LINC01089 suppresses lung adenocarcinoma cell proliferation and migration via miR-301b-3p/STARD13 axis

LUAD cell lines (PC9, H2073, H-1975, A549; ATCC, Manassas, VA, USA) and human bronchial epithelial cell line (BEAS-2B; ATCC) were cultivated in RPMI 1640 medium (Gibco, Grand Island, NY, USA) at 37 °C with 5% CO₂. The penicillin, streptomycin and 10% fetal bovine serum (FBS; Gibco) were used as the medium supplements. PC9 cell line was purchased from Mingzhoubio (Ningbo, Zhejiang, China), and other cell lines were all bought from ATCC. The catalogue numbers of the cell lines are listed as follows: PC9 (MZ-0668); H2073 (CRL-5918); H-1975 (CRL-5908); A549 (CCL-185); BEAS-2B (CRL-9609).

For transfection, A549 and H-1975 cells at 80–90% confluence were seeded into 6-well plates and transfected for 48 h by using Lipofectamine 3000 kit (Invitrogen, Carlsbad, CA, USA). Cells stably transfection were screened by utilizing G418 and then applied in subsequent experiments. The pcDNA3.1/LINC01089, pcDNA3.1/STARD13 and control (pcDNA3.1), STARD13-specific shRNAs (sh-STARD13#1/2) and control (sh-NC), together with miR-301b-3p mimics/inhibitor and control (NC mimics/inhibitor), were all procured from RiboBio (Guangzhou, China). In addition, the primary ADC cells were used for in vivo assays with HLC as a control in the study. The sequences were listed as follows:

In line with the manual of Trizol (Invitrogen), total RNA from A549 and H-1975 cells was obtained, centrifuged and washed. After using the Prime Script™ RT Master Mix (TaKaRa, Otsu, Japan), the synthesized cDNA was subjected to SYBR Green I fluorescent method (TaKaRa) on Applied Biosystems 7900 Real‐Time PCR System (Applied Biosystems, Foster City, CA, USA). The relative quantification of samples was tested by the equation 2^−ΔΔCt. GAPDH or U6 was used as the normalized control. The primer sequences were shown in Table 1.

10 μl of CCK-8 reagents (Dojindo Molecular Technologies, Tokyo, Japan) was added to the medium containing A549 and H-1975 cells for 2 h. The microplate reader (Bio-Tek, Winooski, VT, USA) was applied for monitoring the absorbance at wavelength of 450 nm.

Cell proliferation of A549 and H-1975 was analyzed via EdU incorporation assay kit (Ribobio). LUAD cells were placed in 96-well plates with 100 μl of 50 μM EdU for 3 h, and then treated with 4% paraformaldehyde and 100 μl of 0.5% Troxin X-100 (×100; Sigma-Aldrich, Miamisburg, OH, USA). Following Apollo® 488 fluorescent staining, nuclei were counterstained with DAPI (Beyotime, Shanghai, China). Thereafter, cells were observed and analyzed with fluorescent microscope (Leica, Wetzlar, Germany).

Cell apoptosis of LUAD cells were measured with the help of FITC Annexin V Apoptosis Kit (BD Biosciences, San Jose, CA, USA). A549 and H-1975 cells treated with trypsin were washed in pre-cooled phosphate buffer saline (PBS). 5 × 10⁵ cells were cultured in 100 μl of 1× binding buffer adding 5 μl of PI and 5 μl of FITC Annexin V at room temperature. After 400 μl of 1× binding buffer was added, Flow Cytometer (BD Biosciences) was utilized to determine cell apoptosis rate.

In the wound healing assay, the collected 5 × 10⁵ A549 or H-1975 cells seeded in 24-well plates were cultivated at 37 °C until cells reached 100% confluence after transfection. Thereafter, cells were scraped by 200 μl sterile micropipette tip, and then cultured at 37 °C for 24 h. After being washed for three times in serum-free medium to clear the detached cells, the scratch was imaged by microscope at the time 0 h and 24 h for analysis.

This assay was conducted in 24-well Transwell chamber (Corning, Corning, NY, USA) containing 8 μm pore size polycarbonate membrane filter. A549 or H-1975 cells were seeded in the upper chamber with 500 μl of culture medium without FBS, while lower chamber was filled with 500 μl of complete medium. After 24 h of incubation, cells in the lower side were subjected to 4% formaldehyde (Sigma-Aldrich) and 1% crystal violet (Sigma-Aldrich), and then counted under optical microscope (Olympus, Tokyo, Japan) at ×200 magnification.

The segmentation of nucleus and cytoplasm was performed by PARIS™ kit (Ambion, Austin, TX, USA). 1 × 10⁷ A549 or H-1975 cells were washed on ice and re-suspended in 500 μl pre-cooled cell fractionation buffer for 10 min. The supernatant was reaped as cell cytoplasm after centrifugation, while the nuclear deposit was treated with cell disruption buffer. After collecting nuclear and cytoplasmic fraction, RT-qPCR was performed for quantifying LINC01089, with GADPH and U6 as cytoplasmic and nuclear controls, respectively.

Using Pierce Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific, Waltham, MA, USA), RNA pull-down assay was carried out in A549 and H-1975 cells. Biotinylated LINC01089 probes (LINC01089 biotin probe) were incubated with cell extracts and streptavidin magnetic beads (Invitrogen). The LINC01089 no-biotin probe was used as the control. Finally, the RNA complexes bound to beads were analyzed by RT-qPCR.

The pmirGLO luciferase vectors (Promega, Madison, WI, USA) containing the firefly reporter gene were formed using the wild-type (WT) or mutant (Mut) LINC01089 sequences with or without miR-301b-3p binding sites, termed as LINC01089-WT/Mut. A549 and H-1975 cells were plated to 24-well plates (5 × 10⁴ cells/well), then co-transfected with LINC01089-WT/Mut for 48 h. Renilla luciferase reporter pRLCMV (Promega) acted as the normalized control. Dual-luciferase Reporter assay system (Promega) was applied to estimate the luciferase activity of each group.

The lysed LUAD cells were separated on 10% SDS-PAGE and transferred electrophoretically onto PVDF membranes (Millipore, Billerica, MA, USA). Following the treatment with 5% non-fat milk (Merck KGaA, Darmstadt, Germany), membranes were cultivated with anti-STARD13 (1:2000 dilution; ab126489; Abcam, Cambridge, MA, USA) or anti-GAPDH (1:2000 dilution; ab128915; Abcam) primary antibodies all night, followed by incubation with HRP-labeled secondary antibody (1:2000 dilution; ab6728; Abcam). Finally, the membranes were exposed to ECL chemiluminescence Detection kit (Millipore). GAPDH was an internal control.

Using Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore), RIP assay was carried out in A549 and H-1975 cells using 5 μg anti-AGO2 (03-110; Millipore) or 5 μg anti-IgG antibodies (12-370; Millipore). Anti-IgG group served as a negative control, and cell lysates from RIP lysis buffer were treated with the beads conjugated with above antibodies for 2 h at 4 °C, followed by RNA analysis via RT-qPCR.

Although LINC01089 has been revealed to be down-regulated in breast cancer cells [17], the expression of it in LUAD cells remains unknown. According to the results shown by GEPIA database (http://​gepia.​cancer-pku.​cn/​) and RT-qPCR, LINC01089 was discovered with lower expression in LUAD tissues than that in normal tissues (Additional file 2: Fig.S2A). After that, we observed a significantly down-regulated expression of LINC01089 in LUAD cell lines (PC9, H2073, H-1975 and A549) than that in normal BEAS-2B cell (Fig. 1a). As LINC01089 expression in A549 and H-1975 cells was the lowest, they were kept for follow-up studies. After that, gain-of-function assays were performed to detect the underlying biological function of LINC01089 in LUAD. Before that, RT-qPCR was utilized to detect the overexpression efficiency of LINC01089 in A549 and H-1975 cells (Fig. 1b). Next, CCK-8 assay illustrated that the OD value of cells transfected with pcDNA3.1/LINC01089 was lower when compared with the negative control, suggesting that cell viability of LUAD could be inhibited by LINC01089 overexpression (Fig. 1c). Then, EdU assay revealed that overexpressing LINC01089 led to a decrease in the proliferation of A549 and H-1975 cells since the percentage of EdU positive cells was notably decreased in cells transfected with pcDNA3.1/LINC01089 (Fig. 1d). Additionally, flow cytometry analysis testified that cell apoptosis ability could be enhanced by the overexpression of LINC01089 in A549 and H-1975 cells (Fig. 1e). Moreover, Transwell and wound healing assays uncovered that elevating LINC01089 expression in A549 and H-1975 cells could suppress cell migration (Fig. 1f, g). What’s more, a series of in vivo assays were performed to verify the above functions of LINC01089 in ADC cells. As shown by the results, LINC01089 was found in ADC cells with low expression and the overexpression efficiency of LINC01089 was examined via RT-qPCR (Additional file 1: Fig. S1A-B). After that, functional assays were carried out by us, which demonstrated that LINC01089 overexpression could repress the cell proliferation, migration while enhancing the apoptosis ability of ADC cells (Supplementary Fig. 1C-G). Besides, LINC01089 overexpression could lead to a suppressed tumor growth as well as a decreased tumor weight (Additional file 1: Fig. S1H-J). Taken together, we could confirm that LINC01089 was discovered with low expression in LUAD tissues and cells, and LINC01089 overexpression repressed LUAD cell malignant progression.

In order to explore the potential molecular mechanism of LINC01089 in LUAD, we conducted subcellular fractionation assay to verify the distribution of LINC01089 in LUAD cells. The result demonstrated that LINC01089 was mainly located in the cytoplasm of A549 and H-1975 cells (Fig. 2a). Thus, we speculated that LINC01089 might regulate LUAD progression through sponging miRNA. After searching for the related data from starBase (http://​starbase.​sysu.​edu.​cn/​index.​php), 14 miRNAs were predicted to have a binding relationship with LINC01089 (the detailed information of the 14 miRNAs are listed in Additional file 4: Table S1). After that, RNA pull-down assay revealed an abundant enrichment of miR-301b-3p in LINC01089 biotin probe group, suggesting that miR-301b-3p could bind to LINC01089 in A549 and H-1975 cells (Fig. 2b, Additional file 2: Fig. S2B). Additionally, starBase was applied to predict the binding sites between LINC01089 and miR-301b-3p (Fig. 2c). To investigate how LINC01089 interacted with miR-301b-3p, the expression of miR-301b-3p in LUAD cell lines and BEAS-2B cell and the overexpression efficiency of miR-301b-3p in A549 and H-1975 cells were examined at first (Fig. 2d, e). From our observation, miR-301b-3p overexpression caused a decrease in the luciferase activity of LINC01089-WT whereas no obvious change was observed in the luciferase activity of LINC01089-Mut (Fig. 2f). Furthermore, the expression of miR-301b-3p was notably decreased in A549 and H-1975 cells after the transfection with pcDNA3.1/LINC01089 (Fig. 2g). Taken together, miR-301b-3p directly bound to LINC01089 in LUAD cells.

To further probe the competing endogenous RNA (ceRNA) mechanism of LINC01089, miRWalk database (http://​mirwalk.​umm.​uni-heidelberg.​de/​) was consulted and seven mRNAs that might bind to miR-301b-3p was presented (Additional file 5: Table S2). Only STARD13 was uncovered in LUAD tissues with a significantly low expression compared with that in normal tissues (Fig. 3a, Additional file 2: Figure S2C-H). Hence, we selected STARD13 for following studies. After that, we found that the overexpression of LINC01089 resulted in a conspicuous up-regulation of the mRNA and protein level of STARD13 (Fig. 3b, Additional file 3: Figure S3A). Then, we examined the interference efficiency of miR-301b-3p (Additional file 3: Figure S3B). After miR-301b-3p expression was inhibited in A549 and H-1975 cells, we found that the expression of STARD13 mRNA and its protein level were observably enhanced in A549 and H-1975 cells (Fig. 3c, Additional file 3: Figure S3C). Besides, it was revealed that STARD13 expression in LUAD cell lines was signally reduced in contrast to that in BEAS-2B cell (Fig. 3d, Additional file 3: Figure S3D). In addition, starBase was applied to explore the binding sites between STARD13 3′UTR and miR-301b-3p (Fig. 3e). Next, the overexpression efficiency of STARD13 was examined in A549 and H-1975 cells (Additional file 3: Figure S3E). Besides, luciferase reporter assay demonstrated that STARD13 overexpression could reverse the inhibitory effect caused by overexpressing miR-301b-3p on the luciferase activity of LINC01089-WT. As for the luciferase activity of LINC01089-Mut, no evident changes were noted in different groups (Fig. 3f). Furthermore, an abundant enrichment of LINC01089, miR-301b-3p and STARD13 was captured in anti-AGO2 group, indicating that these RNAs co-existed in RNA-induced silencing complexes (RISCs) (Fig. 3g). Therefore, it was confirmed that LINC01089 could regulate STARD13 expression by sponging miR-301b-3p in LUAD.

With the intention of validating the existence and role of LINC01089/miR-301b-3p/STARD13 axis in LUAD, we first silenced STARD13 in A549 cells by transfecting the cells with sh-STARD13#1/2 (Additional file 3: Figure S3F). As sh-STARD13#1 exhibited better interference efficiency, it was kept for the following assays. Then data from RT-qPCR analysis confirmed that STARD13 depletion or miR-301b-3p overexpression could countervail the facilitating influence of LINC01089 overexpression on STARD13 expression (Additional file 3: Figure S3G). After that, the weakened cell proliferation ability induced by LINC01089 overexpression was rescued by silencing STARD13 or overexpressing miR-301b-3p (Fig. 4a, b). Besides, STARD13 silencing or miR-301b-3p overexpression could offset the facilitating effect on cell apoptosis caused by LINC01089 up-regulation (Fig. 4c). More importantly, wound healing and Transwell assays verified that STARD13 deficiency or miR-301b-3p overexpression counteracted the repressive influence of LINC01089 overexpression on cell migration (Fig. 4d, e). To sum up, LINC01089/miR-301b-3p/STARD13 axis hindered the progression of LUAD cells.