A long non-coding RNA lncRNA-PE promotes invasion and epithelial–mesenchymal transition in hepatocellular carcinoma through the miR-200a/b-ZEB1 pathway

GEO data set analysis

Gene expression data (GSE45267, GSE55092, and GSE40873 profiling data) were downloaded as raw signals from Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo), interpreted, normalized, and log2 scaled using software R.

Tissue samples and cell lines

Human HCC cell lines BEL-7402 and SK-Hep-1, as well as the normal human liver cell line LO2, were obtained from the American Type Culture Collection (ATCC, USA) and were cultured at 37°C in an atmosphere containing 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) plus 10% fetal bovine serum (FBS; Gibco, USA).

HCCs and surrounding control tissue specimens were obtained from 24 patients at General Hospital of PLA (Beijing, P.R. China) after surgical resection with informed consent. The tumor tissues and adjacent normal tissues were frozen in liquid nitrogen after resection. No patient in this study received chemotherapy or radiation therapy before the surgery. Liver samples were fully clinically characterized (Table 1). This study was performed with the approval of the Medical Ethical Committee of General Hospital of PLA, which were previously used in the other study. ¹³

OPEN IN VIEWER

Table 1.

Patient clinical features.

Patient no.	Age	Gender	Tumor size (cm3)	Edmondson Grade	HBsAg	HCV-Ab	Cirrhosis
1	66	F	15 ×7 × 5	3	Positive	Negative	No
2	45	F	6 × 6 × 5.5	3	Negative	Negative	No
3	39	M	12 × 11 × 6	3	Positive	Negative	No
4	51	M	2.5 × 2 × 2	2	Positive	Negative	Yes
5	47	M	5 × 5 × 4	2	Positive	Negative	Yes
6	50	M	16 × 14 × 8	3	Positive	Negative	No
7	52	M	8 × 7 × 4.5	2	Positive	Negative	No
8	64	M	8 × 5 × 4	2	Negative	Positive	No
9	50	F	6 × 5.5 × 2.5	1	Negative	Negative	No
10	52	M	7 × 7 × 7	3	Positive	Negative	Yes
11	56	M	8 × 6.5 × 5	2	Positive	Negative	Yes
12	48	M	2 × 2 × 1.5	2	Positive	Negative	No
13	59	M	8.8 × 8.5 × 8	2	Negative	Negative	No
14	50	M	2.5 × 2.5 × 2	3	Positive	Negative	No
15	52	M	10.5 × 8 × 5	3	Positive	Negative	Yes
16	50	M	2.5 × 1.8 × 1.5	1	Negative	Negative	No
17	41	F	12 × 10 × 8	3	Positive	Negative	No
18	33	M	10 × 8 × 7	4	Positive	Negative	Yes
19	35	F	5 × 4 × 4	3	Positive	Negative	Yes
20	63	M	14 × 10 × 10	2	Positive	Negative	Yes
21	70	M	3.5 × 3.5 × 3	2	Positive	Negative	Yes
22	34	M	10 × 7 × 7	3	Positive	Negative	Yes
23	51	M	2.5 × 2 × 2	2	Positive	Negative	Yes
24	49	M	5 × 4 × 2.5	3	Positive	Negative	Yes

HCV-ab: hepatitis C virus antibodies; HBsAg: hepatitis B surface antigen.

Clinicopathological characteristics of the patients with HCC are presented. All samples were obtained during surgery.

RNA extraction and real-time polymerase chain reaction

Total RNA of cells and tissues was isolated using TRI Reagent (Sigma, USA). First-strand complementary DNA (cDNA) was generated using the ImProm-II Reverse Transcription System (A3800; Promega, USA) and random primers or oligo-dT primers. We used SYBR Premix Ex Taq (RR420A; TaKaRa, Japan) and gene-specific primers for reverse transcription polymerase chain reaction (RT-PCR). The gene-specific primers are shown in the following. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was employed as an endogenous control for mRNA and lncRNA. For microRNA (miRNA) analysis, quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed as described before ¹⁴ and U6 small nuclear RNA (snRNA) was used for normalization. The relative expression levels of RNAs were calculated using the comparative Ct method.

Primers used for real-time PCR are as follows: lncRNA-PE-F: 5’AGGATAGGATGGGCTTTTTCTGT-3’; lncRNA-PE-R: 5’-GCAATTGCCCTAGACCCAGT-3’; pri-miR-200b-F: 5’-CAGCCGTGGCCATCTTACT-3’; pri-miR-200b-R: 5’-CCGCCGTCATCATTTACCAGG-3’; GAPDH-F: 5’-TCAGTGGTGGACCTGACCTG-3’; GAPDH-R: 5’-TGCTGTAGCCAAATTCGTTG-3’; U6-F: 5’-CGCTTCGGCAGCACATATACTA-3’; U6-R: 5’-CGCTTCACGAATTTGCGTGTCA-3’; miR-200b: 5’-TAATACTGCCTGGTAATGATGA-3’; miR-200a: 5’-CATCTTACCGGACAGTGCTGGA-3’; and miR-451a: 5’-AAACCGTTACCATTACTGAGTT-3’.

Small interfering RNAs and miRNAs

Small interfering RNAs (siRNAs) for lncRNA promoting EMT (lncRNA-PE) were purchased from GenePharma (China), the sequences of siRNAs are as follows: silncRNA-PE-1: 5’-CCGGCCCAUAUGUUGAAAUTT-3’ and 5’-AUUUCAACAUAUGGGCCGGTT-3’; silncRNA-PE-2: 5’-CCAUGCGAACACACAGCAATT-3’ and 5’-UUGCUGUGUGUUCAUGGTT-3’. For miR-200a/b overexpression studies, 20 nM miR-200a/b oligonucleotide duplex (miRBase Release 21) was purchased from GenePharma.

RNA immunoprecipitation

Cells were harvested using phosphate-buffered saline (PBS) and lysed in immunoprecipitation (IP) lysis buffer (Life Technologies, USA) containing an RNase inhibitor (RRI, 0.2 U/µL; TaKaRa, Japan) and protease inhibitors (Roche, Switzerland) for at least 10 min at 4°C. To perform IP experiments, the supernatants of lysates were first incubated with antibodies for 4 h at 4°C, and then, beads were added into the mixture and rotated overnight at 4°C. All precipitates were washed thrice with a wash buffer (20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 150 mM NaCl, 5 mM MgCl₂, and 1% NP40) and then boiled with sodium dodecyl sulfate (SDS)-loading buffer for 10 min. For RNA IP, RNAs were extracted by mixing the TRI Reagent with beads and then subjected to reverse transcription and real-time PCR analysis.

Luciferase assay

Dual-luciferase assays were performed according to the Dual-Luciferase Reporter Assay System guidelines (Promega, USA). Cells cultured in 24-well plates were incubated with passive lysis buffer (Promega) for 15 min at room temperature. Cell lysates were transferred into EP tubes together with Luciferase assay reagent II (LAR II), and an equal volume of the Stop&Glo Reagent was added into the EP tubes. The fluorescence signals of cell lysates were analyzed using a GloMax Luminometer (Promega). The data are presented as means ± SD of three independent experiments.

Cell migration and invasion assay

In the wound-healing assay, cells were scratched in the monolayer and cultured under normal conditions. The migrated distances were measured at 0, 24, and 48 h after scratching. For the invasion assay, BEL-7402 and SK-Hep-1 cells were serum starved for 24 h in DMEM containing 0.1% FBS. Serum-starved cells were resuspended in DMEM containing 0.1% FBS, and 2 × 10⁵ cells were added to the upper chamber of each well (6.5 mm in diameter, 8-µm pore size; Corning, USA) coated with 30 mg/cm² Matrigel (ECM Gel; Sigma). Medium containing 0.1% FBS and supplemented with hepatocyte growth factor (HGF; 20 ng/mL) was placed in the lower compartment of the chamber. After 24 h at 37°C, cells on the upper membrane surface were removed by careful wiping with a cotton swab and the filters were fixed by treatment with 95% ethanol for 30 min and stained with 0.2% crystal violet solution for 30 min. Invasive cells adhering to the lower surface of the filter were then counted (five high-power fields/chamber) using an inverted microscope. The migration assay is the same as the invasion assay, except that no Matrigel was used and the permeating time for cells was 12 h.

Western-blot analysis

Cells were lysed with radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors. After bicinchoninic acid (BCA) quantification, equal amounts of protein were separated by 10% SDS–polyacrylamide gel electrophoresis (PAGE), electrophoretically transferred to nitrocellulose membrane (Millipore, USA), and incubated with primary antibodies against the target proteins followed by appropriate secondary antibodies (Pierce, USA). Signals were visualized with SuperSignal^® West Pico Chemiluminescent Substrate (Pierce) and film exposure.

Antibodies for western blotting

For western blot, antibodies against ZEB1 (sc-10570; Santa Cruz Biotechnology, USA), E-cadherin (sc-8426; Santa Cruz Biotechnology), vimentin (sc-6260; Santa Cruz Biotechnology), N-cadherin (sc-7939; Santa Cruz Biotechnology), ZO-1 (sc-10804; Santa Cruz Bio-technology), and GAPDH (10494; Proteintech, USA) were used.

Vector construction

The cDNA encoding lncRNA lncRNA-PE was PCR-amplified by KOD Plus Neo DNA Polymerase (Toyobo, Japan) and subcloned into the pcDNA3 vector (Invitrogen, USA). The promoter region of miR-200a/b was PCR-amplified and subcloned into BglII and NheI sites of pGL3-enhancer (Promega) vector. Primers used for the construction of vectors are as follows: lncRNA-PE-KpnI: 5’-AGCTGGTACCGACGGAGTCTCGCTCTGTCGCCCA-3’; lncRNA-PE-XhoI: 5’-CGCACTCGAGATCATTTAATTTTCCACAATTGAT-3’; miR-200a/b-promoter-NheI: 5’-ACCTGCTAGCCATTCCCGGACAGTCACAG-3’; and miR-200a/b-promoter-BglII: 5’-CAGGAGATCTCCTGGCACAGGAAGTCAGTT-3’.

Isolation of cytoplasmic and nuclear RNA

Cytoplasmic and nuclear RNA were isolated and purified using the RiboTrap Kit (RN1011; MBL, Japan) according to the manufacturer’s instructions.

Statistical analysis

Microsoft Excel software and GraphPad Prism 5 were used to assess statistical differences between experimental groups. The MO-free survival (MFS) was evaluated by the survival pack of R. The median expression level for each gene was used as a cutoff value calculated by Cutoff Finder. ¹⁵ Statistical significance was analyzed using t tests, and p < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, and ***p < 0.001).

LncRNA-PE is upregulated in HCC tissues

We sought to screen for lncRNAs that can promote the progression of HCC and EMT. By combining microarray data from our previous study with data in the article that presented the discovery of lncRNA-atb, ¹² we serendipitously uncovered a lncRNA, BC013423 (1454-bp long). It is upregulated both in HCC ¹⁶ and upon transforming growth factor-beta (TGF-β; treatment. ¹² We inferred that this lncRNA may have a role in promoting HCC progression and EMT and sought further validation. For convenience, we call this lncRNA lncRNA-PE in the later parts of this article. First, we used the online database Coding Potential Calculator ¹⁷ to calculate its coding potential, which was 0.663685, compared with the score of lncRNA SNHG1, 0.951527, and that of MALAT1, 0.348632, which indicated that it had very weak coding potential. To determine the role of this lncRNA in HCC progression, we first verified the expression of lncRNA-PE in HCC cells, which was increased in two HCC cell lines relative to the expression in the normal human liver cell line LO2 (Figure 1(a)). We next analyzed its expression level in HCC tissues compared with their normal counterparts in GSE45267 and GSE55092 and discovered that lncRNA-PE was upregulated in HCC tissues (Figure S1(A) and (B)). Then, we validated its expression levels in 24 pairs of HCC tissues and adjacent liver tissues using quantitative real-time PCR. The lncRNA-PE level was significantly upregulated (>2 fold) in 50% of HCC tissues (12 of 24 patients; Figure 1(b)) compared with adjacent normal tissues. Moreover, Figure S1(C) indicates that the cumulative recurrence-free survivals were significantly associated with lncRNA-PE expression in GSE40873 (log-rank test, p = 0.014). These data indicate that ectopic lncRNA-PE expression may be related to HCC progression and EMT, and lncRNA-PE could be a potential HCC prognosis maker.

OPEN IN VIEWER

Figure 1.

LncRNA-PE levels were upregulated in hepatocellular carcinoma cells and tissues. (a) LncRNA-PE expression was assessed by real-time PCR in hepatocellular carcinoma cell lines and a normal liver cell line. lncRNA-PE levels were normalized to GAPDH levels (*p < 0.05). (b) LncRNA-PE expression was assessed in hepatocellular carcinoma tissues and adjacent normal tissues (n = 24). LncRNA-PE expression was assessed by real-time PCR, and values were normalized to those for GAPDH. Bars represent the ratio between expression in adjacent normal and hepatocellular carcinoma tissues (log scale). (c) The mRNA levels of lncRNA-PE after lncRNA-PE siRNAs were transfected in SK-Hep-1 cells.

LncRNA-PE promotes HCC cell migration and invasion

To explore the role of lncRNA-PE in cell migration and invasion, we hypothesized that there would be a loss-of-function effect if this upregulated lncRNA was inhibited by silencing RNAs (siRNAs). As shown in Figure 1(c), lncRNA-PE levels were significantly decreased in SK-Hep-1 cells transfected with siRNA against lncRNA-PE (Figure 1(c)). Moreover, downregulation of lncRNA-PE significantly inhibited SK-Hep-1 cell migration and invasion (Figure 2(a) and (b)), whereas overexpression of lncRNA-PE in BEL-7402 cells was found to have the opposite effect (Figure S2). These data suggest that lncRNA-PE can promote cell motility and invasion in HCC.

OPEN IN VIEWER

Figure 2.

LncRNA-PE knockdown decreased hepatocellular carcinoma cell invasion and migration. (a) Wound-healing assay used to detect cell migration changes in hepatocellular carcinoma cells. Decreased cell migration was observed in SK-Hep-1 cells after knocking down lncRNA-PE with siRNAs for 2 days, and columns in each groups were normalized with the column of 0 h. (b) Cell migration and invasion assays. For the cell migration assay, SK-Hep-1 cells were transfected with lncRNA-PE siRNAs and normal control RNA and then subjected to Transwell migration assays as described. After 12 h, migrating cells were counted after staining with crystal violet. Migrating cell numbers are the average count of three random microscopic fields. Each bar represents the mean ± SD of the counts from a single representative experiment (*p < 0.05). For the cell invasion assay, cells transfected with lncRNA-PE siRNAs and normal control RNA were subjected to Transwell invasion assays as described. After incubation for 24 h, cells that invaded through the pores to the bottom surface of the membrane were fixed, stained, and counted. Invading cell numbers are the average count of three random microscopic fields. Each bar represents the mean ± SD of the counts from a single representative experiment (*p < 0.05).

LncRNA-PE induces EMT of HCC cells

EMT has been proved to be important in the early events of tumor cell metastatic dissemination by endowing cells with more motility and potential for invasiveness. We therefore investigated whether lncRNA-PE regulates EMT. The expression levels of the epithelial marker (E-cadherin and ZO-1) and the mesenchymal marker (N-cadherin and vimentin) were analyzed by western blot. The results revealed that knockdown of lncRNA-PE reduced N-cadherin and vimentin, whereas induced E-cadherin and ZO-1 at the protein levels (Figure 3(a)). Consistent with this, overexpression of lncRNA-PE had the opposite effects (Figure 3(b)).

OPEN IN VIEWER

Figure 3.

Effects of lncRNA-PE knockdown and overexpression on EMT markers in vitro. (a) The protein levels of four EMT markers after lncRNA-PE siRNAs were transfected into SK-Hep-1 cells. (b) The protein levels of four EMT markers after lncRNA-PE were overexpressed into BEL-7402 cells.

LncRNA-PE induces EMT via the miR-200a/b-ZEB1 pathway

We hope to seek the mechanistic insight into the role of lncRNA-PE in regulating HCC cell motility and EMT. As miR-200a/b can downregulate ZEB1, ⁶ an important EMT trigger, we wonder whether this lncRNA took part in miR-200-ZEB1 pathway. Therefore, we applied siRNA knockdown of lncRNA-PE in SK-Hep-1 cells and found that miR-200a and miR-200b were upregulated, with no influence on miR-451a (Figure 4(a)), with the opposite effects being caused by overexpression of lncRNA-PE (Figure S4(A)). Continuing our investigation of the molecular mechanism by which lncRNA-PE regulated miR-200, we tested whether lncRNA-PE regulated miR-200 via a competing endogenous RNA (ceRNA) mechanism. We analyzed the distribution of lncRNA-PE in the cell, and it was localized both in the nucleus and the cytoplasm (Figure S3(A)), which indicated that it may have the ability to act as an miRNA sponge. The ability to bind with protein AGO2, a key component of the miRNA-containing RNA-induced silencing complex (RISC), is a distinctive feature of ceRNA. ¹⁸ Therefore, we conducted an RNA IP experiment with AGO2 and proved that lncRNA-PE has poor affinity with AGO2 (Figure S3(B)), ruling out the possibility that lncRNA-PE utilized a ceRNA mechanism, and must regulate miR-200 via other mechanisms. We hypothesized that knockdown of lncRNA-PE could either induce the transcription of miR-200 or enhance its stability. We performed a dual-luciferase-reporter assay to ascertain whether the miR-200a/b upregulation observed after knockdown of lncRNA-PE depends upon regulation of the miR-200a/b promoter to enhance its transcription. pGL3-enhancer luciferase reporters containing the miR-200a/b promoter (pGL3-enhancer-miR-200a/b promoter) were constructed, and luciferase plasmids (pGL3-enhancer-miR-200b promoter or the control reporter pGL3-enhancer) were transfected into the different BEL-7402 and SK-Hep-1 cells with either lncRNA-PE or its siRNAs. Knockdown of lncRNA-PE increased the luciferase activity of the pGL3-miR-200b promoter (Figure 4(b)). Reciprocally, overexpression of lncRNA-PE had the opposite effect (Figure S4(B)). Therefore, we draw a preliminary conclusion that lncRNA-PE can inhibit the transcription of miR-200a/b. Next, we tested whether knockdown of lncRNA-PE can alter the expression of primary miR-200b and miR-200a which share the same promoter with miR-200b, and the data showed that the primary miRNA transcript of miR-200b and miR-200a was upregulated when lncRNA-PE was knocked down (Figure 4(a) and (c)), with the opposite effects being caused by overexpression of lncRNA-PE (Figure S4(A) and (C)). Additionally, both the RNA and protein level of ZEB1 were downregulated by siRNA against lncRNA-PE (Figure 4(d) and (e)), and overexpression of lncRNA-PE had the opposite effect (Figure S4(D)–(E)). To test whether lncRNA-PE regulated ZEB1 through miR-200a/b, we overexpressed lncRNA-PE in hepatoma cells; in the meantime, miR-200a/b mimics were transfected to compensate its effects. Results shown that both miR-200a and miR-200b mimics successfully downregulated ZEB1 (Figure S4(F)) and could compensate lncRNA-PE’s effects on ZEB1 (Figure S4(G)), which verified that lncRNA-PE regulated ZEB1 through miR-200a/b.

OPEN IN VIEWER

Figure 4.

LncRNA-PE knockdown can alter EMT via the miR-200b-ZEB1 pathway. (a) The miRNA levels of miR-200b, miR-200a, and miR-451a after lncRNA-PE siRNAs were transfected into SK-Hep-1 cells. (b) Luciferase assays indicated that lncRNA-PE regulated the expression of miR-200b via its promoter. (c) The mRNA level of pri-miR-200b after lncRNA-PE siRNAs was transfected into SK-Hep-1 cells. (d) The mRNA level of ZEB1 after lncRNA-PE siRNAs was transfected into SK-Hep-1 cells. (e) The protein levels of ZEB1 after lncRNA-PE siRNAs were transfected into SK-Hep-1 cells. Each bar represents the mean ± SD of the counts from a single representative experiment (*p < 0.05).