MKRN2 inhibits migration and invasion of non-small-cell lung cancer by negatively regulating the PI3K/Akt pathway

Tumor specimens collected from 261 patients with NSCLC from 2013 to 2017 were retrieved from the Pathology Archive of the First Affiliated Hospital of China Medical University (Shenyang, China). All enrolled patients underwent curative surgical resection without having prior chemotherapy or radiation therapy. This study was approved by the Medical Research Ethics Committee of China Medical University, and informed consent was obtained from all patients.

All specimens were fixed in 10% neutral formalin, embedded in paraffin, and prepared as 4-μm-thick serial sections. Immunostaining was performed according to the streptavidin-peroxidase method. The sections were incubated with anti-MKRN2 rabbit polyclonal antibodies (1:100; HPA037559; Sigma-Aldrich, Shanghai, China) at 4 °C overnight. Sections were washed in phosphate-buffered saline (PBS) and incubated with reagents A and B (EliVsion Reagent; KIT9921; MaiXin, Fuzhou, China), according to manufacturer instructions. Sections were then developed using 3,3-diaminobenzidine tetrahydrochloride (MaiXin), lightly counterstained with hematoxylin, dehydrated in alcohol, and mounted. Two investigators blinded to the clinical data semiquantitatively scored all slides by evaluating the staining intensity and percentages of cells stained in representative areas of each slide. The percentages of stained cells were scored as follows: 1 (1–25%), 2 (26–50%), 3 (51–75%), or 4 (76–100%). Based on the staining intensity, MKRN2 expression was also scored as follows: 0 (no staining), 1 (weak staining), 2 (moderate staining), and 3 (high staining). Percentage scores were assigned as follows: 0 (0%), 1 (1–30%), 2 (31–70%), and 3 (71–100%). The scores of each tumor sample were multiplied to give a final score ranging from 0 to 9, with tumor samples scored > 3 considered MKRN2-high, and those scored ≤3 considered MKRN2-low [14].

All cell lines were purchased from the Cell Bank of the China Academy of Sciences (Shanghai, China) and cultured in medium containing 10% qualified fetal calf serum (FB15015; Clark Biosciences, Richmond, VA, USA), 100 IU/mL penicillin (Sigma-Aldrich, St. Louis, MO, USA), and 100 μg/mL streptomycin (Sigma-Aldrich). HBE cells were cultured in Dulbecco’s modified Eagle medium (DMEM) with high glucose; A549, H1299, H460, H226, and H292 cells were cultured in Roswell Park Memorial Institute 1640 medium; and SK-MES-1 cells were cultured in MEM medium. Cells were maintained in a 5% CO₂ incubator at 37°C [27].

Cell transfection was performed using Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to manufacturer instructions. In MKRN2-knockdown experiments, cells were transfected with MKRN2-specific small-interfering (si)RNA and scrambled control siRNA (Ribobio, Guangzhou, China) for 48 h. For MKRN2 overexpression, cells were transfected with an MKRN2-expression plasmid and the corresponding empty EX05 vector (LongqianBiotech, Shanghai, China).

To inhibit PI3K/Akt signaling, cells were treated with 10 μM LY294002 (MedChemExpress, Monmouth Junction, NJ, USA), an inhibitor that blocks the PI3K/Akt pathway. LY294002 was dissolved in dimethyl sulfoxide (DMSO) and added 12 h after a 36-h transfection, with the same volume of DMSO added to control cells.

Quantitative real-time PCR was performed in a 7900HT fast real-time PCR system (Applied Biosystems, Foster City, CA, USA) using SYBR Green PCR master mix in a total volume of 20 μL under the following cycling conditions: 95°C for 30 s and 45 cycles of 95°C for 5 s and 60°C for 30 s. The dissociation step was used to generate a melting curve and confirm amplification specificity. Relative gene expression was calculated using the 2^−ΔΔCt method, with β-actin used as a reference. Primer sequences were as follows: MKRN2 forward, 5′-AAGGCCTCTGCTTCTGAGAG-3′, and reverse, 5′-GATATCACACGGCATTCTGG-3′; and β-actin forward, 5′-ATAGCACAGCCTGGATAGCAACGTAC-3′, and reverse, 5′-CACCTTCTACAATGAGCTGCGTGTG -3′. All experiments were performed in triplicate.

Total cellular protein was extracted using lysis buffer (P0013; Beyotime Biosciences, Shanghai, China) containing a protease-inhibitor cocktail (B14002; Biotool, Shanghai, China) and in the presence or absence of a phosphatase-inhibitor cocktail (B15002; Biotool) according to manufacturer instructions and quantified using the Bradford method. Proteins (80 μg/lane) were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA), and the membranes were blocked in 5% skim milk (232,100; Becton Dickenson, Franklin Lakes, NJ, USA) in Tris-buffered saline with Tween-20 at room temperature, followed by incubation overnight at 4°C with the appropriate primary antibodies (Table 1). Membranes were washed and then incubated with horseradish peroxidase-conjugated anti-mouse/rabbit IgG (1:2000; ZSGB-BIO, Beijing, China) at 37°C for 2 h. Immune reactivity was detected by enhanced chemiluminescence (Thermo Fisher Scientific, Waltham, MA, USA) using a BioImaging system (UVP, Inc., Upland, CA, USA). Relative protein expression was calculated after normalization to glyceraldehyde 3-phosphate dehydrogenase (GAPDH), which was used as a loading control.

Assays were performed as described previously [28].

Cell migration and invasion assays were performed in 24-well Transwell chambers containing inserts with a pore size of 8 μm (Costar, Washington, DC, USA). For the invasion assay, inserts were coated with 100 μL Matrigel (1:9 dilution; BD Biosciences, San Jose, CA, USA). Cells were trypsinized 24 h after transfection, and 5 × 10⁴ cells in 100 μL medium supplemented with 2% fetal bovine serum (FBS) were transferred to the upper Transwell chamber, whereas 600 μL medium supplemented with 10% FBS was added to the lower chamber. After incubation for 24 h, cells on the upper membrane surface were removed with a cotton tip, and those passed through the membrane were fixed with paraformaldehyde and stained with hematoxylin. The number of migrated/invaded cells was counted in 10 randomly selected fields under a microscope at high magnification. All experiments were repeated independently at least three times under identical conditions.

For proteasome-inhibition assays, cells were transfected with the indicated plasmids, and at 48-h post-transfection, the 26S proteasome inhibitor MG132 (s1748; Beyotime Biosciences) was added at a final concentration of 10 μM for 5 h, after which samples were collected. Cells were lysed in lysis buffer, and cell debris was pelleted by centrifugation at 12,000 rpm for 10 min at 4 °C. Supernatants were collected for IP, which was performed using whole-cell lysates (~ 200 μg protein), 4 μg to 10 μg antibody, and 20 μL protein A/G agarose (P2012; Beyotime Biosciences). Cell lysates were precleared with 20 μL agarose A/G beads by rocking for 1 h at 4 °C, after which beads were removed, and appropriate antibodies were added. Samples were then incubated with 20 μL agarose A/G beads by rocking for 4 h to 6 h at 4 °C, followed by the addition of lysates and incubation with rocking overnight at 4 °C. Immune complexes were collected by centrifugation, followed by washing in cell lysis buffer before analysis. Levels of PI3Kp85α ubiquitination were evaluated by IP of PI3Kp85α using anti-PI3Kp85α antibodies, followed by anti-hemagglutinin (HA) immunoblotting.

All statistical analyses were performed using SPSS 17.0 (SPSS, Inc., Chicago, IL, USA). Kaplan–Meier survival analyses were performed and compared using the log-rank test [29, 30]. Differences between groups were determined by Student’s t test, and a P < 0.05 was considered statistically significant.

Immunohistochemistry analyses of the 261 NSCLC tissues and immunofluorescence analyses of seven NSCLC cells revealed that MKRN2 was localized in the cytoplasm and nucleus (Fig. 1a and c). Additionally, MKRN2-expression levels were significantly lower in NSCLC tissues relative to those in normal lung tissues, alveoli, and bronchial tissues. We then investigated the relationship between MKRN2 protein level and clinical parameters. As shown in Table 2, MKRN2 protein levels were negatively correlated with clinicopathological parameters, including differentiation (P < 0.001), lymph node metastasis (P = 0.002), and pathological TNM stage (P = 0.005), whereas correlations between MKRN2 protein levels and tumor size (P = 0.936), histological type (P = 0.562), and patient age (P = 0.374) and sex (P = 0.318) were not statistically significant. Moreover, Kaplan–Meier analysis showed that high MKRN2 levels were associated with higher rates of NSCLC-patient survival (Fig. 1b). These findings suggest that MKRN2 might have an effect on lung cancer cell migration and invasion.

We then performed western blot analysis to detect MKRN2 protein levels in normal bronchial epithelial cells and commonly used NSCLC cell lines, finding that MKRN2 protein levels were lower in tumor cell lines than in normal bronchial epithelial cells (Fig. 1d). Therefore, we choose A549 and H1299 cell lines, which showed moderate levels of MKRN expression, for subsequent experiments.

Because MKRN2 levels correlated with NSCLC progression, we explored the biological function of MKRN2 in lung cancer cells by genetically manipulating MKRN2 expression. We downregulated or upregulated MKRN2 expression in A549 and H1299 cell lines using siRNA or the EX05-flag MKRN2 vector, respectively. Transwell assays showed knockdown of MKRN2 promoted the migration and invasion of A549 and H1299 cells as compared with control cells, whereas MKRN2 overexpression inhibited this activity by both cell lines (P < 0.05) (Fig. 2a). Consistent with these results, variations in MKRN2 protein levels affected the expression of proteins involved in cell migration and invasion, with levels of RhoA, Rho-associated protein kinase (ROCK)1, and matrix metalloproteinase (MMP)9 upregulated following MKRN2 knockdown, whereas MKRN2 overexpression caused the opposite effect (Fig. 2b and c). These findings demonstrated that MKRN2 suppressed the migration and invasion of lung cancer cells, consistent with our immunohistochemical results.

Based on the relationships between lymph node metastasis and MKRN2 expression in clinical cases and the functions of MKRN2 in NSCLC cell migration and invasion, we investigated the specific biological mechanisms associated with these effects. Previous studies reported associations between MKRN2 expression and the Akt-signaling pathway [31], specifically that MKRN2 acts downstream of PI3K; therefore, we examined whether PI3Kp85α affects MKRN2 expression, with results showing no changes in MKRN2 expression following PI3Kp85α downregulation in A549 and H1299 cells (Fig. 3a). However, evaluation of the effects of MKRN2 expression on activation of the Akt pathway in A549 and H1299 cells revealed that MKRN2 knockdown and overexpression upregulated and downregulated levels of PI3Kp85α, respectively, and that MKRN2 overexpression reduced Akt phosphorylation, whereas MKRN2 knockdown increased Akt phosphorylation in A549 cells (Fig. 3b), with the same effects detected in H1299 cells (Fig. 3c). A possible explanation for the differences in results between the two studies could be that the role of the nervous system differs from that of the tumor. To investigate whether MKRN2 function was mediated through the PI3K/Akt pathway, we evaluated the effects of this pathway using the inhibitor LY294002, which inhibits PI3K expression and reduces Akt phosphorylation. In A549 and H1299 cells transfected with si-MKRN2, LY294002 administration reduced the induction of cell migration and invasion (Fig. 3d), and the negative correlation between MKRN2 expression and RhoA, ROCK1, and MMP9 levels was reversed (Fig. 4a and b). We then verified interactions between MKRN2 and PI3Kp85α in A549 and H1299 cells by co-immunoprecipitation analysis (Fig. 4c), with results confirming this interaction. As shown in Table 3,immunohistochemistry of 34 pairs NSCLC tissues revealed that levels of p-Akt (Ser473) were low in cells exhibiting high levels of MKRN2 expression, whereas p-Akt (Ser473) levels were high in cells exhibiting low levels of MKRN2 expression (Fig. 4d), indicating a negatively correlated relationship between MKRN2 expression and p-Akt (P = 0.03). These results suggested that activation of the PI3K/Akt pathway contributed to the MKRN2-mediated migration and invasion of NSCLC cells.

We then evaluated how MKRN2 interacts with PI3Kp85α to activate the Akt-signaling pathway to promote cancer-cell migration and invasion. Real-time PCR analysis indicated that the PI3Kp85α expression was unchanged, regardless of MKRN2 overexpression or knockdown in both A549 and H1299 cells (Fig. 5a and b). These results indicated that the interaction between MKRN2 and PI3Kp85α occurs at the protein level rather than the mRNA level. Accordingly, we treated A549 and H1299 cells with cycloheximide (CHX) for 0 h to 9 h as a control and with both si-MKRN2 and CHX in the experimental group, with densitometric analysis revealing significantly higher levels of PI3Kp85α upon silencing MKRN2 relative to controls (Fig. 5c). This result indicated that MKRN2 expression promoted PI3Kp85α degradation. Previous studies reported that MKRN2 exhibits E3 ubiquitin ligase activity [25, 26]. Therefore, we hypothesized that MKRN2 might contribute to activation of the Akt pathway by attenuating PI3Kp85α ubiquitination.. Then we transfected flag-tagged MKRN2 and si-MKRN2 into A549 cells along with HA-ubiquitin and treated cells with the 26S proteasome inhibitor MG132. PI3Kp85α ubiquitination levels were then evaluated according to IP of PI3Kp85α using the anti-PI3Kp85α antibody, followed by anti-HA immunoblotting. The results revealed a positive correlation between MKRN2 levels and PI3Kp85α ubiquitination (Fig. 5d) in A549 cells, with similar results obtained in H1299 cells (Fig. 5e). These data suggested that MKRN2 was involved in the ubiquitin-dependent degradation of PI3Kp85α to activate the Akt pathway.