Long non‐coding RNA H19 is responsible for the progression of lung adenocarcinoma by mediating methylation‐dependent repression of CDH1 promoter

Abstract Lung adenocarcinoma is a common histologic type of lung cancer with a high death rate globally. Increasing evidence shows that long non‐coding RNA H19 (lncRNA H19) and CDH1 methylation are involved in multiple tumours. Here, we tried to investigate whether lncRNA H19 or CDH1 methylation could affect the development of lung adenocarcinoma. First, lung adenocarcinoma tissues were collected to detect CDH1 methylation. Then, the regulatory mechanisms of lncRNA H19 were detected mainly in concert with the treatment of overexpression of lncRNA H19, siRNA against lncRNA H19, overexpression of CDH1 and demethylating agent A‐5az in lung adenocarcinoma A549 cell. The expression of lncRNA H19 and epithelial‐mesenchymal transition (EMT)‐related factors as well as cell proliferation, sphere‐forming ability, apoptosis, migration and invasion were detected. Finally, we observed xenograft tumour in nude mice so as to ascertain tumorigenicity of lung adenocarcinoma cells. LncRNA H19 and methylation of CDH1 were highly expressed in lung adenocarcinoma tissues. A549 cells with silencing of lncRNA H19, overexpression of CDH1 or reduced CDH1 methylation by demethylating agent 5‐Az had suppressed cell proliferation, sphere‐forming ability, apoptosis, migration and invasion, in addition to inhibited EMT process. Silencing lncRNA H19 could reduce methylation level of CDH1. In vivo, A549 cells with silencing lncRNA H19, overexpression of CDH1 or reduced CDH1 methylation exhibited low tumorigenicity, reflected by the smaller tumour size and lighter tumour weight. Taken together, this study demonstrates that silencing of lncRNA H19 inhibits EMT and proliferation while promoting apoptosis of lung adenocarcinoma cells by inhibiting methylation of CDH1 promoter.


| INTRODUC TI ON
Lung adenocarcinoma, a frequently occurring subtype of lung cancer, contributes to more than one million deaths on an annual basis globally. 1 Lung adenocarcinoma is characterized by a heterogeneous group of tumours stemming from the smaller airways and occurring in peripheral lung tissues. 2,3 Metastasis, considered as a leading lethal cause of lung adenocarcinoma, can occur, only within months of diagnosis, in diverse organs in a swift manner. 4,5 Major therapies for lung adenocarcinoma conclude the use of erlotinib and gefitinib. 6 However, unfortunately, lung adenocarcinoma is highly refractory to conventional radio-and chemotherapies, which poses a challenge to the treatment efficacy. 7 Emerging evidence has reported the implication of long-chain non-coding RNAs (lncRNAs) in the development as well as progression of lung cancer. 8 LncRNA H19, being an affluent and conserved transcript in the development of mammalian, is found in embryonic as well as extra-embryonic cell lineages. 9 Highly expressed lncRNA H19 was found to be responsible for metastasis of lung adenocarcinoma and to exhibit a negative correlation with the prognosis of patients with lung adenocarcinoma. 10 Moreover, a recent study has demonstrated that lncRNA H19 promotes epithelial-mesenchymal transition (EMT), a crucial process of cancer metastasis, by targeting miR-484 in human lung cancer cells. 11,12 CDH1, a cell adhesion molecule also called E-cadherin, plays a critical part in epithelial phenotype maintenance and functions as a gene that can suppress invasion. 13 DNA methylation is considered to be a common event in lung cancer. 14 As previously reported, silencing of CDH1 gene caused by hypermethylation of its promoter aided in the occurrence of lung primary adenocarcinoma. 15 Intriguingly, lncRNA H19 was reported to inhibit the expression of E-cadherin in bladder cancer and tongue squamous cell carcinoma. 16,17 From all the records mentioned above, it has been suggested that lncRNA H19 and CDH1 might be associated with the progression of lung adenocarcinoma. We downloaded lncRNA H19 expression in lung adenocarcinoma from the Cancer Genome Atlas (TCGA) database and found that compared with normal control tissues, lncRNA H19 was highly expressed in tissues of lung adenocarcinoma. Herein, we conducted this study to explore the effects of lncRNA H19 and CDH1 on lung adenocarcinoma with the involvement of methylation of CDH1, with the hope to raise the life quality of patients with lung adenocarcinoma.

| Ethics statement
This research study was conducted with the approval of the ethics committee of the First Hospital of Qinhuangdao. All patients participating in the study were provided with informed consent documentation, which they all subsequently signed. The animal experiment procedures were performed in accordance with strict protocols approved by the Institutional Animal Care and Use Committee. (b) patients who had undergone no cancer-related treatment, such as radiotherapy and chemotherapy. The cancer tissues and adjacent tissues obtained through surgical resection were immediately placed into a liquid nitrogen tank. Next, four siRNA fragment mixtures were heated at 95°C for 5 minutes and then cooled at room temperature for 20 minutes to form double-stranded fragments. The annealed double-stranded siRNA was further diluted into 10 μmol/L concentration, after which the double-stranded siRNA fragments were inserted into the siRNA expression vector pGFP-V-RS using a vector construction kit for connection at room temperature for 30 minutes, and the siRNA expression plasmid was constructed. Finally, the siRNA expression plasmid was converted to the competent cell DH5a, and three clones were selected in each conversion plate, respectively, for gene sequencing in order to verify whether the inserted fragment sequences in the recombinant clones were consistent with the designed siRNA oligonucleotide sequences. CDH1 overexpression vector was constructed as follows: Complete sequence analysis of the coding region of CDH1 was performed to detect whether there was a particularly complicated secondary structure and repetitive sequences in the CDH1 gene. Subsequently, based on the analysis results of gene sequencing, CDH1 single-stranded oligonucleotide fragments were designed and synthesized, and restriction enzyme sites were added to the ends of the CDH1 sequence. PCR was used to splice the synthesized single-stranded oligonucleotide fragments into a complete gene, after which the synthesized sequences were loaded into the pGFP-V-RS vector and converted into the competent cell DH5a. Next, gene sequencing was performed so as to verify whether the inserted fragment sequences in the recombinant clones were consistent with the requirements. The mutation sites in the gene sequences were repaired by overlapping PCR. Full-length fragments after repair then underwent enzyme digestion by Xho I and EcoR and were connected to the target vector and converted into competent cell DH5a. Finally, sequence information of the target gene fragments in the recombinant clones was verified using sequencing to obtain correct CDH1 overexpression vector.

| Vector construction
The siRNA interference vector and CDH1 overexpression vector were transiently cotransfected: Cells were seeded into each well of the 6-well plate (5 × 10 5 cells/well), followed by culturing overnight, with the cell growth observed. When the cell confluence reached approximately 80%, cell transfection was conducted. Serum-free Opti-MEM was used to wash the cells twice, and 1.5 mL Opti-MEM was added. Next, 250 μL Opti-MEM was used to dilute 3 μg interference plasmid and NC plasmid, respectively, followed by slight mixing and addition of 1 μg overexpression vector. A total of 10 μL Lipofectamine 2000 reagent was diluted by 250 μL Opti-MEM, mixed gently and incubated at room temperature for 5 minutes. The diluted plasmids and Lipofectamine 2000 were gently mixed and allowed to stand at room temperature for 20 minutes. Afterwards, plasmid mixture (500 μL each tube) was slowly added to each cell well and gently mixed. Following 6 hours of culturing in an incubator of 37°C with 5% CO 2 , the previous medium was replaced by complete culture solution (without antibiotics) and further cultured in a 37°C incubator with 5% CO 2 overnight. At the 3rd day, the expression of enhanced green fluorescent protein (EGFP) protein was observed at 488 nm under a fluorescence microscope, and the expressions of ln-cRNA H19 and CDH1 in the cells were determined by reverse-transcription quantitative polymerase chain reaction (RT-qPCR) in order to detect the interference efficiency and overexpression efficiency.

| Methylation-specific polymerase chain reaction (MSP) assay
Lung adenocarcinoma cancer tissues and adjacent tissues (each n = 60) were collected. Initially, the tissue DNA was extracted by conventional phenol-chloroform-isoamyl alcohol method and the concentration and purity of DNA were determined using ultraviolet spectrophotometry. Next, DNA modification and purification were performed with hydroquinone and sodium bisulfite (Sigma-Aldrich, SF) and DNA Clean-up System (Promega). Methylation-specific PCR reaction was as follows: The reaction system was 25 μL, including 2 μL sample DNA, 2.5 μL 10 × PCR buffer, 0.5 μL forward primer and 0.5 μL reverse primer, 2 μL dNTP, 0.2 μL Taq enzyme and 17.3 μL double-distilled water. Reaction conditions were as follows: pre-denaturation at 95°C for 5 minutes, and 35 cycles of denaturation at 94°C for 30 seconds, at 57°C for 45 seconds, at 72°C for 30 seconds and extension at 72°C for 10 minutes. The annealing temperature of methylation reaction was 57°C, and that of non-methylation reaction was 54°C, with double-distilled water used as a blank control and 1000 bp D L2000 as a molecular mass marker. Primer sequences were as follows: (a) methylation   primers: forward primer: 5′-TTAGGTTAGAGGGTTATCGCGT-3;   reverse primer: 5′-TAACTAAAAATTCACCTACCGAC-3; amplification product was 115 bp; (b) non-methylation primers: forward   primer: 5′-TAATTTTAGGTTAGAGGGTTATTGT-3′; reverse primer: 5 ′-CACAACCAATCAACAACACA-3′; amplification product was 97 bp. The obtained PCR product was 5 μL, which then underwent 2% agarose gel electrophoresis at 100 V for 40 minutes. Finally, the electrophoresis results were analysed by a laser scanner (Pharmasia Company). The experiment was repeated three times (also applicable to cell experiments).

| Immunohistochemistry
The specimens were fixed with formaldehyde for more than 24 hours, conventionally dehydrated by gradient alcohol (70%, 80%, 90%, 95%, 100%; 1 minute each), cleared by xylene twice, 5 minutes each, immersed and embedded in paraffin and sliced into 4-μm sections. The sections were baked in a 60°C oven overnight, deparaffinized with xylene, and immersed in gradient alcohol (100%, 95%, 80%, 70%; 5 minutes each) and running water for 5 minutes and washed with phosphate-buffered saline (PBS) three times, 3 minutes each. Next, the sections were heated with 0.01 mol/L citric acid buffer for antigen retrieval (10 minutes) and cooled down to room temperature, followed by 3 PBS washes, 3 minutes each. Endogenous peroxidase was blocked by immersing the sections in 0.3% H 2 O 2 -formaldehyde solution for 20 minutes. Following three times of PBS washes, the sections were blocked with 10% goat serum (36119ES03, Yeasen Company) at room temperature for 10 minutes. CDH1 antibody (1:50; ab1416, Abcam Inc) working solution was added to the sections, followed by incubation overnight at 4°C, with the primary antibody replaced by PBS as NC. Horseradish peroxidase (HRP)-labelled goat anti-rabbit secondary antibody immunoglobulin G (IgG; 1:1000; ab6721, Abcam Inc) was dropped to the sections and incubated at room temperature for 30 minutes. Next, 3,3′-diaminobenzidine (DAB, Beyotime Biotechnology Co.) was added for coloration for 5 minutes, which was observed under a microscope. After that, the sections were rinsed with running water for 5 minutes, counterstained with haematoxylin for 3 minutes, differentiated by 1% ethanol-hydrochloric acid for 5 seconds, rinsed in running water for 10 minutes to return blue colour, mounted by neutral gum, observed under an optical microscope and photographed. The cells with brown-yellow cytoplasm or membrane were considered to be positive. Five fields of vision were randomly selected in each section to calculate the positive rate.

| RT-qPCR
The total RNA of cells was extracted using one-step method with ref- The experiment was repeated three times independently. The aforementioned method was equally applicable to the mRNA determination among the cells.

| Western blot analysis
The obtained tissue specimens were ground into homogenate

| Cell counting kit (CCK)-8 assay
Cell proliferation was detected using CCK-8 (GM-040101-5, Dojindo Laboratories) based on its protocol. Cells in logarithmic growth after transfection were collected, rinsed with PBS, treated with trypsin, washed with additional PBS and suspended. After the cell concentration was adjusted to 5 × 10 3 cells/μL, the cells were transferred to a 96-well plate, with six duplicated wells set for each group. The plate was then placed in a 37°C incubator with 5% CO 2 in which the cells were cultured for consecutive 2 days. At the last 4 hours, 10 μL CCK-8 solution was added to each well for further culturing, after which spectrophotometer (UV-1800A, Macylab) was employed to measure OD value at the wavelength of 450 nm based on the following formula: OD = OD experiment well -OD bank well . The experiment was repeated three times.

| Sphere-forming assay
Human lung adenocarcinoma A549 cell line in each group underwent sphere-forming culture. With the addition of high glucose DMEM and 10% FBS, the A549 cells after amplification transfection were cultured in vitro in a 37°C thermostat with 5% CO 2 . When the cell confluence reached 80%, 0.25% trypsin was used to treat the cells, which were then collected, suspended and counted. Next, the cells were seeded in a 6-well ultra-low attachment plate and cultured in a 37°C incubator with 5% CO 2 . After 6 days of culturing, the tumour spheres grew into 50 μm and the cell suspension was collected. Following low-speed centrifugation at 64 g for 5 minutes, the tumour spheres precipitated at the bottom of the centrifuge tube, while the single cells with lighter weight were still suspended in the supernatant. After the upper cell suspension was removed, PBS was slightly added to resuspend the precipitates. Pure glomus cells were obtained by two repetitions of the above steps, treated with 0.25% trypsin/0.02% ethylenediamine tetraacetic acid (EDTA), incubated at a 37°C incubator for 2 minutes and resuspended. Finally, the cells were counted and the average value was taken from three independently repeated experiments.

| Scratch test
Cells in logarithmic growth were collected after 48 hours of transfection and seeded in a 6-well plate at a density of 1 × 10 6 /well. The

| Immunofluorescence assay
At one week after plasmid in each group was transfected into A549 cells, cell slides were prepared and fixed for 20 minutes in mixed cold fixative freshly prepared with methanol and acetone (volume ratio

| Chromatin immunoprecipitation (ChIP)
A ChIP kit (Merck Millipore) was used to study the enrichment of

| Luciferase reporter gene assay
The relationship between lncRNA H19 and CDH1 was detected using luciferase reporter gene assay. Based on the binding sequence between the sequence of CDH1 promoter and lncRNA H19, the wild target sequence (CDH1 wt) and mutation sequence

| Statistical analysis
The SPSS 21.0 software (IBM Corp., Armonk, NY, USA) was used for data analysis. Measurement data were presented as mean ± standard deviation. The t test was used for data comparison between two groups and one-way analysis of variance (ANOVA) for comparisons among multiple groups. Enumeration data were demonstrated as percentage or ratio and compared by chi-square analysis. P < .05 was considered to be statistically significant difference. We determined methylation of CDH1 in 60 cases of lung adenocarcinoma tissues and 60 cases of adjacent tissues using MSP assay.

| RE
No methylation of CDH1 was detected in all the 60 adjacent tissues, while among the 60 lung adenocarcinoma tissues, 17 cases showed methylation of CDH1 promoter, accounting for 28.33%. Compared with the adjacent tissues, the lung adenocarcinoma tissues exhibited a notable increase in methylation of CDH1 promoter ( Figure 1D, Table 2). The aforementioned results suggest that methylation of CDH1 promoter is correlated with the occurrence of lung cancer.

| A549 cell line is selected for cell transfection with silencing and overexpression plasmids of CDH1 successfully constructed
We performed RT-qPCR ( Figure 2A found in the siRNA-3 group among the four groups, with an inhibitory efficiency more than 70%. Therefore, siRNA-3 was selected for further experimentation. The oe-CDH1 group had a higher CDH1 mRNA expression than the oe-NC group ( Figure 3C). Based on the results of Western blot analysis, the expression of CDH1 in the oe-CDH1 group significantly increased compared with the oe-NC group (P < .05) ( Figure 3D). These results reveal that silencing and overexpression plasmids of CDH1 are successfully constructed.

| Silencing of lncRNA H19, overexpression of CDH1, or demethylation of CDH1 inhibits sphereforming ability and invasion and promotes the apoptosis of A549 cells
In order to investigate the effects of CDH1 and lncRNA H19 on the pro-

| Silencing of lncRNA H19, overexpression of CDH1 or inhibition of CDH1 methylation suppresses migration and invasion and EMT of A549 cells
For investigation into the effects of CDH1 and lncRNA H19 on the EMT process, we further performed scratch test and Transwell assay to ascertain whether silencing of lncRNA H19, overexpression of CDH1 or inhibition of CDH1 methylation could affect migration and invasion of lung adenocarcinoma cells ( Figure 5A, From the aforementioned results, silencing of lncRNA H19, overexpression of CDH1 or inhibition of CDH1 methylation could suppress migration and invasion and EMT of A549 cells.

| Silencing of lncRNA H19 inhibits methylation of CDH1
We downloaded lncRNA H19 expression in multiple cancers from the TCGA database and found that compared with normal tissues, lncRNA H19 was highly expressed in tissues of lung adenocarcinoma ( Figure 6A). We designed probe primers of lncRNA H19, followed by FISH experiment ( Figure 6B) to detect the subcellular location of lncRNA H19. The results showed that lncRNA H19 was located in the nucleus. After Merge, the nucleus had a fluorescent staining, suggesting that lncRNA H19 was highly expressed in the nucleus of the A549 cell line. Next, we input nucleotide sequences (3000 bp) near the CDH1 gene promoter in the Meth Primer software so as to analyse the CpG island in the CDH1 promoter region. The results showed that there was a CpG island in the CDH1 promoter region ( Figure 6C). A biological prediction website was used to predict the target binding site of lncRNA H19, as shown in Figure 6D, and luciferase gene reporter was employed to verify whether CDH1 is a target gene of lncRNA H19. The results showed that in comparison with the NC group, the luciferase activity of the CDH1 wild-type promoter region was significantly reduced in the si-lncRNA H19 group (P < .05), while the luciferase activity of the mutant promoter region did not differ significantly (P > .05) ( Figure 6E), indicating that lncRNA H19 specifically bind to the promoter region of CDH1. Subsequently, we continued to study the relationship between lncRNA H19 and methylation of CDH1 using double luciferase ( Figure 6F), the results of which demonstrated that compared with the NC group, the expression of M-CDH1 in the si-lncRNA H19 was significantly reduced and the expression of N-CDH1 increased.
For further verification, we performed ChIP assay ( Figure 6G).
After incubation with DNMT1 and DNMT3A antibodies, the results showed that the expression of lncRNA H19 and the expression of CDH1 promoter region in the DNMT1 and DNMT3A groups were significantly higher than those in the IgG group. LncRNA H19 induced DNMT1 and DNMT3A, which were enriched around the CDH1 promoter, to promote the methylation of CDH1.
Taken together, silencing of lncRNA H19 could inhibit methylation of CDH1.

| Silencing of lncRNA H19, overexpression of CDH1 or inhibition of CDH1 methylation inhibits tumorigenicity of A549 cells in nude mice
A nude mouse model of tumour formation was established. The tumour length and width were measured at the 1st, 2nd, 3rd and 4th week, respectively. The tumour tissues of nude mice were weighed at the 4th week. The results showed that there were no significant differences in volume and weight of the tumour tissues among the mock, oe-NC, si-NC and DMSO groups at the 1st, 2nd, 3rd and 4th week. Compared with the oe-NC group, the volume and weight of tumour tissues in the oe-CDH1 group decreased significantly (P < .05).

| D ISCUSS I ON
Lung adenocarcinoma is a major contributor to cancer-related death all over the world. 3 LncRNAs are considered as biomarkers as well as molecular targets for cancer therapies. 18  was found to be aberrantly expressed in rats with lung adenocarcinoma, 26 which was consistent with the present study.
One of the key findings of our study was that silencing of lncRNA H19, overexpression of CDH1 or reducing CDH1 methylation sup- diseases, via regulation of cell proliferation. 28,29 In addition, lncRNAs are identified as emerging regulators in metastatic cascade, among which lncRNA H19 has been found to be able to regulate EMT, thereby affecting tumour cell metastasis. 30 Overexpression of lncRNA H19 is responsible for tumorigenesis and progression of diverse cancer types; lncRNA H19, serving as a microRNA sponge, could accelerate the process of EMT in colorectal cancer. 31 Moreover, lncRNA H19 stimulated invasion and migration of let-7 in pancreatic ductal adenocarcinoma, in part through the up-regulation of HMGA2-mediated EMT. 32 The above reports all supported our results that overexpressed lncRNA H19 could accelerate the development of lung adenocarcinoma cells, while by silencing of lncRNA H19, this trend could be significantly reversed. CDH1 is considered as an anti-oncogene in interactions between epithelial cells. 33 Reduced CDH1 expression is referred to as a biomarker for cancer invasion as well as metastasis. 34 DNA methylation is one of the epigenetic changes that account for aberrant regulation of tumour-related genes, which can affect cellular activities such as cell proliferation, apoptosis and metastasis. 35 Consistent with our results, inhibition of CDH1, identified as a critical gene during EMT process, was found to aid in tumour metastasis of breast cancer. 36 Therefore, it has been suggested that lncRNA H19 might affect cellular processes of lung adenocarcinoma cells by regulating methylation of CDH1.
Furthermore, we found that silencing of lncRNA H19 could inhibit methylation of CDH1 promoter in lung adenocarcinoma. Accumulating evidence demonstrates that lncRNAs are able to mediate transcription as well as DNA methylation of CpG island. 37 LncRNA H19 has been revealed to be capable of interacting with histone lysine methyltransferases. 38 Moreover, lncRNA H19 could bind to S-adenosylhomocysteine hydrolase (SAHH) and repressed its expression, thereby leading to DNMT3B, which produced methylate gene subsets. 39 It was reported that silencing of lncRNA H19 could elevate Nctc1 methylation which was mediated by DNMT3B. 40 Additionally, lncRNA H19 was found to be able to regulate expression of Hnf4α as well as its methylation. 41 In the current study, the ChIP assay showed that lncRNA H19 induced DNMT1 and DNMT3A to promote the methylation of CDH1; however, the potential mechanism implicated in link of LncRNA H19 with methyltransferases in lung adenocarcinoma cells still remained complicated.
To conclude, the present study demonstrated that silencing of lncRNA H19 inhibited EMT and proliferation while promoting apoptosis of lung adenocarcinoma cells via inhibition of methylation of CDH1 promoter (Figure 8), which may provide a novel molecular target for treatment of lung adenocarcinoma. However, the specific mechanism by which lncRNA H19 inhibits methylation of CDH1 promoter still lacks elucidation. Therefore, further study is required for validating our findings. In addition, we plan to supplement BSP experiment and to explore the relationship between lncRNA H19 and DNMT3B in the later experiment if possible.

ACK N OWLED G EM ENTS
We would like to acknowledge the reviewers for their helpful comments on this paper.

CO N FLI C T S O F I NTE R E S T
The authors have declared no conflicts of interest.

DATA AVA I L A B I L I T Y
The data used to support the findings of this study are available from the corresponding author upon request.