Introduction

Ovarian cancer is the most lethal gynecologic malignancy worldwide [1]. The most frequent histological subtype is epithelioid ovarian cancer (EOC), accounting for over 90% of all ovarian malignancies [2]. EOC is hardly diagnosed at an early stage and, therefore, called “silent killer”. Only 15% of ovarian cancer is localized to the ovary, while about 62% occurs as distant disease [3]. Accordingly, even though significant improvements in surgical techniques and neoadjuvant chemotherapy, the 5-year overall survival (OS) for patients with the advanced disease is far from satisfactory [4]. Post-operative recurrence and drug resistance are primary causes [3]. Therefore, understanding the molecular mechanisms involved in EOC development and identification of novel therapeutic targets are urgently required.

Epithelial to mesenchymal transition (EMT) and cell proliferation are known to be modulated by PI3k/Akt/mTOR and Ras/MAPK pathways. Interestingly, both the two signaling pathways converge on mitogen-activated protein kinase (MAPK)-interacting kinase 1/2 (MNK1/2)–eukaryotic initiation factor 4E (EIF4E) axis [5]. EIF4E plays key roles in the control of mRNA translation, which can be phosphorylated and activated by MNKs [6]. The significance of EIF4E phosphorylation in cancer cell proliferation and transformation has been previously reported [7,8,9]. However, the role of its upstream regulator, MNKs, seems to be dispensable for normal cell growth and development [10]. On the other hand, MNK1, the primary isoform of MNKs which can be regulated by ERK and p38, was reported to be up-regulated in glioblastoma [11], lymphoma [12], and breast cancer [9]. However, whether MNK1 play roles in EOC is still unclear.

Since MNK1 is a downstream modulator of both mTOR and MAPK pathways, we speculated that the variable expression of MNK1 may play critical roles in EOC development and chemoresistance. In the current study, we firstly evaluated the levels of MNK1 in EOC tissues, which showed positive correlation with p-EIF4E levels. Also, increased expression of MNK1 indicates poor clinical outcomes of EOC patients, according to Kaplan–Meier survival and Cox regression analyses. Moreover, silencing or inhibition of MNK1 significantly down-regulated the proliferation capacity of OVCAR-5 cells, revealing the potential values of MNK1 inhibition in chemotherapy development for EOC patients.

Methods

Patients and specimens

We enrolled 138 patients who underwent surgical resection of primary EOC between June 2002 and June 2015 in Yidu Central Hospital of Weifang. Paraffin-embedded resected tumor tissues were obtained from the Pathology Department of our Hospital. None of the patients had serous pre-operative complications or chemotherapies before surgery. The clinicopathologic data was obtained retrospectively. The median age of the cohort was 56.5 years, with an age range of 31–85 years. The International Federation of Gynecology and Obstetrics (FIGO) stage, pre-operative CA125 level, histological type, pathological grade, and existence of lymph node metastasis of all the 138 cases were retrieved. This study was approved by The Ethics Committee of Yidu Central Hospital of Weifang. Written informed consent forms were obtained from all patients or their family members who participated in this study. The detailed patient’s characteristics are summarized in Table 1.

Table 1 Correlation between MNK1 level and clinical features of EOC patients
Full size table

Antibodies and reagents

Anti-p-EIF4E (Ser209) antibody was obtained from Abcam. MNK1 (C4C1), EIF4E (C46H6), Cyclin-D1 (92G2), and MMP9 (D6O3H) antibodies were purchased from Cell Signaling. Caspase-3 (C-6), c-Myc (6A10), E-cadherin (G-10), and β-actin (I-19) antibodies were obtained from Santa Cruz Biotechnology. CGP57380 (MNK1 inhibitor) was from Sigma. SiRNAs were from Shanghai Gene Pharma Co., Ltd. (Shanghai, China).

Immunohistochemistry (IHC) and IHC evaluation

We performed immunohistochemical staining for MNK1 and p-EIF4E on 5-μm sections from the specimens. Sections were firstly incubated for 60 min at 65 °C, deparaffinized and rehydrated using a graded ethanol series. Sections were then placed in citrate buffer (pH 6.4) and microwaved for 15 min to retrieve antigen. After cooled to room temperature and washed with PBS for three times, the sections were incubated in 3% H2O2 for 20 min for quenching of endogenous peroxidase, followed by blocked in goat serum in 37 °C for 45 min. Primary antibody incubation was then performed overnight at 4 °C. The second day, sections were incubated with IgG Antibody-Fab-HRP for 30 min. Finally, DAB and Hematoxylin staining were carried out.

The IHC staining results were observed under light microscope (Olympus Corp, Tokyo, Japan). MNK1 was primarily located in nucleus of EOC cells, with slightly staining in cytoplasm. The existence of p-EIF4E was mainly observed in cytoplasm of tumor cells. Semi-quantitative expression levels were evaluated by two independent pathologists based on the average intensity and percentage of positively stained tumor cells. Staining intensity was scored as follows: 0 (negative staining), 1 (lightly yellow), 2 (deep yellow), and 3 (yellow brown). The percentage of stained tumor cells was graded as 0 (no positive cells), 1 (0–25% positive cells), 2 (25–50% positive cells), 3 (50–75% positive cells), and 4 (75–100% positive cells). Ultimate IHC score was calculated by multiplying intensity score with percentage score. IHC score >4 was classified as high expression level, while ≤4 was defined as low expression level.

Cell culture and transfections

Human ovarian carcinoma OVCAR-5 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; GE Healthcare, Logan, UT, USA), and cultured at 37 °C in a 5% CO2 atmosphere.

The construct of human MNK1 plasmid, pWZL-Neo-Myr-Flag-MNK1, was a gift from William Hahn and Jean Zhao (Addgene plasmid # 20536) [13]. For the transfection, approximately 2 × 105 cells were seeded in 6-well plates and cultured in serum-free DMEM to 40% confluence by the time of transfection. Overexpression of MNK1 was conducted by Lipofectamine 3000 reagent (Thermo Fisher Scientific, USA) according to the manufacturer’s instructions. After 4 h of incubation, the cells were cultured in DMEM with 10% FBS and subjected to other assays.

The silencing experiments were also performed using Lipofectamine 3000 reagent. The targeting siRNAs and negative control siRNA were designed as follows:

  • MNK1 siRNA: 5′-UGCCCAUCUCUAUAGGUUUUU-3′ [14, 15]

  • EIF4E siRNA: 5′-UACAUUAAUCGGUAGCAGGAA-3′ [16]

  • Negative control siRNA (nc-siRNA): 5′-AAACCUAAGUAGAGUGGACUU-3′

RNA isolation and real-time PCR

RNA was isolated using an TRIzol Reagent Kit (Invitrogen) according to the manufacturer’s instructions. Aliquots of 1 μg total RNA were then used for quantitative reverse transcription real-time PCR analysis (qRT-PCR) by TaqMan Gene Expression Assays (Life Technologies, CA, USA). GAPDH was used an internal standard.

Specific primers for RNA quantification were designed as follows:

  • MNK1 (F): 5′-GATTCCTCTGAGACTCCAAGTTAA-3′

  • MNK1 (R): 5′-ACGCTTCTTCTTCCTCCTCTT-3′

  • MMP9 (F): 5′-CCTGGGCAGATTCCAAACCT-3′,

  • MMP9 (R): 5′-CAAAGGCGTCGTCAATCACC-3′.

  • c-Myc (F): 5′-GCCACGTCTCCACACATCAG-3′,

  • c-Myc (R): 5′-TCTTGGCAGCAGGATAGTCCT T-3′.

  • Cyclin-D1 (F): 5′-CACGCGCAGACCTTCGTT-3′,

  • Cyclin-D1 (R): 5′-GCGGATTGGAAATGAACTTCA-3′.

  • GAPDH (F): 5′-GGGTGTGAACCATGAGAAGT-3′,

  • GAPDH (R): 5′-GACTGTGGTCATGAGTCCT-3′.

Western blot

Cells were harvested with glycerol lysis buffer, boiled for 10 min, fractionated by SDS-PAGE and transferred to PVDF membranes. Membranes were blocked in 5% BSA in Tris-buffered saline with 1% Tween (TBST) and then incubated with primary antibodies. After washed, blots were incubated with secondary antibodies (Santa Cruz Biotechnology, USA; 1:5000 dilution) for another 45 min and finally developed using chemiluminescence reagents (Santa Cruz biotechnology, USA). Blots were later scanned and semi-quantified with Image J software.

Cell proliferation assays

The Cell Counting Kit-8 (CCK-8) assay kit (Dojindo, Kumamoto, Japan) was used to determine the impact of transfection or CGP57380 on cell proliferation. Briefly, transfected cells were plated in 96-well plates at a density of 5 × 103 cells per well for designated time points. 10 μl CCK-8 solution was added to the cells for 2 h at 37 °C, and the viability of the cells was then measured at 450 nm using a microplate reader according to the manufacturer’s instructions. Each condition was performed in triplicate, and the experiments were repeated for three times.

Statistical analysis

All statistical analyses were performed with SPSS 18.0 software. A Chi square test of cross-tab was used to evaluate the correlations between the expression of MNK1 and patient’s clinicopathologic characteristics. Survival analyses were performed with Kaplan–Meier method and log-rank test. Cox proportional hazard model was applied to evaluate the independent prognostic values of different variables. Student t test was used to analyze the significance of in vitro assays (compared with control group). P < 0.05 was considered as statistically significant.

Results

Expression of MNK1 and p-EIF4E in clinical EOC tissues

MNK1 was primarily detected in nucleus of tumor cells, while p-EIF4E was mainly located in cytoplasm (Fig. 1a–d). As described in “Methods”, high expression of MNK1 was observed in 62 cases (44.9%), and 49 cases (35.5%) showed high p-EIF4E levels. Moreover, as a result from correlation test, we found that the level of MNK1 was positively correlated with p-EIF4E level (Fig. 1e). We also collected 7 pairs of fresh-frozen tumor tissues and adjacent normal tissues to perform the RT-qPCR assay, which revealed that MNK1 was significantly higher in tumor tissues (Fig. 1f).

Fig. 1
figure 1

Expression of MNK1 and p-EIF4E in EOC tissues. a Representative IHC staining of MNK1 low expression in serous EOC. b Representative high expression of MNK1 in nucleus: the percentage of positive cells was scored as 4, and the staining intensity was scored as 3, resulting in the final IHC score with 12. d Representative low p-EIF4E level in EOC tissue. d Representative high p-EIF4E level in cytoplasm: the percentage of positive cells was scored as 4, and the staining intensity was scored as 2, resulting in the final IHC score with 8. e Cross-tab showed the significant correlation between MNK1 and p-EIF4E levels (P = 0.012). f RT-qPCR experiments revealed that the mRNA level of MNK1 was significantly higher in tumor tissues than adjacent normal tissues (P = 0.042). Scale bar 100 μm. Asterisk indicates significant difference with P < 0.05

Full size image

Patients characteristics and their correlations with MNK1 expression levels

For the 138 EOC patients, 55 cases (39.9%) were diagnosed as FIGO stage I–II, and the other 83 cases (60.1%) with stage III–IV. Based on the histological type of EOC, we classified the patients into serous EOC (117 cases, 84.8%) and mucinous EOC (21 cases, 15.2%). As for pathological grades, 42 cases (30.4%) were with grade I, while 59 (42.8%) and 37 cases (26.8%) were with grades II and III, respectively. Fifty-seven patients (41.3%) occurred lymph node metastasis at the time of primary surgery.

Among those, 66 cases (47.8%) were confirmed as cancer-specific death until the end of our follow-up, and the other 72 (52.2%) were censored due to discontinued follow-up or died of reasons other than EOC or survived by the date of last follow-up. The detailed case characteristics are summarized in Table 1.

We performed Chi square test to see whether the MNK1 level has any relationship with patients clinicopathological characteristics. High MNK1 level was identified to be correlated with higher FIGO stage and positive lymph node metastasis, which revealed that MNK1 may play critical roles in EOC development and metastasis.

Univariate and multivariate analyses for prognosis of patients with EOC

Kaplan–Meier survival analysis was carried out to estimate the prognostic values of different variables. As assumed, higher MNK1 level indicates poor OS of EOC patients (66.8 ± 4.8 vs. 90.4 ± 6.1 months, P = 0.037; Table 2; Fig. 2a). In addition, advanced FIGO stage and positive lymph node metastasis also indicated unfavorable clinical outcomes (P = 0.013 and P = 0.001, respectively; Fig. 2b, c). To better evaluate the role of MNK1 in EOC progression, we next analyzed the effect of MNK1 expression in progression-free survival (Fig. 2d), which revealed that high expression of MNK1 may also help predict disease relapse (P < 0.001).

Table 2 Kaplan–Meier survival analysis for EOC patients
Full size table
Fig. 2
figure 2

Kaplan–Meier survival analysis of EOC patients. MNK1 expression level (a), FIGO stage (b), and lymph node metastasis (c) can affect the overall survival of EOC patients. In addition, high expression of MNK1 was also correlated with disease recurrence (d)

Full size image

The three prognosis factors were then subjected to Cox multivariate analysis to figuring out the independent predictive factors. Importantly, levels of MNK1 can act as an independent prognostic factor for the OS of EOC patients (HR = 1.793, 95% CI = 1.064–2.655, P = 0.038; Table 3). FIGO stage and lymph node metastasis also showed statistical significance as independent prognostic factors (P = 0.040 and P = 0.009, respectively).

Table 3 Multivariate analysis of the prognostic factors for EOC
Full size table

MNK1 regulates the mRNA translation of proliferation-related proteins through phosphorylating EIF4E

We performed MNK1-overexpression and siRNA experiments to further investigate mechanisms of MNK1 in regulating EOC development. MNK1 silencing down-regulated the phosphorylation level of EIF4E without changing total-EIF4E level (Fig. 3a). Taking into consideration that EIF4E plays key roles in regulating protein translation, we examined the expression of proliferation-related proteins. The protein expression of Cyclin-D1, c-Myc, and MMP-9 protein all enhanced upon MNK1 overexpression; even though the mRNA levels of those molecules showed no significant changes (Fig. 3b). These results are consistent with the opinions that EIF4E mainly regulates the mRNA translation process instead of gene transcription. In addition, MNK1 also regulates E-cadherin level, indicating that MNK1 may take participate in the cell adhesion and migration. On the other hand, MNK1 showed no significant effects on the protein level of caspase-3, showing that it may not directly modulate cell apoptosis.

Fig. 3
figure 3

MNK1 regulates EOC cell proliferation through phosphorylating EIF4E. a Transfected OVCAR-5 cells were stimulated with EGF (50 ng/mL) for 30 min, and the protein levels were subsequently examined by Western blot. MNK1 can phosphorylate EIF4E without affecting total EIF4E level. MNK1 overexpression up-regulated the protein expressions of proliferation-related proteins, such as Cyclin-D1, c-Myc and MMP-9, whereas EIF4E silencing attenuated those effects. In addition, MNK1 can also negatively regulate the expression of E-cadherin, but showed no effect on Caspase-3. The effect of MNK1 inhibitor (CGP57380) was also tested paralleled. b MNK1 or EIF4E overexpression showed no direct modulation towards the mRNA levels of Cyclin-D1, c-Myc or MMP-9, according to the qRT-PCR results. c MNK1 overexpression significantly promoted OVCAR-5 cell proliferation; on the other hand, MNK1 or EIF4E silencing resulted in decreased cell viability. d Transfected cells were grown in DMEM and treated with or without 10 μmol/l CGP57380 for 5 days; cell density was then measured. CGP57380 showed comparable inhibition effects with MNK1-siRNA, indicating the potential of MNK1 as a novel chemotherapy target. In addition, the migration (e) and invasion (f) capacities of OVCAR-5 cells were both inhibited by either MNK1-siRNA or CGP57380. Asterisks indicate significant difference with P < 0.05

Full size image

Moreover, EIF4E silencing can attenuate, if not abolish, the effects of MNK1 overexpression in modulating changes of those downstream proteins (Fig. 3a). Therefore, we came to the conclusion that MNK1 can mediate the expression of proliferation-related proteins through up-regulating the EIF4E phosphorylation level.

MNK1 silencing or inhibition can down-regulate EOC cell proliferation, migration and invasion

Proliferation of OVCAR-5 cells was tested upon MNK1-silencing or MNK1-overexpression. Consistent with protein expression patterns, MNK1 overexpression significantly promoted the cell proliferation process, while EIF4E-siRNA simultaneously can abolish this effect (Fig. 3c).

The in vitro results showed that MNK1 had critical functions in regulating EOC cell proliferation. Therefore, we used CGP57380, a specific MNK1 inhibitor, to further test the potential of MNK1 in chemotherapy development for EOC patients. Meaningfully, CGP57380 showed comparable effects with MNK1-silencing on inhibiting OVCAR-5 cell proliferation (Fig. 3d). We also performed scratch assay and Tranwell assay to investigate the effects of MNK1 on cell migration and invasion, respectively. Our results showed that both the migration (Fig. 3e) and invasion (Fig. 3f) capacities of OVCAR-5 cells were impaired by either MNK1-siRNA or CGP57380 treatment.

Discussion

Despite transcriptional process used to be considered as the principal control for gene expression, it is now widely approved that modulation of mRNA translation is critically involved in determining the cellular phenotype [17, 18]. Control of mRNA translation allows for rapid and spatially specific expression of proteins, which is particularly important in cancer cells [19].

EIF4E can direct ribosomes to the cap structure of mRNAs, thus functioning as the rate-limiting component of the eukaryotic translation apparatus and is involved in the mRNA-ribosome binding step of eukaryotic protein synthesis [20]. MNK1/2 was reported to phosphorylate EIF4E under the docking of EIF4G and subsequently regulate the binding activity of EIF4E [21, 22]. Moreover, MNK1-EIF4E axis was revealed as common downstream effectors of both PI3k/Akt/mTOR and Ras/MAPK signaling pathways [22,23,24], therefore, drawing considerable attentions in investigating its functions in tumor developing and treatment.

In the current study, we focused on exploring the roles of MNK1 in EOC, the most frequent cause of female cancer mortality in Western Europe and the United States [25]. Accordingly, our clinical study demonstrated that the expression level of MNK1 can serve as an independent prognostic factor for EOC; on that higher MNK1 level indicated poorer clinical outcomes. In addition, the expression level of MNK1 is positively correlated with that of p-EIF4E; however, we did not find any prognostic significance of p-EIF4E in our cohort; even it was reported that high p-EIF4E levels indicated better OS according to Noske’s study [26]. This is really interesting since that more frequent expression of total EIF4E was reported to be observed in advanced-stage EOC specimens [27]. More rigorous and larger sample analysis may be needed to illuminate the exact roles of EIF4E and its modification in EOC.

One reason for poor prognosis of EOC is due to the cancer resistance under consecutive chemotherapy, suggesting that the identification of novel rational targets for new effective therapeutic modalities is required [28]. We, therefore, tested whether MNK1 can be helpful in possible therapeutic strategies. Cell experiments demonstrated that MNK1 silencing can significantly down-regulate the expression of proliferation-related proteins, whereas up-regulate the E-cadherin level, revealing the indispensable roles of MNK1 in ovarian cancer cell proliferation and transformation. Moreover, MNK1 inhibitor, CGP57380, also showed significant inhibition for the EOC cell viability, providing new evidence for the potential of MNK1 as a chemotherapy target.

Conclusion

High MNK1 expression in EOC tissues indicates poor clinical outcomes, and inhibition of MNK1 provides novel directions for future thermotherapy development towards EOC.