5 are both positive regulators of ESCC cell proliferation.

Depletion of MTH1 and NUDT5 induced cell cycle arrest in ESCC cell lines and affected the expression of cycle-related proteins in KYSE50 and KYSE70 cells

We further performed flow cytometry analysis to investigate whether or not the proliferation inhibition was related to the cell cycle arrest. As shown in Figs. 3A–3L, with MTH1 and NUDT5 depletion, the cells in the G0/G1 phase increased significantly and showed a decline in the S and G2/M phases relative to those in the control groups. This finding suggested that depletion of MTH1 and NUDT5 effectively induced G0/G1 cycle arrest in ESCC cell lines. These results showed that the inhibition of proliferation induced by depleting MTH1 and NUDT5 might occur through cell cycle arrest.

To verify the mechanism underlying the involvement of MTH1 and NUDT5 in G0/G1 arrest, we analyzed the expression of cycle-related proteins involved in the regulation of the G1 and S phases by Western blotting. After blocking the expression of MTH1 and NUDT5, the Cyclin D level was decreased (Fig. 3M). Conversely, the levels of p16, p21 and p27 were markedly increased. The changes in cycle-related proteins were more obvious in KYSE50 cells than in KYSE70 cells (Fig. 3M). The results demonstrated that the effects of MTH1 and NUDT5 on G0/G1 phase arrest were related to those proteins specifically targeted in the G1/S transition.

Depletion of MTH1 and NUDT5 inhibited the migration, invasion and EMT of ESCC cells in vitro

To evaluate the effects of MTH1 and NUDT5 on ESCC cell migration and invasion, Transwell assays were performed. The Transwell migration assays without Matrigel suggested that knockdown of MTH1 or NUDT5 inhibited the migration of KYSE50 and KYSE70 cells. The Transwell invasion assays with Matrigel showed consistent results (Student’s t-test, p < 0.05) (Figs. 4A–4P).

Our data showed that MTH1 and NUDT5 are associated with the migration and invasion of ESCC cells. We therefore explored whether or not EMT of ESCC cells was regulated by MTH1 or NUDT5. We found that a cadherin switch occurred due to the knockdown of MTH1 or NUDT5. Specifically, after the knockdown of MTH1 and NUDT5, E-cadherin expression markedly increased, whereas N-cadherin expression markedly decreased (Fig. 4Q), indicating that EMT was inhibited in the MTH1- and NUDT5-deficient cell lines. Consistent with N-cadherin, the expression of Vimentin (another mesenchymal marker) was also markedly reduced. After blocking MTH1 and NUDT5, the MMP7 expression was found to be markedly reduced (Fig. 4Q). These results suggest that silencing MTH1 and NUDT5 inhibited migration and invasion via the inhibition of EMT.

The depletion of MTH1 and NUDT5 suppressed ESCC cell aggressiveness via inactivation of the MAPK/MEK/ERK pathway

We next investigated the mechanisms underlying how MTH1 and NUDT5 are involved in the signaling pathway for ESCC cell aggressiveness. We noted that phosphorylation of core members MEK1/2 and ERK1/2 in MAPK pathway was significantly decreased when MTH1 or NUDT5 was knocked down (Fig. 4R). The results in Fig. 4 suggest that MAPK/MEK/ERK activation might underlie the phenotypes induced in ESCC cells by MTH1 and NUDT5.