This article is mentioned in:

Introduction

microRNAs (miRNA) are small non-coding RNA molecules that inhibit gene expression at the transcriptional and posttranscriptional level by binding to the 3′-untranslated region (3′-UTR) of target mRNAs (1–4). miRNAs can bind to partially complementary recognition sequences of mRNA, subsequently causing mRNA degradation or translation inhibition, thus effectively silencing their target genes (5–7). Bioinformatic analysis of known miRNAs suggests that the majority of mRNAs can be targeted by miRNAs and that a single miRNA can regulate several hundred genes (8,9). miRNAs have been reported to participate in many important cellular processes, such as apoptosis, cell differentiation and proliferation, tumor suppression, development and metabolism (3,7–14). In recent years, more and more miRNAs have been detected by microarray analysis or other advanced technologies. At the same time, more protein factors have been confirmed to affect the expression of miRNAs, such as p53 (12,15,16). Thus, in order to elucidate the molecular mechanisms associated with non-small cell lung cancer (NSCLC) cell cycle arrest, identification of the regulatory network of miRNAs and proteins is critical.

To identify miRNAs which are differentially expressed in NSCLC and corresponding non-tumor lung tissues, miRNA solexa analysis was performed. Seven miRNAs were chosen for further study. All of the candidate miRNAs which have been verified in our laboratory play an important role in NSCLC cell cycle arrest. Potential target genes of 7 miRNAs were predicted by TargetScan (Table I). All proteins can regulate the cell cycle. To further identify the miRNAs that may regulate the expression of p53 or be regulated by p53 protein, we performed a prediction using software and constructed two expression vectors of p53 (with or without 3′-UTR). Only miR-150 was predicted to bind to the 3′-UTR of p53 by TargetScan software. The p53 expression vector contained the coding sequence (cd) only (pcDNA3.1-p53) and cds with 3′-UTR containing the binding sequences of miR-150 (pcDNA3.1-p53-3′-UTR) were constructed, respectively. Our results showed that miR-150 targets the 3′-UTR of p53 and reduces G1 phase arrest in the H1299 cell line triggered by p53. miR-34a, miR-184, miR-181a and miR-148 were significantly upregulated in the H1299 cells transfected by pcDNA3.1-p53. Moreover, the expression of miR-34a and miR-184 was consistent with p53 in the NSCLC cell lines, including SPCA-1, H1299, A549 and HCC827. These findings suggest that miR-150, p53 protein and relevant miRNAs may be members of a regulatory network in NSCLC tumorigenesis.

Table I Potential target genes of 7 miRNAs predicted by TargetScan.

Table I

Potential target genes of 7 miRNAs predicted by TargetScan.

miRNA	Target gene(s)	Published in TargetScan
miR-34a	CCNE2	2005, 2007, 2009
	CDKN1C	2009
miR-184
miR-181a	CAPRIN1	2005, 2007, 2009
	CCNJ/K	2009
	CCNT2	2005, 2007, 2009
	CCNG1
	CDK8	2009
miR-148	CDK5R1	2005, 2007, 2009
	CDK6/8/13/19	2003, 2005, 2007, 2009
miR-10a	CNNM4	2005, 2007
	CDK6	2009
miR-182	CCNJ/Y	2005, 2007, 2009
	GSPT1
	CNNM2/3	2007, 2009
	CNGA3
	CCND2/3
	CDKN1C
	CDK6	2009
	CCNE2	2005, 2007, 2009
miR-34c	CCNJL	2009
	CCND1	2005, 2007, 2009

Materials and methods

Cell culture

Human cell lines (SPCA-1, A549, HCC827, 95-D, HEK293T and BEAS-2B) were obtained from the Cell Bank of the China Academy of Sciences (Shanghai, China). H1299 was from the American Type Culture Collection (ATCC, Manassas, VA, USA). SPCA-1, A549, HCC827 and H1299 cell lines were derived from an NSCLC cell line, while 95-D is a small-cell lung cancer cell line with high metastatic potential. Human bronchial epithelial (BEAS-2B) cells were cultured in LHC-9 medium. A549 cells were cultured in F12K medium (Gibco, Gaithersburg, MD, USA). All the other lung cancer cells were cultured in RPMI-1640 medium (Gibco). Human embryonic kidney cells (HEK293T) were cultivated in Dulbecco’s modified Eagle’s medium (DMEM, Gibco). All the media were supplemented with 10% fetal bovine serum (FBS; HyClone Laboratories, Logan, UT, USA), 100 U/ml penicillin and 100 μg/ml streptomycin. Cells were cultured at 37°C in 5% CO2.

Clinical cancer samples

Human lung cancer samples were obtained from the Department of Oncology, Shanghai Chest Hospital affiliated to Shanghai Jiao Tong University, Shanghai, China, under ethical assessment.

Construction of recombinant expression vectors

The 3′-UTR of the tumor protein TP53 (p53) gene containing the miR-150 binding site (665 bp) was subcloned downstream of the firefly luciferase reporter gene in the pGL3 vector (Promega, Madison, WI, USA) and designated as pGL3-p53-3′-UTR. The plasmid pGL3-p53-3′-mUTR which contained the mutated binding site of miRNA-150 in the 3′-UTR was also constructed. The cds of p53 with or without the miR-150 binding sequence were cloned into the pcDNA3.1 (−) plasmid and named pcDNA3.1-p53 and pcDNA3.1-p53-3′-UTR, respectively. The primer sequences used in this study are shown in Table II.

Table II Primer sequences used in this study.

Table II

Primer sequences used in this study.

Plasmid	Primer sequences	Restriction enzyme
pGL3-p53-3′-UTR	Forward: CTAGTCTAGATCAGTCTACCTCCCGCCATAA Reverse: CCGGAATTCTGACAACTCCCTCTACCTAACCAG	XbaI EcoR1
pGL3-p53-3′-mUTR	Forward: CTGTGAGGGATGTCTAGCATATGTAAGAAATGTTCTTGCAGTTAAGGG Reverse: TTTCTTACATATGCTAGACATCCCTCACAGTAAAAACCTTAAAATCTAAGC
pcDNA3.1-p53	Forward: CCGCTCGAGATGGAGGAGCCGCAGTCAGA Reverse: CCGGAATTCCAAAACCCAAAATGGCAGGG	XhoI EcoR1
pcDNA3.1-p53-3′-UTR	Forward: CCGCTCGAGATGGAGGAGCCGCAGTCAGA Reverse: CCGGAATTCCCCTACCTAGAATGTGGCTGATTG	XhoI EcoR1

Luciferase assay

For reporter assays, HEK293T cells cultured in 24-well plates were transiently cotransfected with 400 ng luciferase vector pGL3-p53-3′-UTR or pGL3-p53-3′-mUTR and either miR-150 mimics or miRNA negative control (miRNA-NC). To determine the transfection efficiency, 20 ng pRL-SV40 (Promega) was cotransfected as the control. Reporter assays were performed at 36 h post-transfection using the Dual-luciferase assay system (Promega).

Quantitative real-time PCR (qRT-PCR) analysis of miRNAs and target genes

Total RNA was extracted from the cell cultures using TRIzol reagent (Bio Basic Inc., Toronto, Canada) according to the manufacturer’s instructions. Reverse transcription was performed using the M-MLV Reverse Transcriptase cDNA Synthesis kit (Takara, Dalian, China). A cDNA library of miRNAs was synthesized by the QuantiMir cDNA kit (Takara). U6 snRNA and the housekeeping gene 18S RNA were used as the endogenous control for miRNA and mRNA, respectively. The target genes and controls were treated under the same condition and analyzed by qRT-PCR using SYBR Premix Ex Taq™ (Takara) according to the manufacturer’s protocol.

Western blot analysis

Protein for western blot analysis was precipitated according to the standard protocol (17–20). Equal amounts of protein samples were subjected to SDS-polyacryl-amide gel electrophoresis (SDS-PAGE) and then transferred to a PVDF membrane. The membrane was soaked in Tris-buffered saline (TBS)-Tween buffer containing 5% low-fat milk for 60 min with gentle shaking and then incubated with a specific antibody overnight followed by washing and incubating with a secondary antibody and the final chemiluminescence ECL (Thermo Scientific, Rockford, IL, USA) detection of the bands. Protein bands were quantitated by densitometric analysis using Image Lab analysis software and expressed as the fold of the control after being normalized to GAPDH. The primary antibodies used were rabbit anti-p53 (1:1,000) and mouse anti-GAPDH (1:1,000). The secondary antibodies were rabbit anti-mouse (1:10,000) and mouse anti-rabbit (1:10,000). All antibodies were purchased from Cell Signaling Technology.

Cell cycle analysis

Cells were fixed in 70% ethanol for 12 h at 4°C. After washing with phosphate-buffered solution (PBS), cells were treated with RNase A (50 μg/ml) and stained with propidium iodide (PI; 25 μg/ml) for 30 min at 37°C. Samples were analyzed using MoFlo XDP flow cytometer (Beckman Coulter, Inc., Brea, CA, USA) and the distribution of cell cycle phases was determined using FlowJo software. The phase ratio (%) was calculated as the percentage of cells in the G1/S/G2 phase.

Statistical analysis

Results are expressed as the group means ± SEM and analyzed using GraphPad Prism 5 software, using t-tests for 2-group comparisons and one-way ANOVA for three or more group comparisons. A P<0.05 was considered to indicate a statistically significant result.

Results

miR-150 directly targets the p53 gene by interaction with the 3′-UTR

TargetScan and Pictar are two types of software broadly used on-line to predict the targets of miRNAs. Generally, the software was used to predict the targets of miRNA. In the present study, we used it to predict the target miRNA of p53. Results showed that only miR-150 targeted the 3′-UTR of p53. To confirm this, pGL3-p53-3′-UTR containing the miR-150 binding sequence and pGL3-p53-3′-mUTR were constructed (Fig. 1A and B). Analysis of the luciferase activity showed that the activity of miR-150 mimics cotransfected with pGL3-p53-3′-UTR was obviously inhibited when compared to miRNA-NC. However, the activity of miR-150 mimics cotransfected with pGL3-p53-3′-mUTR exhibited no difference when compared with miRNA-NC. Results of the luciferase activity assay indicated that mutated 3′-UTR affected the binding of miR-150 (Fig. 1C).

Figure 1 miR-150 directly targets the p53 gene. (A) miR-150 binding sequence in the 3′-UTR of p53. (B) Mutation of p53 3′-UTR was generated on the complementary site for the seed region of miR-150. *Indicates the mutant nucleotide. (C) Relative luciferase activity in HEK-293T cells tranfected with pGL3-p53-3′-UTR or pGL3-p53-3′-mUTR and either miRNA-NC or miR-150 mimics. The normalized luciferase activity in the control group was used as the relative luciferase activity. (D-F) The expression of p53. Both samples were cotransfected with pcDNA3.1-p53-3′-UTR and either miRNA-NC or miR-150-mimics. Data in D were obtained by qRT-PCR. 18S was used as an internal control. Data in E were analyzed by western blot analysis. GAPDH was used as an internal control. Data in F showed the expression level of p53 protein analyzed by Image Lab software. (G and H) Expression of miR-150 and p53 was detected by qRT-PCR in 13 lung cancer patient tissues, respectively. All experiments were at least repeated in triplicate with similar results. *P<0.05, ***P<0.001.

To further investigate whether miR-150 affects the expression of p53 at both the transcriptional and translational levels, we constructed an expression vector, pcDNA3.1-p53-3′-UTR, which contained the miR-150 binding sequence. The vector was cotransfected into H1299 cells with miR-150 mimics or miRNA-NC. The expression level of p53 mRNA in the miR-150 mimic-transfected H1299 cells was significantly decreased by 47% when compared with that in the miRNA-NC-transfected cells (Fig. 1D). Moreover, the expression level of p53 protein was significantly inhibited by 60% (Fig. 1E and F).

Expression of miR-150 and its target p53 was also detected in NSCLC patient tissue samples. The clinicopathological characteristics of 13 NSCLC patients are shown in Table III. The expression of miR-150 in stage T2 tissue samples was higher than that in T1 stage tissue samples. The corresponding target gene p53 was correlated with miR-150 expression (Fig. 1G and H). These data indicate that miR-150 directly targets p53 in NSCLC by binding to the 3′-UTR of the p53 gene.

Table III Data of the NSCLC patients.

Table III

Data of the NSCLC patients.

No.	Gender	Age (years)	Specimen type	Histologic type	Histology	Lymphatic invasion	pTNM
1	M	83	Wedge resection	Adenosquamous carcinoma	GX		T1bN0M0
2	M	67	Lobectomy	Adenocarcinoma	G3		T1bN0M0
3	M	62	Lobectomy	Adenosquamous carcinoma	G2		T2bN0M0
4	M	54	Lobectomy	Squamous cell carcinoma	G3		T2bN2M0
5	F	64	Lobectomy	Adenocarcinoma	G2		T1bN0M0
6	F	60	Lobectomy	Adenocarcinoma	G2	Present	T2aN2M0
7	F	52	Lobectomy	Adenocarcinoma (some BAC)	G2		T2bN0M0
8	M	61	Lobectomy	Adenocarcinoma	G1–G2		T2bN2M0
9	M	62	Lobectomy	Squamous cell carcinoma	G2	Present	T2aN1M0
10	F	45	Lobectomy	Adenocarcinoma	G2		T2aN0M0
11	F	67	Lobectomy	Adenocarcinoma	G2	Present	T1bN1M0
12	M	61	Lobectomy	Adenocarcinoma	G2		T2bN0M0
13	F	54	Lobectomy	Mucoepidermoid carcinoma	G3	Absent	T1bN0M0

Overexpression of miR-150 inhibits the cell cycle arrest by targeting p53

Cell cycle analysis was performed after transfection with pcDNA3.1-p53 or pcDNA3.1 for 48 h. Results showed that the cells transfected with pcDNA3.1-p53 were significantly arrested in the G1 phase when compared to the control which was transfected with empty vector pcDNA3.1 (Fig. 2). The expression vector pcDNA3.1-p53-3′-UTR was then cotransfected into H1299 cells with the miR-150 mimics or miRNA-NC. Cell cycle analysis was also performed 48 h later. Both of the miR-150 mimics- or miRNA-NC-cotransfected samples exhibited an obviously cell cycle arrest in the G1 phase when compared to the control which was transfected with pcDNA3.1. However, when compared to the pcDNA3.1-p53-3′-UTR- and miRNA-NC-cotransfected samples, miR-150 mimics cotransfected with pcDNA3.1-p53-3′-UTR inhibited cell cycle arrest (Fig. 3). These results indicate that miR-150 inhibits the cell cycle arrest triggered by p53.

Figure 2 p53 affects the cell cycle. (A) Cell cycle distribution of H1299 cells transfected with pcDNA3.1. (B) Cell cycle distribution of H1299 cells transfected with pcDNA3.1-p53. Cell cycle distribution was measured by flow cytometry. Cell cycle analysis was performed at 48 h post-transfection by staining DNA with propidium iodide prior to flow cytometry. (C) The phase ratio (%) of cells transfected with pcDNA3.1 or pcDNA3.1-p53. All experiments were at least repeated in triplicate with similar results. *P<0.05.

Figure 3 miR-150 affects the cell cycle distribution by targeting p53. (A-C) Cell cycle distribution measured by flow cytometry. Cell cycle analysis was performed at 48 h post-transfection by staining DNA with propidium iodide prior to flow cytometry. (D) The phase ratio (%) of cells transfected with pcDNA3.1 or pcDNA3.1-p53-3′-UTR with either miRNA-NC or miR-150-mimcs. All experiments were at least repeated in triplicate with similar results. *P<0.05.

Expression level of miRNAs in the H1299 cell line transfected with pcDNA3.1-p53

H1299 cell lines have a homozygous partial deletion of the p53 gene, and lack expression of p53 protein. To identify miRNAs which were differentially expressed after p53 ectopic expression in the H1299 cell line pcDNA3.1-p53 was transfected into H1299 cells. To avoid targeting by miRNAs in the 3′-UTR, the pcDNA3.1-p53 contained cds only. Western blot analysis was performed to detect the expression of p53 protein in the H1299 cell line transfected with pcDNA3.1 or pcDNA3.1-p53. The data showed that the p53 protein was significantly expressed in the H1299 cells transfected with pcDNA3.1-p53, but was not detectable in the control (transfected with pcDNA3.1) (Fig. 4A). qRT-PCR was then performed to identify miRNAs. Results showed that the level of miR-34a, miR-184, miR-181a and miR-148 expression were significantly upregulated by 2.8-, 2.5-, 2.2- and 1.7-fold of the control (Fig. 4B). However, the expression levels of miR-10a, miR-182 and miR-34c demonstrated no difference when compared with the control. The expression values were normalized to the levels of U6 RNA. In particular, all of the upregulated miRNAs play a cancer-suppressor role in lung cancer tumorigenesis which has been previously reported (21–27). Thus, these data indicate that p53 protein promotes the expression of miRNAs, particularly tumor suppressors miR-34a and miR-184.

Figure 4 Expression of p53 protein and miRNAs in H1299 cells transfected with pcDNA3.1-p53. (A) The expression level of p53 protein. Lanes 1 and 2 indicate control (transfected with pcDNA3.1) and pcDNA3.1-p53, respectively. (B) Expression of miRNAs. Data in A were obtained by western blot analysis. GAPDH was used as an internal control. Data in B were obtained by qRT-PCR. U6 was used as an internal control. Each assay was performed in triplicate. **P<0.01.

Expression level of miR-34a, miR-184 and p53 was relevant in NSCLC cell lines

To confirm that the expression level of miR-34a, miR-184 and p53 was relevant, 5 lung cancer cell lines (A549, H1299, 95-D, SPCA-1 and HCC827) were chosen as the samples and normal lung cell line (BEAS-2B) as the control. qRT-PCR analysis was performed to detect the expression levels of miR-34a, miR-184 and p53. The data showed that the expression of miR-34a and miR-184 was consistent with p53 except for that in the 95-D cell line (Fig. 5). Notably, all of the other 4 lung cancer cell lines originated from NSCLC. Altogether, these results indicate that p53 protein affects the expression of miR-34a and miR-184.

Figure 5 Expression of miR-34a, miR-184 and p53 in lung cancer cell lines. Expression of (A) miR-34a, (B) miR-184 and (C) p53. Data were performed by qRT-PCR. The expression level of each gene is the fold-change relative to the expression level of BEAS-2B. U6 and 18S were used as the internal control in A, B and C respectively. Each assay was performed in triplicate. *P<0.05, **P<0.01, ***P<0.001.

miR-150, p53 protein and miRNAs are members of a regulatory network in NSCLC tumorigenesis

As the results confirmed, miR-150 targets the 3′-UTR of p53. Overexpression of p53 can significantly enhance the expression of miR-34a, miR-184, miR-181a and miR-148. In particular, the expression of miR-34a and miR-184 was increased higher than 2-fold of the control. The targets of miR-34a have been previously reported (21,28,29). The protein cyclin E2 (CCNE2) is a key regulator in the cell cycle, and it is a potential target of miR-34a (Table I). miR-181a and miR-148 regulate the expression of CDC73 and CDK1, respectively. Both CDC73 and CDK1 can affect the G1 phase in the cell cycle (24,30–36). Thus, miR-150, p53 protein, the relevant miRNAs and their targets may consist of a complicated regulation network in NSCLC tumorigenesis (Fig. 6).

Figure 6 Role of miR-150, p53 and relevant miRNAs in the regulatory network among miRNA and protein factors. miR-150 inhibits the expression of p53, however, p53 promotes the expression of miR-34a, miR-184, miR-148 and miR-181a.

Discussion

H1299 cells have a homozygous partial deletion of the p53 gene, and lack expression of p53 protein (37). In the present study, we detected the expression variation of miRNAs when ectopic expression of p53 was present in the H1299 cell line. We found that miR-34a, miR-184, miR-181a and miR-148 expression was significantly upregulated. miR-34a and miR-181a have been reported as tumor-suppressor genes in neuroblastoma cells, urothelial bladder carcinoma, human brain glioma cells, head and neck squamous cell carcinoma and breast cancer (21–23,38–40). miR-34a can target many protein factors, such as the Notch-1 signaling pathway, Bcl-2, SIRT-1 and CDK1 and then promote the process of cell apoptosis and inhibit the cell cycle and proliferation (22,23,41,42). miR-181a can target k-ras, a typical oncogene (39). However, miR-184 and miR-148 have not been thoroughly studied. These results suggest that p53 protein may regulate the expression of various miRNAs which play a tumor-suppressor role in NSCLC cell lines. Yet, how p53 protein affects the expression of miRNAs is still unknown.

miR-150 was the only predicted miRNA which binds to the 3′-UTR sequence of p53. The luciferase activity analysis showed that the activity of miR-150 mimics cotransfected with pGL3-p53-3′-UTR was inhibited obviously compared to miRNA-NC. Western blot analysis also showed consistent results that the translation of p53 protein was inhibited significantly when miR-150 mimics were cotransfected with pcDNA3.1-p53-3′-UTR. miR-150 also reduces the cell cycle arrest triggered by p53. These results suggest that miR-150 may promote lung cancer tumorigenesis by targeting p53.

In conclusion, we confirmed that p53 is a direct target of miR-150, and overexpression of p53 promotes the expression of miRNAs including miR-34a, miR-184, miR-181a and miR-148. Our findings suggest that miR-150, p53 protein and relevant miRNAs consist of a complicated regulatory network in NSCLC tumorigenesis.

Acknowledgements

This study was in part supported by grants from the Innovation Program of Shanghai Municipal Commission of Sciences and Technology (11ZR141220), the National Natural Science Foundation (31170750), the National Key Research and Development Program of China (2011CB811304), the National Basic Research Program of China (2011CBA01105) and the Program of Baoshan District Commission of Sciences and Technology, Shanghai (CXY-2011-32).

References