LINC01936 inhibits the proliferation and metastasis of lung squamous cell carcinoma probably by EMT signaling and immune infiltration

RNA-seq data acquisition and preprocessing

The RNA-seq expression profiles of LUSC were downloaded from The Cancer Genome Atlas (TCGA, Version: V31.0). The TCGA-LUSC dataset contained an RNA expression profile from 502 LUSC tissues and 49 adjacent normal tissues. The lncRNA microarray dataset (GSE88862) was obtained from the Gene Expression Omnibus (GEO), which included three LUSC tissues and three adjacent normal tissues (Cheng et al., 2017).

Differential expression analysis

Differential expression analysis of lncRNAs between LUSC tissues and normal tissues were conducted using the DEseq2 (Love, Huber & Anders, 2014), edgeR (Robinson, McCarthy & Smyth, 2010) and limma (Ritchie et al., 2015) R packages. For the GEO dataset, the online analytical tool GEO2R was used to identify differentially expressed lncRNAs. The cut-off criteria were |log₂FoldChange| > 2 and adj.p–value < 0.05. VENNY 2.1.0 was employed to draw Venn diagrams.

Differential expression analysis of LINC01936 in LUSC

The expression data of LINC01936 in LUSC tissues and normal tissues were extracted and processed by the Perl program (Version: V5.30.2). For the LINC01936 expression matrix from 502 LUSC tissues and 49 adjacent normal tissues, the beeswarm and limma packages were utilized to differentiate the expression level of LINC01936 between the tumor group and the normal group.

Prediction of subcellular localization of LINC01936

The lncLocator was used to predict the subcellular localization of lncRNA (Cao et al., 2018). The fasta sequence of LINC01936 (NR_122048.1) was downloaded and then uploaded to IncLocator. The five subcellular localizations (ribosomes, cytoplasm, cytosol, exosomes and nucleus) of LINC01936 were predicted.

Cell lines and culture conditions

LUSC cell lines (NCI-H1703, KNS-62) were purchased from Procell (Hubei, China). NCI-H1703 cells were cultured in RPMI 1640 medium (Gibco, Waltham, MA, USA) and KNS-62 was cultured in DMEM/High Glucose (Eallbio, Beijing, China) in a humidified incubator at 37 °C with 5% CO₂. All medium supplemented with 10% Fetal Bovine Serum (FBS; Tianhang Biological, Zhejiang, China) and 1% Penicillin-Streptomycin (Solarbio, Beijing, China).

The synthesis and transfection of overexpression plasmid and siRNA

The full-length cDNA of LINC01936 was subcloned into the pcDNA3.1 vector to generate the pcDNA3.1/LINC01936 plasmids by Sangon Biotechnology (Shanghai, China) and verified by direct sequencing using the primer 5′- TGGGAGGTCTATATAAGCAGAG-3′(F). The empty pcDNA3.1 was used as negative control. The target sequence for LINC01936-specific small interfering RNA (siLINC01936; 5′-GGCAGACAUCCCACGGAAATT-3′) and control siRNA (siNC; 5′-UUCUCCGAACGUGUCACGUTT-3′) were synthesized by GenePharma (Shanghai, China). The vectors and siRNA were transfected into LUSC cells by using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA, USA).

RNA Extraction and Reverse Transcription-quantitative PCR (RT-qPCR)

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and 2 ug RNA was reverse transcribed into cDNA by using Revert Aid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Grand Island, NY, USA) in accordance with the manufacturer’s instructions. LncRNA expression was measured using the ABIPRISM® 7900HT Fast Real-Time PCR system (Applied Biosystems, Foster City, CA, USA), and the reaction was performed using the Power SYBR Green PCR Master Mix (Thermo Fisher Scientific, Grand Island, USA). The cycling program was as follows: pre-denaturation at 95 °C for 2 min, followed by 45 cycles of denaturation at 95 °C for 15 s, annealing for 2 min, and final extension at 60 °C for 1 min. Three replicates per reaction were used in RT-qPCR, and the results were analyzed using the 2^−ΔΔCt method. GAPDH was used as an internal reference. All primers for RT-qPCR amplification of candidate genes were listed in Table 1.

Cell proliferation analysis

After transfected with LINC01936 overexpression plasmids or siRNA for 24 h, LUSC cells transfected for 24 h were seeded onto a 96-well plate at the density of 5 × 10³ cells/well and cultured for 24, 48 or 72 h. For Cell Counting Kit-8 (CCK8) assay, 10 μL of CCK8 (Dojindo Molecular Technologies, Kumamoto, Japan) reagent was added to each well and the cells were incubated at 37 °C for 2 h. The optical density (OD) value was measured in an Infinite M200 PRO Microplate Reader (Tecan, Männedorf, Switzerland). For MTT assay, LUSC cells (2.5 × 10³/well) was seeded on 96-well plate and incubated. After 24, 48 or 72 h, 100 μL MTT (Mackline, Shanghai, China) solution (0.5 mg/mL) was added to each well and incubated for 4 h. The supernatant was then sucked away and DMSO was added to melt the crystal. The OD₄₉₀ was measured using a microplate reader. Each assay was performed in triplicate.

Apoptosis assay

The nuclear morphology of the apoptotic cells was visualized by Hoechst 33342 staining. After transfection with LINC01936 overexpressed plasmids for 24 h or si-LINC01936 for 48 h, NCI-H1703 cells were fixed with 4% paraformaldehyde (biosharp, Anhui, China) for 30 min and then stained with Hoechst 33342 solution (Solarbio, Beijing, China) for 20 min. The cells were observed under an IX71 inverted fluorescence microscope (Olympus, Tokyo, Japan). The apoptotic cells were quantified by Image J software (version 1.52a).

Migration and invasion assays

Transwell chambers (Corning Incorporated, Corning, NY, USA) with or without Matrigel were used for cell invasion and migration. LUSC cells (5 × 10⁴) were added to the upper chambers in a total volume of 100 μL medium and 600 μL culture medium with 20% FBS was added to the lower chambers. LUSC cells were incubated for 24 h in migration assay and 48 h in invasion assay. The cells on upper chamber were swiped using cotton swabs and the cells on lower chamber were fixed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet (Solarbio, Beijing, China) for 15 min. The migrated and invaded cells were counted using an IX71 inverted fluorescence microscope.

Functional and pathway enrichment analysis and targets prediction of lncRNA

Pearson correlation coefficient was used to screen the mRNAs co-expressed genes with LINC01936, with |r| > 0.65 and p < 0.05 as the thresholds criteria. The ggplot2 R package and the DAVID 6.8 database were used for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Based on the median of the expression of LINC01936, LUSC samples were divided into high- and low-expression groups. The gene set variation analysis (GSVA) was carried out to study the pathways related to LINC01936 using GSVA R package. The gene set “2.cp.kegg.v6.2.symbols.gmt” from Molecular Signature Database served as the reference. The scores of LINC01936 binding to co-expression genes were calculated using RNA-Protein Interaction Prediction (RPISeq) (http://pridb.gdcb.iastate.edu/RPISeq/) (Muppirala, Honavar & Dobbs, 2011). Interaction probabilities generated by RPISeq. The probabilities (RF and SVM scores) >0.8 were considered as interacting gene.

Tumor-infiltrating immune analysis

Spearman correlation analysis was performed using ggpubr package in the R platform to analyze the correlation between the expression of LINC01936 and the abundance of immune cell (T cells, Natural Killer cells, Macrophages, Dendritic cells, CD8+ T cells and B cells) infiltration in LUSC. RNA-seq data was used for estimating the relative infiltration abundance of different immune cells.

Statistical analyses

Statistical analysis and visualization of results were conducted using SPSS (Version: 22.0; SPSS Inc., Chicago, IL, USA) and GraphPad Prism (Version: 8.0; GraphPad Software, La Jolla, CA, USA). The difference in the expression of LINC01936 between the tumor and normal groups was analyzed using the Wilcoxon rank-sum test. Differences of LINC01936 expression between the overexpressed/silencing group and the control group were analyzed using Student’s t-test. p < 0.05 was considered statistically significant.

Identification of differentially expressed lncRNAs in LUSC

By using DESeq2, edgeR, and limma packages, 2,139 (1,603 upregulated and 536 downregulated), 2,138 (1,702 upregulated and 436 downregulated), and 1,159 (529 upregulated and 630 downregulated) differentially expressed lncRNAs were identified in the TCGA-LUSC dataset, respectively (Figs. 2A–2C). For GEO GSE88862 dataset, 341 differentially expressed lncRNAs (206 upregulated and 135 downregulated) were identified (Fig. 2D). The intersection analysis of the two datasets revealed 52 co-differentially expressed lncRNAs, with 29 upregulated and 23 downregulated (Figs. 2E and 2F). Among these differentially expressed genes, the top four down-regulated lncRNAs were SMIM25, FENDRR, SFTA1P, and LINC01936 (Table 2) and the top four up-regulated ones were LINC01133, CASC9, SOX21-AS1, and ESRG (Table 3). After searching the literatures, we found that seven genes, except for LINC01936, were all contributed to the development of lung cancer by several pathways (Chang et al., 2020; Gao et al., 2019; Jafari et al., 2022; Jiang et al., 2020; Ning et al., 2021; Wang et al., 2021a; Xiong et al., 2019). Due to this, LINC01936 was used for further analysis.

The expression and localization of LINC01936 in LUSC

The expression of LINC01936 in LUSC tissues was significantly downregulated in LUSC tissues as compared to that in normal tissues (p < 0.001, Fig. 3A). In addition, the subcellular localization of LINC01936 was predicted using lncLocator, and result showed that LINC01936 was mainly localized in the cytoplasm and organelles with a total score of 0.89 (Fig. 3B).

The effect of LINC01936 on the proliferation and metastasis of LUSC cells

To study the biological function of LINC01936 in LUSC cells, we measured the RNA expression in LINC01936 was overexpressed or knockdown by transfecting with pcDNA3.1/LINC01936 plasmid and siLINC01936, respectively. LINC01936 and siLINC01936 were successfully transfected into LUSC cells (Figs. 3C and 3D). The effect of LINC01936 on cell proliferation and viability was evaluated by performing CCK-8 assay and MTT assay. The results indicated that LINC01936 overexpression reduced cell proliferation and viability (Figs. 4A–4D) and the silence of LINC01936 promoted cell proliferation and viability (Figs. 5A–5D). Transwell assays were conducted to investigate the effects of LINC01936 on cell migration and invasion. The results showed that the overexpression of LINC01936 significantly reduced the migration and invasion abilities of LUSC cells (Figs. 6A and 6B). In contrast, the silence of LINC01936 enhance the cell migration and invasion (Figs. 6C and 6D). Taken together, these results suggest that LINC01936 suppresses the proliferative and metastatic abilities of LUSC cells.

LINC01936 promotes the apoptosis of LUSC

To investigate the effect of LINC01936 on cell apoptosis, nuclear fluorescence staining was performed using Hoechst33342. Results showed that the number of bright blue regions, which indicating condensed DNA and fragmented nuclei, was significantly higher in pcDNA3.1/LINC01936 transfected group than in control group (p < 0.01, Figs. 7A and 7B) and the silence of LINC01936 inhibited cell apoptosis (Figs. 7C and 7D). These findings confirmed that LINC01936 enhanced the apoptosis of LUSC cells.

Functional enrichment and targets of LINC01936 analysis

GO analysis revealed significant enrichment of LINC01936 in biological processes related to cell matrix adhesion, regulation of endothelial cell proliferation and regulation of angiogenesis (Fig. 8A). The cell components were mainly concentrated in focal adhesion, bicellular tight junction and collagen trimer (Fig. 8B). Molecular functions were significantly enriched in extracellular matrix structural constituent, integrin binding and growth factor binding (Fig. 8C). KEGG pathway enrichment analysis indicated that LINC01936 was mainly involved in cell adhesion molecules, vascular smooth muscle contraction, cGMP−PKG signaling pathway and leukocyte transendothelial migration (Fig. 8D). To further investigate the potential mechanism of LINC01936 in LUSC, Gene Set Variation Analysis (GSVA) was performed on the TCGA-LUSC RNA-seq data (Fig. 8E). The high expression of LINC01936 significantly deactivation the epithelial mesenchymal transition (EMT).

Next, the downstream targets of LINC01936 were predicted. Prior studies suggested that lncRNAs in the cytoplasm can bind to cytoplasmic RNAs to affect the expression patterns of downstream targets (Montes et al., 2015; Suravajhala et al., 2015). Correlation analysis showed that LINC01936 was correlated with the expression of transcription factor 21 (TCF21) (r = 0.758, p < 0.001), sodium voltage-gated channel alpha subunit 7 (SCN7A) (r = 0.752, p < 0.001), amine oxidase copper containing 3 (AOC3) (r = 0.738, p < 0.001), C1q and TNF related 7 (C1QTNF7) (r = 0.734, p < 0.001), Cas scaffold protein family member 4 (CASS4) (r = 0.711, p < 0.001), phospholipase A2 group V (PLA2G5) (r = 0.697, p < 0.001), microfibril associated protein 4 (MFAP4) (r = 0.696, p < 0.001), gap junction protein alpha 5 (GJA5) (r = 0.684, p < 0.001), RAS like family 12 (RASL12) (r = 0.678, p < 0.001), mesenchyme homeobox 2 (MEOX2) (r = 0.677, p < 0.001), indolethylamine N-methyltransferase (INMT) (r = 0.675, p < 0.001), T-box transcription factor 5 (TBX5) (r = 0.668, p < 0.001), Heat shock protein B7 (HSPB7) (r = 0.666, p < 0.001), and podocan (PODN) (r = 0.659, p < 0.001) in LUSC tissues (p < 0.001, Fig. 9A). Further interaction analysis presented that LINC01936 might bind to TCF21, AOC3, RASL12, MEOX2 and HSPB7 with RF and SVM scores greater than 0.8 (Figs. 9B and 9C).

LINC01936 is correlated with immune cells infiltrating

To further identify the specific cell types that played a major role in the process of LUSC, the potential association between LINC01936 expression and immune cell infiltration was analyzed based on TCGA-LUSC transcriptomic data. The results showed that LINC01936 was highly expressed in all infiltrating immune cells (Fig. 10A) and the expression of LINC01936 was significantly correlated with T cells (r = 0.30, p < 0.05), Natural Killer cells (r = 0.11, p < 0.05), Macrophages (r = 0.45, p < 0.05), Dendritic cells (r = 0.34, p < 0.05), CD8+ T cells (r = 0.24, p < 0.05) and B cells (r = 0.30, p < 0.05) in LUSC (Fig. 10B). LINC01936 expression was also significantly correlated with the proportion of infiltrating immune cells, which may affect the occurrence and development of LUSC.