RETRACTED ARTICLE: Long non-coding RNA GATA6-AS1 upregulates GATA6 to regulate the biological behaviors of lung adenocarcinoma cells

LncLocator (http://​www.​csbio.​sjtu.​edu.​cn/​bioinf/​lncLocator/​) [20] was used for determination of the subcellular location of GATA6-AS1. miRDB database [21] was used to reveal the miRNAs (181 miRNAs, data not shown) that potentially bind with GATA6 and the binding site of GATA6 and miR-4530. Regrna2 database [22] was used to reveal the miRNAs (57 miRNAs, data not shown) that potentially bind with GATA6-AS1 and the binding site of GATA6-AS1 and miR-4530.

Thirty-five clinical LUAD tumor tissues and paired adjacent non-tumor tissues were collected from Liaocheng People’s Hospital. All the tissues were collected during surgical procedures and stored in liquid nitrogen or at − 80 °C for future use. Written informed consents were signed by all the patients and the study was approved by the Ethics Committee of Liaocheng People’s Hospital. All methods were carried out in accordance with relevant guidelines and regulations. Clinical characteristics of LUAD patients were provided in Table 1.

The four LUAD cell lines (A427, A549, H1975, HCC827) and human lung cell line (BEAS-2B) were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). All the cells were maintained in RPMI-1640 medium (Gibco, Carlsbad, CA, USA), added with 10% heat-inactivated fetal bovine serum (Gibco) in a moist air atmosphere with 5% CO₂ at 37 °C.

For the purpose of overexpressing GATA6-AS1 or GATA6, the pcDNA3.1-GATA6-AS1 or pcDNA3.1-GATA6 vectors were purchased from Invitrogen (Carlsbad, CA, USA), while the empty pcDNA3.1 vector served as a negative control (NC). The specific short-hairpin RNAs (shRNAs) targeting GATA6 (sh-GATA6#1: 5’-CCGGCACCACAACTACCACCTTATGCTCGAGCATAAGGTGGTAGTTGTGGTGTTTTTTG-3′ and sh-GATA6#2: 5′-CCGGATTCCCATGACTCCAACTTCCCTCGAGGGAAGTTGGAGTCATGGGAATTTTTTTG-3′) and negative control (sh-NC) were synthesized by GenePharma. (Shanghai, China) for GATA6 knockdown. MiR-4530 mimics and corresponding NC mimics were also purchased from GenePharma. A549 and H1975 cells went through transfection for approximately 48 h using Lipofectamine 2000 (Invitrogen), followed by being collected and utilized for following experiments.

Total RNA was isolated and extracted from LUAD cells by TRIzol reagent (Invitrogen) according to the recommendations of manufacturer. Next, RNA concentration was detected and then reverse transcribed to single-stranded complementary DNA with a Reverse Transcription System Kit (Takara, Dalian, China). Subsequently, RT-qPCR reactions were carried out using Universal SYBR Green Master (Roche, Basel, Switzerland). The quantifications of GATA6-AS1, GATA6, and miR-4530 were measured by the 2^−ΔΔCt method [23]. GAPDH and U6 acted as the internal references for normalization, respectively. Primer sequences were listed in Table 2.

For conducting the assay, the Cell-Light EdU DNA Cell Proliferation Kit (RiboBio, Guangzhou, China) was adopted based on the specification of manufacturer. Transfected cells were cultured with EdU for 2 h, fixed with 4% paraformaldehyde, stained with Apollo Dye Solution and mounted with DAPI (Sigma-Aldrich, St. Louis, Missouri, USA). Finally, samples were imaged under a fluorescence microscope (Leica, Mannheim, Germany) and were manually counted. EdU positive cells were calculated as the number of EdU (green) positively stained cells/the number of DAPI (blue) positively stained cells in several randomly selected fields.

The treated cells (with the density of 1 × 10³ cells/well) were plated into 6-well plates for 2 weeks of incubation, and the medium was replaced every 3 days. Thereafter, the cells were immobilized by paraformaldehyde for 30 min and stained with crystal violet (Aladdin, China) for 20 min at room temperature. After purification by phosphate buffered saline (PBS) twice, the viable cells were visualized and counted from 5 randomly identified fields under a microscope (Nikon, Tokyo, Japan).

The cell apoptosis was investigated using the TUNEL apoptosis assay kit (Beyotime, Shanghai, China) as instructed by the manufacturer. In short, the LUAD cells were cleared with PBS for three times and mounted with 4% paraformaldehyde for 30 min. Next, the abovesaid cells were treated with the PBS possessing 0.3% Triton X-100 at room temperature for 5 min. Subsequently, the TUNEL detection solution was added to the cells. Apoptotic cells were manually counted in 5 randomly selected fields under a light microscope (Olympus Corporation). TUNEL positive cells were calculated as the number of TUNEL (green) positively stained cells/the number of DAPI (blue) positively stained cells.

Flow cytometry was conducted to examine cell apoptotic ability utilizing an Annexin V-Fluorescein Isothiocyanate (FITC) Apoptosis Detection Kit (BioLegend, San Diego, CA, USA). After 48 h of transfection, cells were extracted by trypsin without EDTA, and washed thrice with phosphate-buffered saline. After centrifugation and resuspension in 100 μL of flow cytometry binding buffer, the cells were treated with 5 μL of Annexin-V-FITC and 5 μL of propidium iodide solution in the dark for 15 min. The apoptotic ratio was measured with a flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

The transfected cells were embedded into a 6-well plate with 1 × 10³ cells per well and cultivated overnight. Thereafter, a scratch was created with 10 μL pipette tip as the cells were grown to approximately 90% confluence. After being purified thrice with PBS, the cells were imaged from 5 fields in each group via a light microscope (Olympus, Tokyo, Japan). Twenty-four hours later, images were captured again at the same fields.

The cell invasion assay was performed with a 24-well Transwell chamber (Corning, NY, USA). Transfected A549 and H1975 cells (4 × 10⁴ cells) were plated into the Matrigel-precoated upper chamber. The lower chamber was filled with 600 µL RPMI-1640 containing 10% FBS. After incubation for 24 h, cells on the upper side of the membrane were removed using clean swabs, and cells on the underside were captured under a Leica DM IL LED inverted microscope. The number of invaded cells was counted in 5 randomly selected fields.

Total protein was separated and collected from the transfected cells with a lysis buffer (Thermo Scientific, Massachusetts, USA). The protein concentration was detected by the BCA (bicinchoninic acid) Protein Assay kit (Thermo Scientific). The proteins were isolated with 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, followed by transferring onto a polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA, USA). Afterwards, the membrane was sealed with 5% non-fat milk for 10 min and incubated for 12 h at 4 °C with primary antibodies of anti-GAPDH (1:2500, ab9485, Abcam) and anti-GATA6 (1:1000, ab175927, Abcam). Subsequently, the PVDF membrane was washed with TBST buffer for three times and probed with HRP-conjugated secondary antibody for 1 h at room temperature. At last, the immunoreactive bands were observed via chemiluminescence (Millipore) and were quantified using ImageJ 1.52 software. GAPDH was used for normalization.

The nuclear and cytoplasmic parts of transfected cells were divided and rinsed according to the protocol of the Cytoplasmic & Nuclear RNA Purification Kit (Norgen). The expression levels of GAPDH, U6 and GATA6-AS1 in nuclear and cytoplasm fractions of cells was analyzed via RT‐qPCR.

An RIP assay was conducted with the Magna RIP RNA-Binding Protein Immunoprecipitation kit (Millipore). Moreover, cell lysates were collected in RNA immunoprecipitation assay (RIPA) lysis buffer (Millipore) with magnetic beads. Next, anti-Ago2 (ab186733, Abcam) or anti-IgG (ab205718, Abcam) was coincubated for 12 h with magnetic beads at 4 °C for the purpose of acquiring immunoprecipitation complex. Later, the immunoprecipitated RNA was gathered and purified using TRIzol reagent (Takara, Dalian, China). Finally, the results were detected by RT-qPCR.

Partial sequences of GATA6-AS1 and GATA6 3′ untranslated region (3′UTR) including wide type or mutant type miR-4530 binding sites were inserted into dual-luciferase reporter vector (pmirGLO; Promega, Madison, WI, USA) to produce GATA6-AS1-WT, GATA6-AS1-MUT and GATA6-WT, GATA6-MUT. Subsequently, the constructed reporter plasmids were respectively transfected into treated cells with miR-4530 mimics. Afterwards, a dual-luciferase reporter assay system (Promega) was adopted to determine the luciferase activity of cells.

The statistical analysis was performed utilizing Graphpad Prism 5.02 software (La Jolla, CA, USA). All data were obtained from at least three independent experiments and showed as the means ± standard deviation. Student’s t-test was utilized to measure differences statistically between two groups, while statistical differences among multiple groups were compared by one-way analysis of variance followed by Tukey's post hoc test. The p < 0.05 was indicative of significant difference.

To investigate the functions of GATA6-AS1 in LUAD, the expression pattern of GATA6-AS1 in LUAD tissues was discovered from the database of GEPIA (http://​gepia.​cancer-pku.​cn/​index.​html). The finding indicated that GATA6-AS1 expression was significantly decreased in LUAD tissues (n = 483) compared with adjacent noncancerous tissues (n = 347) (Fig. 1a). Thereafter, RT-qPCR revealed that GATA6-AS1 expression was downregulated in 35 LUAD tissues compared to that in 35 non-tumor tissues (Fig. 1b). GATA6-AS1 expression was lower in LUAD cells (A427, A549, H1975 and HCC827) than normal human lung cells (BEAS-2B) (Fig. 1c). Since GATA6-AS1 expression was decreased in A549 and H1975 cells, the two cell lines were selected for the subsequent assays. Later, the expression of GATA6-AS1 was significantly elevated in A549 and H1975 cells with transfection of pcDNA3.1-GATA6-AS1, as suggested by RT-qPCR (Fig. 1d). Afterwards, an EdU assay was conducted to analyze the suppressive effect of overexpressed GATA6-AS1 on the proliferative ability of A549 and H1975 cells (Fig. 1e). Moreover, a colony formation assay was carried out, showing that cell proliferation was inhibited after increasing GATA6-AS1 expression (Fig. 1f–g). Next, we carried out the TUNEL assay to detect the influence of upregulated GATA6-AS1 on cell apoptosis. The findings showed that the upregulation of GATA6-AS1 led to a significant increase of the percentage of apoptotic cells (Fig. 1h). Furthermore, the results of flow cytometric analysis suggested that GATA6-AS1 overexpression increaseed the apoptotic rate of A549 and H1975 cells (Fig. 1i). Later, wound healing assay and Transwell invasion assay were conducted to examine whether overexpressed GATA6-AS1 affects migration and invasion capacities of A549 and H1975 cells. The findings demonstrated that GATA6-AS1 overexpression repressed cell migration and invasion compared with negative control group (Fig. 1j–l).

As the above-mentioned experiments revealed that GATA6-AS1 served as a tumor suppressor in LUAD, GATA6, the cognate sense transcript of GATA6-AS1, drew our attention. We firstly searched the database of GEPIA, finding that GATA6 was low-expressed in LUAD tissues (n = 483) compared to adjacent noncancerous tissue samples (n = 347) (Fig. 2a). GATA6 was downregulated in 35 clinical LUAD tissues compared to that in 35 non-tumor tissues (Fig. 2b). In addition, the results of RT-qPCR and western blot revealed the downregulation of GATA6 expression in LUAD cells compared with normal human lung cells at both mRNA and protein levels (Fig. 2c). Afterwards, we wondered if GATA6-AS1 exerts regulatory effects on GATA6. RT-qPCR analysis revealed that the expression of GATA6 was promoted in A549 and H1975 cells upon the transfection of pcDNA3.1-GATA6-AS1 (Fig. 2d). Meanwhile, the increased protein level of GATA6 in A549 and H1975 cells after overexpressing GATA6-AS1 was depicted by western blot (Fig. 2e). Thereafter, the putative subcellular location was revealed from LncLocator database, suggesting that GATA6-AS1 was mainly located at cytoplasm of LUAD cells (Fig. 2f), indicating the posttranscriptional regulation of GATA6-AS1 on GATA6. This result was further supported by a subcellular fractionation assay, which showed the most percentage of GATA6-AS1 in the cytoplasm of A549 and H1975 cells (Fig. 2g).

The ceRNA mechanism is a typical posttranscriptional mechanism, and lncRNAs were widely reported to regulate their cognate sense transcripts via the ceRNA mechanism [24‐27]. We hypothesized that GATA6-AS1 regulates GATA6 via the ceRNA mechanism. Subsequently, an RIP assay was conducted to investigate whether GATA6-AS1 exists in RNA-induced silencing complexes (RISCs). As illustrated by Fig. 3a, the abundant enrichment of GATA6-AS1 suggested that GATA6-AS1 might act as the molecular sponge of specific miRNA in the ceRNA regulatory network. Next, two miRNAs (miR-4530 and miR-1520) sharing potential binding sites with GATA6-AS1 were screened out with the screening condition of miRDB [21] and RegRNA2 [22] databases (Fig. 3b). Afterwards, the expression levels of miR-4530 and miR-1200 were examined using RT-qPCR, suggesting that miR-4530 was overexpressed in LUAD cells (Fig. 3c, d). Moreover, RT-qPCR analysis revealed that the expression of miR-4530 was elevated in A549 and H1975 cells with transfection of miR-4530-mimics (Fig. 3e). In addition, the expression level of GATA6 was suppressed by miR-4530 overexpression and then reversed by upregulated GATA6-AS1 in A549 and H1975 cells, as indicated by RT-qPCR (Fig. 3f). GATA6-AS1 can suppress the expression of miR-4530, as revealed in Fig. 3g. Overexpression of miR-4530 rescued the stimulating effects of GATA6-AS1 on GATA6 expression (Fig. 3h). Thereafter, the binding sequences between GATA6-AS1 and miR-4530, or between GATA6 and miR-4530 were depicted via prediction from miRDB and RegRNA2 (Fig. 3i). Subsequently, a dual‑luciferase reporter assay revealed that miR-4530 upregulation inhibited the luciferase activities of GATA6-AS1-WT and GATA6-WT, while the luciferase activities of GATA6-AS1-MUT and GATA6-MUT were unchanged following the same transfection in A549 and H1975 cells (Fig. 3j, k). Later, as shown in the RIP assay, the enrichment of GATA6-AS1, miR-4530 and GATA6 was increased in anti-Ago2 precipitated products in A549 and H1975 cells, indicating that GATA6-AS1, miR-4530 and GATA6 co-existed in RISC (Fig. 3l). In conclusion, the GATA6-AS1 served as a ceRNA against miR-4530 to upregulate GATA6.

For exploring the underlying functional mechanism of GATA6-AS1 in LUAD, a series of rescue assays were conducted. First, the expression of GATA6 was repressed significantly in A549 and H1975 cells after the knockdown of GATA6 (Fig. 4a). Subsequently, an EdU assay revealed that the GATA6-AS1 upregulation-reduced apoptotic ability of A549 and H1975 cells was restored by silencing GATA6 (Fig. 4b, c). Additionally, the inhibited cell proliferation by GATA6-AS1 overexpression was countervailed after depleting GATA6, as suggested by a colony formation assay (Fig. 4d). Later, the TUNEL assay was performed to illustrate the inhibitory effect of GATA6 knockdown on cell apoptosis previously promoted by upregulated GATA6-AS1 (Fig. 4e). Furthermore, the apoptotic rate of A549 and H1975 cells increased by overexpressed GATA6-AS1 and then suppressed by GATA6 depletion, as elucidated by flow cytometry analysis (Fig. 4f). Thereafter, wound healing and Transwell assay revealed that the GATA6-AS1 overexpression-inhibited migration and invasion capacities of A549 and H1975 cells were counteracted after downregulation of GATA6 (Fig. 4g–h). Therefore, GATA6-AS1 inhibited malignant progression of LUAD cells via upregulating GATA6.