MiR-744 facilitates non-small cell lung cancer progression by direct transcriptional regulation and indirect MAPK pathway activation mediated upregulation of C-FOS

Background: Metastasis is the leading cause of lung cancer associated death. Here, we focused on the function and molecular mechanism of miR-744 and its potential clinical application in lung carcinoma. Methods: The clinical cohort and data from TCGA were analyzed for the correlation of miR-744 and outcomes. Multiple NSCLC cell lines and a xenograft model were applied for the functional studies. Reporter assays were explored for transcriptional regulatory mechanism. Results: It was conrmed that the overexpression of miR-744 was signicantly correlated with lymph node metastasis and poor prognosis in NSCLC. Both in vitro and in vivo studies revealed that miR-744 overexpression aggravated the growth, invasion and metastasis of NSCLC cells. MiR-744 enhanced the resistance of lung cancer cells to radiotherapy and paclitaxel. MiR-744 positively regulated C-FOS by directly binding to the promoter of C-FOS and -1195 to -1227 and -298 to -323 bp upstream of C-FOS gene were the direct and ecient miR-744 binding sites. MiR-744 could also indirectly upregulate c-FOS protein by activating MAPK pathway. Conclusion: Our ndings uncover the function and a novel mechanism of miR-744 in mediating growth and metastasis of NSCLC cells. Our data suggests that miR-744 may serve as a possible therapeutic target for NSCLC.


Introduction
Despite recent advances in diagnostic and therapeutic measures, survival improvement is still a serious challenge in lung cancer [1]. Although various driven gene targets have been reported, less than 50% of lung cancer would bene t from these drugs. To date, however, promising molecular genetic alterations with clinical or prognostic signi cance in lung cancer have remained elusive. Therefore, it is necessary to further clarify the underlying molecular mechanisms and to identify speci c biomarkers or targets of lung cancer.
Intensive researches in past two decades have uncovered the presence and importance of miRNAs in human carcinogenesis [2]. In our previous miRNA microarray analysis, we demonstrated for the rst time that the expression of miR-744 was decreased signi cantly by metastasis associated gene 1(MTA1) knockdown [3] in non-small cell lung cancer (NSCLC). Meanwhile, we also veri ed the oncogenic potential of miR-744 in nasopharyngeal carcinoma tumorigenesis and metastasis [4]. Above results provided new insights into the molecular mechanisms underlying carcinogenesis and indicated the possibly essential role of miR-744 in tumor development. But little is known about the expression, exact biologic functions and mechanisms of miR-744 in NSCLC. Here, we demonstrate that miR-744 is signi cantly overexpressed in lung cancer and associated with a poor prognosis. MiR-744 can promote lung cancer cells growth, invasion and metastasis both in vitro and in vivo. It also can increase the resistance of lung cancer cells to paclitaxel and radiation.
C-FOS, de ned as an oncogene, plays a critical role in the tumorigenesis [5], and a better understanding of the epigenetic regulation of C-FOS is crucial. Importantly, we provide evidence for the rst time that miR-744 is a positive regulator of C-FOS by directly binding to its promoter region in lung cancer. Interestingly, we found that miR-744 could upregulate C-FOS protein level by activating the MAPK pathway and consequently enhanced the malignant phenotype of lung cancer cells. Our data rst demonstrated the regulatory mechanism of oncogene C-FOS by miR-744 at the level of transcription, which partly explained the proto-oncogene role of miR-744 in NSCLC.

Cell lines and clinical samples
The human lung cancer cell lines A549, H520 were obtained from ATCC (Manassas, VA), the H1299, SPC-A-1 cells and immortalized lung epithelial cell line HBE were purchased from Shanghai Cell Bank of Chinese Academy of Science (Shanghai, China). All cell lines were cultured in RPMI-1640 medium (Corning, NY, USA) containing 10% fetal bovine serum (FBS, GIBCO, South America, NY, USA) under a 5% CO 2 atmosphere at 37 °C. 15 matched non-small cell lung cancer (NSCLC) tissues and their adjacent noncancerous tissues (NCTs) were collected from Nanfang Hospital and stored in liquid nitrogen. All of the human samples were obtained with informed consents and histologically con rmed by hematoxylin and eosin (H&E) staining. This study project was approved by the Clinical Research Ethics Committee of Nanfang Hospital. Tissue chips consisted of 87 cases of human NSCLC were acquired from Outdo Biotech CO. Ltd. (Shanghai, China). The clinical pathological characteristics of 87 patients and 456 patients with NSCLC from The Cancer Genome Atlas (TCGA) were summarized in Supplementary Table 1 and Table 2. In Situ Hybridization (ISH) ISH experiments was performed according to the protocol described previously [6]. Brie y, probes for miR-744, positive control (U6) and negative control were purchased from Exiqon, Vedbek, Denmark. The intensity of cytoplasmic staining was scored manually semiquantitatively. The dominant staining intensity in tumor cells was scored as (S1 Fig): 0 = negative; 1 = weak; 2 = moderate; 3 = strong. A score of 3 was considered as high miR-744 expression and < 3 as low miR-744 expression.

Construction and stable transfection of lentivirus vector
The precursor sequence of miR-744 (MI0005559) was synthesized, annealed and then inserted into the AgeI/EcoR1 site of GV209 vector (GeneChem, Shanghai, China) to construct a vector continuously expressing miR-744, named as Lv-miR-744. The GFP vector was used for control. PLV-Luci-control was also constructed by GeneChem (Shanghai, China). A549 and H520 cells were co-transfected with Lv-miR-744 or Lv-control vector and pLV-Luci-control. The clones with GFP expression were isolated under puromycin (2 mg/mL) selection for 2 weeks. Plasmid construction and transient transfection AntagomiR 744, antagomir negative control (antagomir-NC), agomiR 744, agomir negative control (agomir-NC) were designed and synthesized by RiboBio (Guangzhou, China). Si c-FOS and si control were designed and synthesized by Genepharma (Shanghai, China). The cells were seeded in 6-well plates at 30-50% density. Transient transfection was performed with Lipofectamine 3000 reagents according to the manufacturer's instructions (Invitrogen, USA). For all the experiments, cells were collected at 48 h after transfection for further research.
RNA extraction and quantitative real-time PCR (qRT-PCR) analyses Total RNA was extracted from cell lines using TRIzol® Reagent (Invitrogen, CA, USA) according to the manufacturer's instruction. The reverse transcription of RNA for miR-744 and c-FOS mRNA was performed using SYBR® PrimeScript™ miRNA RT-PCR Kit (TaKaRa, Dalian, China) and PrimeScript™ RT reagent Kit (TaKaRa, Dalian, China) separately. The real-time quantitative RT-PCR assay was carried on a Mx3005P real-time PCR instrument (Stratagene, USA) with SYBR® Premix Ex Taq™ kit (TaKaRa, Dalian, China). Small nuclear RNA U6 (U6 snRNA) or Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control for miRNA or mRNA quanti cation respectively. The sequences of the primers used for qRT-PCR were listed in Supplementary Table 3.
Endogenousβ-Actin protein was used as internal control. Microarray analysis Followed by RNA extraction with TRIzol reagent, cells transfected with antagomiR-744 or antagomir-NC were analyzed by Beijing CapitalBio Corporation with Affymetrix PrimeView™ Human Gene Expression Array.

EdU assay
After transfection for 48 h, cells were used to measure DNA synthesis with a Cell Light™ EdU imaging detecting kit (RiboBio, Guangzhou, China) according to the manufacturer's instructions.

CCK8 cell assay
After transfection, the cells were seeded in 96-well plates (5000 per well). After the cells were attached to the culture dish, paclitaxel was added to each well with the following concentration gradients (0-16 nmol/L or 0-32 nmol/L). 10 µl CCK8 was added into each well after 78 h and was incubated for 2 h. The absorbance values (OD) were measured at a wavelength of 450 nm using a microplate reader.

Colony formation assay
The cells were seeded at a density of 200-10000 cells/well in a 6-well plate. The cells were irradiated with different doses (0, 2, 4, 6 and 8 Gy) of X-rays. After 10 days incubation in RPMI-1640 medium with 20% FBS at 37 °C, 5% CO 2 , the cells were xed in 4% paraformaldehyde and stained with 0.5% crystal violet in 20% methanol. The colonies were counted. The survival curves were tted by a linear-quadratic model [S = exp (-αD-βD 2 )] using GraphPad Prism 5.0 (GraphPad Prism, San Diego, CA, USA). Cell migration and invasion assays Cell invasion assay was performed using Transwell chamber (Corning, NY, USA) in a 24-well plate.
Transfected cells (5-10 × 10 4 cells per 100 µl) in serum-free medium were seeded into the upper chamber with 12.5% matrigel-coated membrane (BD Bioscience, San Jose, CA), while the lower chamber was lled with 500 µl fresh complete medium. After being incubated for 18-24 h at 37℃ with 5% CO2, non-invading cells were removed from the upper surface of the lter by wet cotton swab. Invading cells that adhered to the lower surface of the chambers were xed in methyl alcohol and stained with hematoxylin. The invading cells were manually counted at 200 × magni cation under a inverted microscope. Five visual elds of each chamber were randomly chosen and photographed. Similar upper inserts without the matrigel precoating were used for the migration assay.

Wound healing assay
As described in previous study [4],when cells were grown to approximately 90% con uency (after 48 h of transfection), an arti cial wound was created with a 10 µl pipette tip. The cells were then cultured in fresh medium without FBS. To visualize wound healing, images were taken at 0 h and 36 h. The relative percentage of wound healed was calculated as (the width of wound at 0 h -the width of wound at 36 h)/the width of wound at 0 h. Promoter reporter construction and luciferase assays The pGL3 vectors containing human c-FOS promoter with wild type or mutant putative miR-744 binding site were synthesized and con rmed with DNA sequencing by HuaAnPingKang Co., Ltd (Shenzhen, China). The sequences of these wild type or mutant promoter constructs were listed in Supplementary Table 4-7. The cells were plated onto 24-well plates at a density of 5 × 10 3 cells/well for transfection study. A mixture of 150 ng c-FOS promoter constructs (Luc-C-FOS-A, Luc-C-FOS-B, Luc-C-FOS-C, Luc-C-FOS mut−1 , Luc-C-FOS mut−2 , Luc-C-FOS mut−3 , Luc-C-FOS mut−4 ) or responding control vectors, 100 nmol agomiR 744 or agomir negative control (agoNC) and 5 ng pRL-TK vector (Promega, WI, USA) were co-transfected into cells for each well using Lipofectamine 3000, and then incubating for 48 hours. Dual luciferase reporter assay system (Promega, WI, USA) was used to perform the luciferase assay according to the manufacturer's instruction.

Xenograft tumor growth assay
Four to six weeks old athymic male BALB/c nude mice were purchased from the Guangdong Medical Laboratory Animal Center (GDMLAC, Foshan, China), and were group-housed under speci c pathogen-free (SPF) conditions. The experimental protocols were approved by the Institutional Animal Care and Use Committee at our University. A549 or H520 cells stably overexpressing miR 744 or the control vector (8 × 10 5 cells) were injected subcutaneously into the dorsal ank of mice (n = 5 per group). A549 or H520 cells, transfected with antagomiR-744 or antagomir-NC, were also used for xenograft experiments following the procedures described above. After 8 days of implantation of tumor cells, tumor size was measured every 3 days and tumor volumes were calculated with the following formula: V= (L × W 2 )/2, where L is the length and W is the width of each tumor. At the end of experiments (about after 4-6 weeks), the mice were sacri ced and tumors were dissected and weighed. In vivo metastasis assay A549 or H520 cells (2 × 10 6 ) stably overexpressing miR 744 or NC were injected into the tail veins of BALB/c nude mice (n = 5 per group). After 30 days, the lung colonization capacity of the cancer cells was analyzed based on the normalized bioluminescence imaging. Then the mice were sacri ced and the lung tissues were dissected and xed in 4% formaldehyde overnight. The xed samples were embedded in para n and stained with hematoxylin and eosin (HE). Lung metastasis index was calculated as our previous study described [4]: metastatic tumor areas/total lung areas.

Statistical analysis
Statistical analyses were performed using GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA) and SPSS 13.0 (SPSS Inc., Chicago, USA). Each experiment was repeated at least three times and data were expressed as the mean ± SE. Two-tailed unpaired Student's independent samples t test was used to assess differences among groups in in vitro or in vivo studies. Brown-Forsythe method was used for the relative mRNA expression of each cell lines. Spearman's correlation was used to analyze the relationship between miR 744 and c-FOS mRNA expression. Kaplan Meier survival analysis and cox regression were performed to calculate overall survival (OS) and estimate the prognostic factors. P value < 0.05 was considered statistically, and *p < 0.05, **p < 0.01, ***p < 0.001.

Results
MiR-744 is up-regulated in NSCLC and associated with poor clinical outcome We rst detected the expression of miR-744 in 15 matched NSCLC tissues and their adjacent tissues and found that miR-744 was up-regulated in cancer tissues (Fig. 1A). MiR-744 expression was then quanti ed in 87 formalin-xed para n-embedded (FFPE) NSCLC tissues (training set) and 456 primary NSCLC from TCGA (validation set). Notably, miR-744 expression was signi cantly associated with lymph node metastasis (S1-S2 Table). Survival analyses revealed that patients with higher level of miR-744 presented shorter OS (Fig. 1B,training set: p = 0.005; validation set: p = 0.002). The multivariate Cox regression analyses in two cohorts both indicated that higher expression of miR-744 is an independent unfavorable prognostic marker for OS (S8-S9 Table,

MiR-744 enhances NSCLC cells malignant phenotype and radio/chemoresistance in vitro
To investigate the biological functions of miR 744 in the development and progression of lung cancer, we next performed gain and loss of function studies in lung cancer cells lines by transient transfection with miR-744 agomir or antagomir ( Fig. 2A). We found that overexpression of miR 744 remarkably increased the proliferative, migrative and invasive ability of lung cancer cell lines, while knockdown endogenous miR-744 had the opposite effect (Fig. 2B-2D and S3 Fig).
To investigate the effect of miR-744 on the sensitivity of NSCLC cells to paclitaxel and radiation, A549 and H520 cell lines were transfected with agomiR 744 or antagomiR-744 or corresponding control to perform CCK8 assay and colony formation assay. Cell viability assays indicated that knockdown of miR-744 could improve the response to paclitaxel of A549 cells and H520 cells. Conversely, miR-744 mimics decreased paclitaxel-mediated inhibition of cell proliferation (Fig. 2E). The survival curves in the colony formation experiment demonstrated that agomiR-774 transfected A549 and H520 groups were more radioresistant than the corresponding control groups (Fig. 2F and S10 Table). Taken together, miR-744 enhanced the chemoresistance and radioresistance of lung cancer cells.

C-FOS was directly targeted at transcriptional level and indirectly upregulated through the activation of MAPK pathway by miR-744 in NSCLC cells
To gain insights into the molecular mechanism of miR-744 in the promotion of malignant phenotype of lung cancer cells, we analyzed the mRNAs differentially expressed before and after miR-744 downregulation by using genome-wide expression pro le chip (GSE146601). A total of 11 differentially expressed mRNA were obtained, and all of them were down-regulated (S11 Table, screening criteria: Qvalue ≤ 0.05, and fold change ≥ 1.8 or fold change ≤ 0.55). The results of qRT-PCR showed that only the mRNA level of C-FOS was changed more than twice in different transient antagomiR-744 or agomir-744 transfected cell lines ( Fig. 3A and S4 Fig). All of these con rmed that direct binding of miR-744 to the promoter region − 1195 to -1227 and − 298 to -323 bp upstream to the TSS led to the transcription activity of C-FOS increased (Fig. 3D). The degree of association between the expression of miR-744 and other 20503 genes was analyzed by calculating Pearson's coe cient and Spearman's rank correlation coe cient in 195 cases of NSCLC from TCGA database. 1155 genes with P value less than 0.05 were selected for pathway annotation. The results showed that the expression of miR-744 was related to cell cycle regulation, MAPK pathway, WNT pathway, cytoskeleton regulation, chemokines, local adhesion, JAK/STAT and p53 pathways (S6 Fig). c-FOS was contained in MAPK pathway based on KEGG PATHWAY Database (https://www.genome.jp/kegg/pathway.html). To validate the activation of MAPK pathway by miR-744, we detected the expression of GRB2, Raf1, total ERK1/2, ERK1/2 phosphorylation and c-FOS in A549, SPC-A-1, H1299, H520 cells transfected with agomiR-744 or agomir-NC by western blot. As Fig. 3E shown, a prominent increase in the expression of GRB2, Raf1, ERK1/2 phosphorylation, c-FOS was observed, except for the expression of total ERK1/2. Taken together, these results indicated an indirectly regulation of c-FOS by miR-744-mediated activation of MAPK pathway. C-FOS mediated the promotion of miR-744 on NSCLC cells migration, invasion and proliferation To evaluate whether malignant phenotype of lung cancer cells enhanced by miR-744 was mediated via C-FOS, we performed co-transfection of si-C-FOS plasmid or si-control with agomiR-744 or agomir-NC respectively in A549, SPC-A-1, H1299 and H520 cells and the transfection effect was veri ed by qRT-PCR and western blot (Fig. 4A). It was noted that si-C-FOS partially or completely reversed the effect of agomiR-744 on the migration, invasion and proliferation of lung cancer cells (Fig. 4B-4D and S7 Fig).
miR-744 functions as a tumor promoter in lung cancer metastasis and growth in vivo.
Antagomir has been reported to be able to interfere with microRNAs expression for a long time and been used in animal experiments [37]. To nd out whether the expression of miR-744 could be continuously inhibited by antagomiR-744, we observed the interference effect of three concentrations of antagomiR-744 for 27 days and select a concentration of 125 nm for the further study (Fig. 5A). We also developed Lv-miR-744 to stably overexpress miR-744 in lung cancer cells for in vivo study (Fig. 5B). The growth promoting effect of miR-744 was con rmed as expected in xenograft tumor models. Wheras, tumor growth was signi cantly inhibited in antagomiR-744 group compared to antagomir-NC. (Fig. 5C and 5D). More prominently, signi cant enhancement of lung metastasis was also observed by bioluminescent imaging, which was further veri ed by H&E staining (Fig. 5E-5H) Accordingly, the expression of C-FOS presented the same trend as that of miR-744 (S8 Fig). Collectively, these data further demonstrated that miR-744 played critical role in the development of NSCLC growth and metastasis, and targeting miR-744 with antagomiR-744 might serve as an effective therapeutic approach to attenuate NSCLC progression.

Discussion
It is well known that miRNAs have emerged as key mediators of tumorigenesis and development, as they broadly participate in cellular processes, including invasion, proliferation, senescence, extracellular matrix (ECM) synthesis and death [38][39][40]. MiR-744 was one of the microRNAs that changed after the interference of a metastasis related gene in NSCLC cells, which was the rst time that miR-744 caught our attention [3].
Up to now, it has been observed that miR-744 was a promising tumor molecular marker in some tumor tissues or serum (Table 1), including chronic lymphocytic leukemia [8], hepatocellular carcinoma [12], osarcoteosma [18] and gastric cancer [31]. However, there is still no study on the expression of miR-744 in lung cancer. We rst found that the expression of miR-744 in NSCLC tissues was signi cantly higher than that in paracancerous tissues. To make the clinical signi cance of miR-744 more clear, we used different methods to detect the expression of miR-744 in two NSCLC cohorts. It was con rmed that the high expression of miR-744 was signi cantly related to lymph node metastasis and poor prognosis of NSCLC, and was an independent prognostic molecular marker of NSCLC. This is comparable to ndings in the studies on pancreatic cancer [19][20][21], prostate cancer [23,24] and nasopharyngeal carcinoma [4,36]. All of these suggested that miR-744 might play an important role in the process of NSCLC progression and metastasis, and need to be further studied.
Although miR-744 has been mentioned in a few NSCLC studies, they were limited to the indirect description of miR-744 expression on the cell level [15,16]. Only two studies preliminarily revealed that miR-744 was down regulated in EGFR TKI primary resistant patients' plasma [17] and might function as a tumor suppressor in vitro [14]. In the current study, we provided evidence at both the cellular and animal levels that miR-744 acted as an oncogene and promoted the proliferation, migration, invasion and metastasis of NSCLC cells in vitro and in vivo, which was consistent with the ndings in previous clinical specimen-based study. At present, there are limited studies about the role of miR-744 in chemosensitivity [9,11,20,29,33] and radiosensitivity, and the reported in uence of miR-744 on the sensitivity of cancer cells to different drugs was quite different [9,11]. Our study con rmed that miR-744 could reduce the sensitivity of lung cancer cells to paclitaxel, a commonly used chemotherapy drug in clinical practice, and miR-744 also increased the resistance of cells to radiotherapy.
The role of miRNAs partially relied on their regulation of the expression of many cancer-related genes through post-transcriptional repression [40], and they silenced genes through pairing between limited bases in the seed region and complementary sequences in target mRNA 3'-UTR. At present, there are reports that miRNAs could target 5'-UTRs [41], coding regions [42], promoters [43], or gene terminal downstream sequences [44] to silence gene expression, which revealed the possibility of multi-mode effect of miRNAsmediated gene regulation. However, the regulatory mechanism of miR-744 in lung cancer remains to be elucidated.
c-FOS, as a major component of activator protein (AP)-1 complex, has been involved in cell differentiation, proliferation, angiogenesis, invasion, and metastasis [45,46]. Here, we found that c-FOS was an important target gene positively regulated by miR-744, which mediated the promotion of miR-744 on the malignant phenotype of NSCLC cells. This phenomenon has aroused our research interest, because this result could not be explained by the classical regulatory mechanism of targeting 3'-UTR. In our previous study, we also proposed and con rmed that ARHGAP5, as a critical proto-oncogene, is regulated at the transcription level by miR-774 through direct binding with its promoter region, and contributed to the promotion of miR 744 on NPC invasion and migration [4]. In this study, we also con rmed the direct binding sites between miR-744 and C-FOS promoter by luciferase reporter gene analysis. In addition, high expression of miR-744 could activate MAPK signaling pathway, and C-FOS is just the downstream gene of this pathway. This newly disclosed miR-744/MAPK/C-FOS axis further revealed the mechanism of NSCLC progression (Fig. 6).
Our previous studies have con rmed that c-Jun could regulate miR-744 at the transcription level in both NPC and lung cancer cells [47], because C-JUN can combine with C-FOS to form AP-1 early response transcription factor [48,49], so whether there is a positive feedback between miR-744/C-FOS axis and C-JUN/miR-744 axis needs further study. In conclusion, we presented here that miR-744 could directly and indirectly up-regulate the expression of C-FOS, and then promoted the malignant phenotype of NSCLC cells.
The newly identi ed miR-744/C-FOS axis provided new insights into the progression of lung cancer. At the same time, it added new contents for understanding the biological function of miR-744 gene and its role in tumorigenesis and development, and also provided potential targets for prognosis prediction and drugs intervention in NSCLC.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.

Figure 6
Schematic diagram of miR-744/MAPK/C-FOS axis mediating NSCLC progression. miR-744 positively regulated C-FOS by directly binding to the promoter of C-FOS and also indirectly upregulate C-FOS protein by activating MAPK pathway, thus mediating the growth and metastasis of NSCLC.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. S1Table.docx S2Table.docx