INTRODUCTION

The incidence of liver cancer ranks sixth among all cancers, and the death rate ranks third in cancer-related death for both sexes worldwide [1]. Uncontrolled tumour metastasis, frequent intrahepatic propagation, and extrahepatic progression are the main causes for the poor prognosis of liver cancer. Although surgical treatment and liver-protecting drugs have made great progress over the past 20 years, the survival of liver cancer patients is largely threatened by cancer recurrence and cancer metastasis [2–4]. The development of liver cancer stems from a variety of abnormal biological processes, such as genetic mutations, epigenetic changes, and abnormal regulation of coding or non-coding genes [5]. Therefore, better understanding of the mechanisms underlying liver cancer development and metastasis may help to identify potential targets for novel therapeutic interventions.

Epithelial-to-mesenchymal transition (EMT) is a dynamic and reversible phenotypic switching process in which epithelial cells lose their polarity and undergo transition into a motile mesenchymal phenotype [6]. The EMT process has been linked as a driver of metastatic dissemination by conferring migratory and invasive capacity to cancer cells. Thus, EMT is widely thought to be a critical switch for tumour cell invasiveness. The important hallmarks of EMT include the loss of E-cadherin expression and increased expression of non-epithelial cadherins, such as N-cadherin and vimentin [7]. E-cadherin, a single-membrane-spanning protein and E-cadherin-based adherens junctions are obligatory for the establishment of polarised epithelia [8]. Decreased expression of E-cadherin causes loss of cell adhesion and contributes to cell dissociation, increased motility, and invasion. EMT-inducing transcription factors, mainly represented by the SNAIL, TWIST, and ZEB protein families, regulate the induction of EMT by repressing the transcription of epithelial cell-specific genes while activating mesenchymal cell-specific ones [9]. However, the exact mechanisms are still in debate [10]. Several zinc-finger transcription factors, including Snail1, Snail2, ZEB1, and ZEB2, directly bind to the E-boxes in the E-cadherin promoter to repress its transcription [11–14].

Clinical and epidemiological studies have shown that matrix metalloproteinases (MMPs) play a significant role in tumour invasion, neoangiogenesis, and metastasis, and therefore they represent ideal targets for cancer therapy [15]. MMPs are zinc-dependent proteolytic enzymes that degrade the ECM and basement membrane during invasion and migration [16, 17]. The basement membrane separates the epithelial and mesenchymal compartments, representing the first barrier to invasion. MMP2 and MMP9 both degrade type IV collagen in the basement membrane and thus help in cancer cell invasion [18].

Apoptosis is a critically important biological process that plays an essential role in cell fate and homeostasis. Previous studies indicated that apoptosis in liver cancer is induced by microRNAs targeting mitochondrial family proteins, death receptor-mediated pathways, and in a drug-induced autophagy manner [19–21]. Mitochondria play a major role in the intrinsic apoptotic pathway whereby stress signals converge on mitochondria to induce mitochondrial outer membrane permeabilisation, which is controlled by pro-apoptotic and anti-apoptotic members of the Bcl-2 family [22]. Anti-apoptotic proteins include Bcl-2, Mcl-1, Bcl-XL, Bcl-w, and Bfl-1/A1 while the pro-apoptotic proteins include Bax, Bak, Bim, Bid, and Puma [23]. Cells deficient in both Bax and Bak are extremely resistant to mitochondria-mediated apoptosis, and thus one of these proteins is thought to be necessary for mitochondrial outer membrane permeabilisation [24].

Hox transcript antisense RNA (HOTAIR) is a long noncoding RNA that localises in the nucleus and regulates gene expression by altering chromatin status and epigenetic modifications [25–27]. The expression of HOTAIR is up-regulated in a variety of tumours, and its upregulation is significantly correlated with poor clinical outcomes in breast, colon, and non-small cell lung cancers [28–31]. Previous studies also showed that HOTAIR is an important molecule in regulating tumour invasion and metastasis. In addition, a recent review [32] showed that HOTAIR is a potent regulator of mitochondrial-induced apoptosis. Multivariate analysis revealed that HOTAIR is an independent prognostic factor for predicting the overall survival of liver cancer patients, which indicates that HOTAIR might serve as a potential biomarker for liver cancer progression [33, 34]. Therefore, elucidating the molecular mechanisms of how HOTAIR regulates the proliferation, invasion, migration, and apoptosis may help in using it as a predictive biomarker and lead to the development of novel therapeutic approaches.

In this study, we emphasised the importance of HOTAIR as an independent indicator in predicting overall survival of HCC patients. Our results revealed that the HOTAIR/Snail2/ /E-cadherin signal pathway plays a role in the invasion and migration of liver cancer cells. Also, HOTAIR depletion induces mitochondrial-induced apoptosis in liver cancer cells.

Materials and methods

Patients

Hepatocellular carcinoma patients (n = 373) who underwent surgery were examined in this study. Samples, of hepatocellular cancer tissues, and normal liver tissues, were taken from uninvolved matched fragments from the same patients. Data were analysed on the basis of the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/). Clinical-pathological data of patients included values for age, sex, histologic grade, stage, TNM classification, residual tumour, vital status, and HOTAIR expression are provided in Table 1.

Cell culture

Liver cancer cell lines including HepG2, SNU-387, SMMC97L, SMMC97H, and normal hepatocyte cell line Chang liver cells were purchased from the American Type Culture Collection (Manassas, VA, USA), and HCCLM3 was purchased from the Liver Cancer Institute of Fudan University (Shanghai, China). We also provided a statement of authentication (STR profiling) of HepG2 and SNU-387 cells to confirm that there is no contamination of this cell line (available on reasonable request from the corresponding author). These cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Carlsbad, CA, USA) supplemented with 10% foetal bovine serum (FBS, Gibco) and incubated at 37°C in a humidified environment containing 5% CO2. All the cell lines were maintained in the recommended culture conditions.

Total RNA extraction, reverse transcription, and quantitative real-time polymerase chain reaction (RT-qPCR)

The total RNA was extracted by Trizol reagent (Invitrogen, Waltham, MA, USA), according to the manufacturer’s protocol, and transcribed into cDNA using EasyScript One-Step gDNA Removal and cDNA Synthesis SuperMix (Transgen, Beijing, China, Cat: AE311). The expression levels of the mRNAs were determined by RT-qPCR using SYBR Green (Bimake, Beijing, China) in a real-time PCR detection system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). All data represent the average of 3 repeated experiments. The primers used are listed in Table 2.

Cytoplasmic and nuclear protein extraction

We used the MinuteTM Cytoplasmic and Nuclear Extraction Kit (Invent Biotechnologies, Shanghai, China) for cell protein extraction following the manufacturer’s protocol.

Western blotting and antibodies

Cells were harvested and incubated in RIPA lysis buffer (Beyotime, Shanghai, China) for 45 min at 4°C. The concentration of proteins was measured using the Coomassie G250 assay (Beyotime Biotechnology). Equal amounts of protein lysates per well were separated by SDS-PAGE and transferred to PVDF membrane (Millipore, Billerica, MA, USA), and then they were blocked in 5% non-fat dry milk for 2 h at room temperature. After incubating with various primary antibodies overnight at 4°C, membranes were incubated with horseradish-peroxidase-conjugated secondary antibodies at room temperature for 2 h. Membranes were then incubated in ECL reagents (Thermo Scientific, Waltham, MA, USA), and images were captured by Syngene BioImaging (Synoptics, Cambridge, UK). βeta-actin was used as a control. Antibodies against E-cadherin (1:1000), N-cadherin (1:1000), Bcl-2 (1:1000), Mcl-1 (1:1000), Bak (1:1000), Bid (1:1000), and cleaved-caspase3 (1:1000) were purchased from Proteintech (Chicago, IL, USA). Antibodies against Snail2 (1:1000), MMP2 (1:1000), MMP9 (1:1000) were purchased from Cell Signaling Technology (Boston, MA, USA).

Knockdown and overexpression of HOTAIR

Short hairpin RNA sequences targeting human HOTAIR (sh-HOTAIR) and a non-target sequence (sh-NC) were constructed by GenePharma (Shanghai, China). The sequence of HOTAIR shRNA is 5’-GAACGGGAGTACAGAGAGA-3’. The sequence of non-target sequence (sh-NC) is 5’-TTCTCCGAACGTGTCACGT-3’. A full-length human HOTAIR expression vector was purchased from Addgene (LZRS-HOTAIR, #26110; https://www.addgene.org). The negative control for LZRS-HOTAIR is LZRS vector.

Real-time cell analysis

The real-time cell analysis (RTCA) S16 System (ACEA Biosciences, San Diego, CA, USA) was used to monitor cellular proliferation status. Briefly, cells were seeded into a special 16-well electronic plate (16-E-Plate). Then, the plate was placed into a special station and connected to an electronic sensor analyser by electrical cables. Then the station was placed in a CO2 culturing incubator. The more cells on the electrodes, the greater the change in electrode impedance. A unitless parameter termed Cell Index is used to measure the relative change in electrical impedance to represent cell status. The accompanying software was used to carry out the dimensionless impedance-based Cell Index. The Cell Index value changes with time and reflects the number of cells inside the well.

Flow cytometry

Flow cytometry analysis was performed to identify and quantify the apoptotic cells. The assay was done by using an Annexin V-FITC/PI apoptosis detection kit (BD Biosciences; San Diego, CA, USA, Cat: 556547) according to the manufacturer’s protocol. The cells (105 cells/well) were seeded in 6-well plates. The cells were washed twice with cold phosphate-buffered saline (PBS) and then resuspended in 100 μL binding buffer and 5 μL Annexin V-FITC solution, and 5 μL PI were added to the mixture, which was then co-incubated for 15 min at room temperature in the dark. Finally, flow cytometry (Millipore Guava EasyCyte Cytometer) was conducted to differentiate apoptotic cells.

The measurement of mitochondrial membrane potential (MMP) was analysed using a Mitochondrial Membrane Potential Assay Kit with JC-1 (Beyotime Biotechnology). Briefly, cells were stained with JC-1 for 20 minutes at 37°C, protected from light. Then the cells were washed and resuspended in 1× staining buffer and subsequently analysed using a Guava Easycyte flow cytometer (Guava Easycyte, Millipore).

Cell transfection

Cell transfection was conducted using TurboFect Transfection Reagent (Thermo Scientific, Cat: R0531). 400 µL serum-free culture medium was put in a 1.5 mL tube, 4 µg of plasmid was added and mixed by gently pipetting. Then 6 µL of transfection reagent were added and pipetted gently to mix. Finally, the test tube was left at room temperature for 20 min, and the mixer including medium and plasmid were added into the plate. The RNA samples were harvested after 24 h and protein samples after 48 h.

Cell migration and invasion assays

Cell invasion ability was detected using 24-well chemotaxis chambers (Corning, Ithaca, NY, USA, Cat: 3422). The cells were washed twice with PBS, resuspended in 100 μL of serum-free medium, and added to the upper chambers. The lower chambers were filled with 600 μL medium containing 20% foetal bovine serum (FBS). The cells were incubated for 24 h in the upper chamber coated with a mixture of serum-free medium and Matrigel (50:1; BD Biosciences, Cat: 356234).

Scratch assay

Briefly, the cells were allowed to reach 90% confluence for 48 h and then were serum deprived for 6 h. Later, a denuded area was produced by scratching the inside diameter of the well with a 10-µL pipette tip, and the wells were then washed with PBS. Subsequently, we used an LX71 inverted microscope (Olympus, Tokyo, Japan) to take a photograph of the scratch at 0 h and then again at the same position after 48 h, comparing the changes in the healing of the scratches between the transfected and untransfected groups.

Immunofluorescence

After cells were grown on 6-well plates to confluency and transfected using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instruction, cells were seeded on sterile coverslips and cultured for 48 h. The cells were then fixed with 4% paraformaldehyde for 20 min and permeabilised with 0.25% Triton-X 100 for 10 min, followed by blocking with 3% bovine serum albumin for 1 h. Immunofluorescence staining was conducted with antibodies against Snail2 (1:200, Proteintech, Chicago, IL, USA). Hoechst 33342 (10 µg/mL, Sigma) was used to stain the nucleus. Cells were examined under an Olympus FV1000 confocal laser microscope. Images were taken under a 50× lens.

Colony formation assays

Cells were seeded in 12-well plates at a concentration of 2 × 102 cells per well. Next, they were cultivated in DMEM/1640 with 10% FBS for about 2 weeks. During this period, we replaced medium every 4 days. After colonies were fixed by using 4% paraformaldehyde, we stained them with 0.2% crystal violet (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 15 min. Finally, the number of colonies was manually counted.

Bioinformatical methods

All statistical analyses and visualisations were carried out using R (version 3.5.1). RNA-seq data, clinical information, and phenotype data, including 373 tumours and 50 matched normal samples, were downloaded from TCGA-LIHC data of UCSC Xena (https://xenabrowser.net/datapages/). Box plots were used to evaluate the expression of HOTAIR in the TCGA-Liver Hepatocellular Carcinoma (LIHC) dataset. A receiver operating characteristic curve (ROC) was drawn to assess the diagnostic significance of HOTAIR expression by using the pROC package [35]. UCSC Genome Browser (http://genome.ucsc.edu/cgi-bin/hgGateway) was used to analyse promoter regions of Snail2.

Statistical analysis

We calculated the relationship between HOTAIR and E-cadherin (CDH1) as well as HOTAIR and Snail2 through Pearson’s test using R (version 3.5.1). Kaplan–Meier survival curves (with the log-rank test) were constructed to explore the prognostic value of HOTAIR in overall survival by using the Survival package in R (version 3.5.1). In accordance with the threshold identified by the ROC curve, which was used to perform ROC analysis, the optimal HOTAIR cut-off point was determined (3.225), and then liver cancer patients were separated into low and high expression groups. The ROC curve image can be found in the supplementary material (Suppl. Fig. 1). All data were displayed as the mean ± SD for 3 independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.001 were considered statistically significant. Data were analysed by GraphPadPrism Version 6.0 software (GraphPad Software Inc., San Diego, CA, USA). The ANOVA test was applied for the comparison of multiple groups, as seen in Fig. 1B, and Dunnett’s test was performed as the post hoc test.

Figure 1. HOTAIR expression in liver cancer tissues and cells; A. The differential expression of HOTAIR between primary liver cancer tissues and normal liver tissues. Normal liver tissues were taken from uninvolved matched fragments from the same patients. Data were analysed on the basis of the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/); B. Expression of HOTAIR in the normal hepatocyte cell line Chang liver cells and different liver cancer cell lines was evaluated by RT-qPCR and normalised to beta-actin gene expression; C. Kaplan–Meier analysis showed that patients with tumours with higher HOTAIR expression had a poor prognosis; D. Expression of HOTAIR in SNU-387 cells after transfection with LZRS-HOTAIR plasmid. E. Expression of HOTAIR in SNU-387 and HepG2 cells after transfection with sh-HOTAIR plasmid. The data shown were representative of 3 independent experiments. Bars represent means and whiskers — standard deviation SD, n = 3. ns: no statistical significance, **P < 0.01, and ***P < 0.001 vs. sh-NC/control group.

Results

HOTAIR expression levels in liver cancer tissues and cells

The analysis of HOTAIR was conducted to compare the difference in HOTAIR expression between primary liver cancer tissues and normal liver tissue via box plots. The results showed that HOTAIR mRNA was upregulated in liver cancer tissues (P = 0.005, Fig. 1A). Kaplan–Meier survival curves showed that high expression of HOTAIR was related to worse overall survival (P = 0.021, Fig. 1C). We then used RT-qPCR to detect the expression of HOTAIR in 5 liver cancer cell lines (HepG2, SNU-387, HCCLM3, SMMC97L, and SMMC97H) along with normal hepatocyte cell line Chang liver cells. The expression of HOTAIR in liver cancer cells was significantly higher compared with that in normal Chang liver cells (P < 0.01, Fig. 1B). We selected HepG2 and SNU-387 cells, which showed higher expression of HOTAIR, for subsequent experiments. Finally, we used HOTAIR overexpression (LZRS-HOTAIR) and knockdown (sh-HOTAIR) plasmids and then tested the transfection efficiency separately by RT-qPCR. The results showed that the transfection efficiency could meet the experimental needs (Fig. 1D, E).

HOTAIR promotes the proliferation of liver cancer cells

To examine the effects of HOTAIR on the growth of liver cancer cells, we modulated HOTAIR expression in SNU-387 and HepG2 cells using HOTAIR overexpression (LZRS-HOTAIR) and knockdown (sh-HOTAIR) plasmids. We confirmed overexpression and knockdown of HOTAIR using these plasmids (Fig. 1D, E). Real-time cell analysis platform results showed that the proliferation rate of SNU-387 cells in the HOTAIR overexpression group was significantly accelerated and the doubling time was shortened compared with the controls (Fig. 2A). In the HOTAIR knockdown group, the proliferation rate of both SNU-387 and HepG2 cells was decreased and the doubling time was prolonged (Fig. 2B, C) with a much greater effect in HepG2 cells.

We performed colony formation assay to investigate the in vitro tumourigenesis ability in SNU-387 and HepG2 cells with HOTAIR upregulation and downregulation. The results showed that the colony-forming ability of SNU-387 cells in the LZRS-HOTAIR group was enhanced compared with controls (Fig. 2D), while the colony-forming ability of SNU-387 and HepG2 cells in the sh-HOTAIR knockdown group was reduced compared with controls (Fig. 2E, F). Together, these results indicate that HOTAIR promotes the proliferation of liver cancer cells in vitro.

Figure 2. HOTAIR promotes proliferation of liver cancer cells. Cell proliferation was determined by real-time cell analysis (RTCA) for real-time monitoring every 5 min. The number of cells inside the well was displayed as Cell Index. The arrow indicates the point of plasmid transfection; A. SNU-387 cells were transfected with LZRS-HOTAIR plasmid; B, C. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid. The effect of HOTAIR on the doubling time was calculated by the RTCA system. D, F. Colony formation assay was performed, and colonies were counted and imaged (magnification: 50×). The data shown were representative of 3 independent experiments. Bars represent means and whiskers standard deviation, n = 3. **P < 0.01, and ***P < 0.001 vs. control/sh-NC group.

HOTAIR promotes invasion and migration of liver cancer cells

To investigate the role of HOTAIR in regulating the invasion and migration of liver cancer cells, we used transwell chambers to detect the invasion and migration abilities of SNU-387 and HepG2 cells in the LZRS-HOTAIR and sh-HOTAIR groups. Both the cell invasion and migration abilities were enhanced after overexpression of HOTAIR in SNU-387 cells (Fig. 3A). In contrast, after depletion of HOTAIR in SNU-387 and HepG2 cells, the invasion and migration abilities of SNU-387 (Fig. 3B) and HepG2 cells (Fig. 3C) were decreased. We also performed scratch assays and confirmed that the migration ability of cells was significantly increased after overexpression of HOTAIR in SNU-387 cells (Fig. 3D). Knockdown of HOTAIR in SNU-387 and HepG2 cells led to decreased migration ability of SNU-387 (Fig. 3E) and HepG2 cells (Fig. 3F). Together, these data demonstrated that HOTAIR contributes to the invasion and migration of liver cancer cells.

Figure 3. HOTAIR promotes invasion and migration of liver cancer cells. Cell invasion and migration were examined by transwell assay and wound healing assay; A, D. SNU-387 cells were transfected with LZRS-HOTAIR plasmid after 36 h. Transwell assay and wound healing assay indicated that LZRS-HOTAIR increased invasion and migration of SNU-387 cells. B, C, E, F. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid after 36 h. Transwell assay and wound healing assay showed that sh-HOTAIR decreased SNU-387 and HepG2 cells invasion and migration. Bars represent means and whiskers standard deviation, n = 3. **P < 0.01, and ***P < 0.001 vs. control/sh-NC group. Magnification of all microphotographs: 100×.

HOTAIR promotes invasion and migration by regulating the nuclear localisation of Snail2 in liver cancer cells

To examine the possible mechanisms through which HOTAIR promotes invasion and migration in liver cancer cells, we used western blot and RT-qPCR to examine changes in EMT-related molecules at the protein and transcriptional levels. Upon overexpression of HOTAIR in SNU-387 cells, the EMT epithelial marker E-cadherin protein level was down-regulated, and the mesenchymal markers N-cadherin and Snail2 expression levels were up-regulated. Importantly, MMP2 and MMP9 protein levels were also increased, especially MMP9, compared with the CON group (Fig. 4A, 5A). The RT-qPCR results were mostly consistent with the western blot data, except for N-cadherin, the protein level of which was significantly up-regulated whereas N-cadherin mRNA levels were not appreciably up-regulated (Fig. 4D).

Figure 4. HOTAIR participates in epithelial-mesenchymal transition (EMT) and promotes invasion and migration by regulating the nuclear localisation of Snail2 and the expression of metalloproteinase 2 (MMP2) and MMP9 in liver cancer cells. Western blot and RT-qPCR were applied to analyse expression of E-cadherin, N-cadherin, Snail2, MMP2, and MMP9 protein and mRNA levels, respectively; A. SNU-387 cells were transfected with LZRS-HOTAIR plasmid. Representative western blots of E-cadherin, N-cadherin, and Snail2 expression; B, C. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid. Western blots document the protein expression of E-cadherin, N-cadherin, and Snail2; D. SNU-387 cells were transfected with LZRS-HOTAIR plasmid. RT-qPCR was used to detect mRNA expression of E-cadherin, N-cadherin, Snail2, MMP2, and MMP9. β-actin was used as a loading control for RNA and total protein levels; E, F. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid. mRNA expression of E-cadherin, N-cadherin, Snail2, MMP2, and MMP9 were determined by RT-qPCR; G, J. SNU-387 cells were transfected with LZRS-HOTAIR plasmid. Snail2 and laminAC nuclear localisation was examined by western blot and immunofluorescence (magnification: 200×); H, I, K, L. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid. Snail2 and laminAC nuclear localisation were examined by western blot and immunofluorescence (magnification: 200×). The data shown were representative of 3 independent experiments. Bars, SD (n = 3), **P < 0.01, and ***P < 0.001 vs. control/sh-NC group.

Consistent with the overexpression results, upon knocking down HOTAIR in SNU-387 and HepG2 cells, the E-cadherin protein level was up-regulated, and the mesenchymal markers N-cadherin and Snail2 expression levels were both decreased. MMP2 and MMP9 protein levels were also reduced compared with the sh-NC group (Fig. 4B, C, 5B, C). The RT-qPCR data were mostly consistent with the western blot results except for N-cadherin in HepG2 cells and Snail2 in SNU-387 cells (Fig. 4E, F). These results showed that overexpression of HOTAIR promoted EMT in SNU-387 cells, while EMT in SNU-387 and HepG2 cells was reversed to some extent after HOTAIR knockdown.

We next examined the eFxpression of Snail2 in the nuclear fractions of HOTAIR overexpression or knockdown cells. Upon overexpression of HOTAIR in SNU-387 cells, the nuclear expression level of Snail2 was up-regulated (Fig. 4G). In contrast, the nuclear expression level of Snail2 was reduced in both SNU-387 and HepG2 cells after knockdown of HOTAIR (Fig. 4H, I). We also used immunofluorescence experiments with co-localisation of the target protein Snail2 (red) and the nucleus of Hoechst-stained cells (blue) to assist in observing changes in the nuclear localisation of Snail2, and the results were consistent with western blot data (Fig. 4J–L). Taken together, our results suggest that HOTAIR may promote the degradation of extracellular matrix by enhancing the expression of MMP2 and MMP9 and repress the transcription of E-cadherin by enhancing the nuclear localisation of Snail2, thereby promoting invasion and migration of liver cancer cells.

Figure 5. The histogram shows the relative density of the bands with respect to β-actin in Figure 4A–C.

HOTAIR knockdown promotes mitochondria- -induced apoptosis in liver cancer cells

To explore the effects of HOTAIR on apoptosis, we examined apoptosis using Annexin V-FITC/PI staining and western blotting. The rates of apoptosis in SNU-387 cells (Fig. 6A) and HepG2 cells (Fig. 6B) were increased after transfection of sh-HOTAIR for 48 h. We also used JC-1 fluorescent staining to measure the MMP. The results showed that depletion of HOTAIR further induced the decrease in MMP of SNU-387 (Fig. 6C) and HepG2 (Fig. 6D) cells. The sh-HOTAIR plasmid was transfected into SNU-387 and HepG2 cells by liposome transfection. Forty-eight hours later, the cells were fixed and the nuclei of SNU-387 and HepG2 cell lines transfected with the sh-HOTAIR plasmid were stained with Hoechst33342. The results showed that the nuclei of the cells had an abnormal morphology, with chromatin condensation, nuclear condensation, and even nuclear division, as compared with the control group (Fig. 6E, F). We also found that the expressions of anti-apoptotic protein Bcl-2 and Mcl-1 were both decreased and the expression level of the pro-apoptotic protein Bak and Bid was up-regulated in response to HOTAIR depletion. In addition, HOTAIR knockdown induced the expression of cleaved caspase 3, which partially explained the HOTAIR-mediated effects on liver cancer cell apoptosis (Fig. 6G, H, 7A, B).

Figure 6. HOTAIR knockdown promotes mitochondria-induced apoptosis in liver cancer cells. The control and experimental liver cells were treated as described in Methods and analysed after transfection; A, B. SNU-387 and HepG2 cells were transfected with sh-HOTAIR plasmid for 36 h, apoptosis was assessed by staining for Annexin V/PI and analysed by flow cytometry; C, D. The mitochondrial membrane potential (MMP) was assessed by JC-1 staining after transfection with sh-HOTAIR plasmid in SNU-387 and HepG2 cells for 48 h; E, F. Cell morphology was observed by confocal microscopy 48 h after transfection with sh-HOTAIR plasmid in SNU-387 and HepG2 cells, arrows represent apoptotic cell nuclei (magnification: 100×); G, F. The expression levels of Bak, Cleaved-caspase3, Bcl2, Bid, and MCL-1 proteins were analysed by western blotting after transfection with sh-HOTAIR plasmid of SNU-387 and HepG2 cells for 48 h. The data shown were representative of 3 independent experiments.

Figure 7. The histogram shows the relative density of the bands with respect to β-actin in Figure 6G, F.

Discussion

An increasing number of studies have provided evidence that non-coding RNAs are key factors involved in gene regulation that influence normal and cancer cell phenotypes [36, 37], as well as having an important function in cancer pathogenesis. Our results showed that HOTAIR was overexpressed in hepatocyte cell lines. Whereas in clinical cases, HOTAIR expression was low in hepatocellular carcinoma cases, we hypothesise that there may be individual differences due to the insufficient and unrepresentative size of this selected sample. Many previous studies have shown that HOTAIR expression is upregulated in many cancer types including hepatocellular carcinoma. And in this study, we found that the expression of HOTAIR was significantly up-regulated in the hepatocellular carcinoma cell lines HepG2, SNU-387, HCCLM3, SMMC97, and SMMC97H, which is consistent with the results of many previous studies. Therefore, we can still reasonably infer that high levels of HOTAIR are closely associated with the progression of hepatocellular carcinoma, suggesting that HOTAIR inhibition may be an effective therapeutic strategy for hepatocellular carcinoma. Related studies have confirmed that HOTAIR promotes the proliferation, invasion, and metastasis of liver cancer cells through the ceRNA competition mechanism [38]. Also, it has been reported that HOTAIR functions as a miR-23b-3p sponge to positively regulate ZEB1 [39], a zinc-finger transcription factor [39] that is predominantly expressed in the stroma of several tumours [40]. These results indicate that the regulation of HOTAIR may be related to the zinc-finger transcription factor family.

Previous studies showed that Snail2 is up-regulated in lung cancer tissues compared with non-cancerous tissues and is indispensable for EMT [41]. The Snail2 transcription factor regulates the transcription of E-cadherin through various mechanisms. Snail2 directly binds to the E-boxes of the E-cadherin promoter to repress its transcription. Snail2 also participates in the regulation of E-cadherin by Snail2/G9a/HDACs axis [42]. In addition, Snail2 interacts with HDAC6 and then recruits HDAC6 and PRC2 to the promoter of E-cadherin to inhibit the expression of E-cadherin in colorectal cancer [43]. Based on these studies, we tried to interfere with the nuclear localisation of Snail2 to reduce or block the regulation of E-cadherin regulated by Snail2. When the nuclear localisation of Snail2 increases, binding of Snail2 to the E-boxes of target gene promoters also increases, resulting in repressed transcription of downstream target genes, such as E-cadherin, which is crucial in maintaining epithelial polarity [44]. Our results indicate that HOTAIR may promote invasion and migration by regulating the nuclear localisation of Snail2. MMP2 and MMP9 are 2 important members of the MMP family and are key factors in hepatocellular carcinoma (HCC) cell invasion and metastasis [45, 46]. Even though our results confirmed that HOTAIR regulated the expression of MMP2 and MMP9, whether this regulation is direct or indirect requires further experimental identification.

We examined the correlation between HOTAIR and E-cadherin (CDH1) by analysing information of liver cancer patients in the TCGA database but did not find a significant correlation (Suppl. Fig. 2A). E-cadherin is synthesised as a precursor and then undergoes cleavage by proprotein convertases, which is essential for E-cadherin maturation and cell adhesion [8]. Importantly, EMT transcriptional factors also cooperate with various enzymes to repress the expression of E-cadherin and regulate EMT at the epigenetic and post-translational level [47]. Thus, we hypothesised that HOTAIR may also regulate E-cadherin expression at the level of post-translational modifications, but the specific mechanism remains to be identified.

We also investigated the correlation between HOTAIR and Snail2. To examine the transcriptional regulator of Snail2, we used ENCODE Histone Modification Tracks in the UCSC Genome Browser and found H3K4me3 and H3K27ac enrichment peaks in the Snail2 promoter region (Suppl. Fig. 3). These results indicate that the transcriptional regulation of Snail2 is a complex process regulated by multiple genes and multiple factors. When the western blot data were applied to a large liver cancer patient database for comparison, we found no obvious correlation between HOTAIR and Snail2 (Suppl. Fig. 2B).

While the changes in N-cadherin protein levels with HOTAIR modulation were consistent with the effects of HOTAIR on EMT, the N-cadherin gene levels were not consistent with the protein level. The possible reason for this might be that cadherins typically undergo post-translational regulation through processing, trafficking, or stabilisation [48]. For example, N-cadherin levels in chick trunk neural crest cells are regulated by these mechanisms prior to EMT [49]. These findings indicate that the N-cadherin protein level may not always be consistent with the transcriptional level of N-cadherin.

Several reports showed that apoptosis in liver cancer cells is induced by microRNAs targeting mitochondrial proteins, death receptor-mediated pathways, and in a drug- -induced autophagy manner [19–21]. For example, some drugs can induce apoptosis in liver cancer cells, like metformin, glibenclamide, and chrysin [50–52]. Our data indicate that the mitochondrial-related cell death pathway was involved in a HOTAIR-mediated apoptosis. After HOTAIR knockdown, the mitochondrial membrane potential was decreased and the expression of the mitochondrial pro-apoptotic protein Bak was up-regulated, while the expressions of mitochondrial anti-apoptotic proteins Bcl2 and Mcl-1 were decreased. These findings suggest the possibility for treatment of liver cancer by targeting the HOTAIR-mediated mitochondria apoptosis pathway.

Conclusions

In summary, we demonstrated a role for HOTAIR in promoting proliferation, invasion, and migration and inhibiting mitochondrial-related apoptosis in liver cancer. These findings contribute to a better understanding of the importance of dysregulated HOTAIR in liver cancer progression. Our results also suggest HOTAIR as a novel target for the treatment of liver cancer and provide a rationale for the potential development of long non-coding RNA-based targeted approaches for the treatment of liver cancer.

However, the way how HOTAIR regulates Snail2 needs to be discussed in the future, and we did not utilise the HOTAIR expression and clinical data to construct a predictive model. The related molecular mechanisms and multiomics of HOTAIR still need to be elucidated in future studies.

In conclusion, our results provide strong evidence that HOTAIR plays critical roles in promoting proliferation, invasion, and migration and inhibiting mitochondrial-related apoptosis in liver cancer progression.

Article information and declarations

Data availability statement

Original contributions presented in the study are included in the article, and further inquiries can be directed to the corresponding author.

Ethics statement

This study was approved by Medical Ethics Committee of Taizhou Hospital, and patients/participants provided their written informed consent to participate in this study.

Author contributions

Xinzi Zheng conducted data collection and analysis and contributed to the writing and revision of the manuscript. Renyin Cui contributed to the study design, data interpretation, and manuscript revision. Yan Jiao contributed to the manuscript revision. Dongxia Chu provided critical input in the study design, data interpretation, and manuscript revision. Bingrong Wang assisted in data collection and analysis and contributed to the writing and revision of the manuscript. Na Li served as the corresponding author, overseeing the entire project, providing guidance in all facets of the study, and finalising the manuscript. All authors have reviewed and approved the final version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82173310; 81702793); the Fundamental Research Funds for the Central Universities (No.2132020KJC017); the Norman Bethune Program of Jilin University (2022B24); the Outstanding Young Teachers Training Program of Jilin University.

Conflict of interest

The authors had no conflicts of interest to declare in relation to this article.

Supplementary material

Supplementary material is available on the Journal’s website. This includes:

Supplementary Figure 1. The optimal HOTAIR cut-off point was established based on ROC curve.

Supplementary Figure 2A. The correlation between HOTAIR and E-cadherin (CDH1) in transcriptional level by using TCGA database; B. The correlation between HOTAIR and Snail2 in transcriptional level. Data were downloaded from TCGA database.

Supplementary Figure 3. The epigenetic modification of the promoter of Snail2. Relevant data stems from UCSC Genome Browser database.

This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, allowing to download articles and share them with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially.