The FD-LSC-1 human LSCC cell line [a gift from Professor Liang Zhou (20)] was cultured in BEGM™ Bronchial Epithelial Cell Growth Medium (Lonza Group Ltd.) supplemented with 10% fetal bovine serum (Biological Industries). 293T cells (China Center for Type Culture Collection) and the TU-177 human LSCC cell line (Shanghai Bioleaf Biotech Co., Ltd.) were cultured in DMEM supplemented with 10% fetal bovine serum. The cultures were maintained at 37°C in 5% CO₂.

Human prostaglandin reductase 1 (PTGR1), DHCR24 and solute carrier family 38 member 2 (SLC38A2) overexpression plasmids were generated by inserting their respective CDS sequence into the p3xFLAG-CMV-10 vector (Merck KGaA). The FSCN1 overexpression plasmid was generated by inserting the FSCN1 coding sequence into the pCMV-HA vector (Clontech; Takara Bio USA, Inc.). Empty p3xFLAG-CMV-10 vector was used as a negative control. For DHCR24 overexpression, FD-LSC-1 and TU-177 cells were seeded in six-well plates at a density of 5×10⁵ cells/well. A total of 5 μl Lipofectamine 3000^® (Invitrogen; Thermo Fisher Scientific, Inc.) and 2.5 μg DHCR24 overexpression vector or empty p3xFLAG-CMV-10 vector were mixed in 250 μl Opti-MEM™ (Gibco; Thermo Fisher Scientific, Inc.) and added to each well. Following 6 h of incubation at 37°C with 5% CO₂, the medium was replaced with fresh medium. At 48 h following transfection, the cells were collected for use in subsequent experiments.

siRNA targeting FSCN1 (si-FSCN1) and scrambled siRNA (si-NC) were synthesized by GenePharma Co., Ltd. The siRNA sequences were as follows: si-FSCN1 sense 5′-GCAAGAAUGCCAGCUGCUACU-3′ and antisense 5′-AGUAGCAGCUGGCAUUCUUGC-3′; and si-NC sense, 5′-UUCUCCGAACGUGUCACGUTT-3′ and antisense, 5′-ACGUGACACGUUCGGAGAATT-3′. For transfection, TU-177 cells were seeded in six-well plates at a density of 5×10⁵ cells/well. A total of 5 μl Lipofectamine 3000^® (Invitrogen; Thermo Fisher Scientific, Inc.) and 100 nM siRNAs were mixed in 250 μl Opti-MEM™ (Gibco; Thermo Fisher Scientific, Inc.) and added to each well. Following 6 h of incubation at 37°C with 5% CO₂, the medium was replaced with fresh medium. At 48 h following transfection, the cells were collected for use in subsequent experiments.

TU-177 cells in which FSCN1 was knocked down and control TU-177 cells were generated using siRNA transfection, followed by total RNA extraction using TRIzol^® reagent (Thermo Fisher Scientific, Inc.). Total RNA was quantified using a NanoDrop ND-2000 spectrophotometer (Thermo Fisher Scientific, Inc.), and the RNA integrity was assessed using Agilent Bioanalyzer 2100 (Agilent Technologies, Inc.). Sample labeling, microarray hybridization and washing were performed according to the manufacturer's instructions (Agilent Human lncRNA Microarray 2018 Version, 4×180k, Design ID: 085630). Briefly, total RNA was transcribed into double-stranded cDNA (RNA Spike In Kit, one-color, Agilent p/n 5188-5282; Agilent Technologies, Inc.), synthesized into cRNA (RNA Spike In Kit, one-color, Agilent p/n 5188-5282; Agilent Technologies, Inc.), and labeled with Cyanine-3-CTP (Low Input Quick-Amp Labeling Kit, one-color. Agilent p/n 5190-2305; Agilent Technologies, Inc.). The labeled cRNAs were hybridized onto the microarray. Following washing, the arrays were scanned using the Agilent Scanner G2505C (Agilent Technologies, Inc.).

Feature Extraction software (version 10.7.1.1, Agilent Technologies) was used to analyze array images to obtain raw data, and GeneSpring (version 13.1, Agilent Technologies) was then used to complete the basic analysis of the raw data. First, the raw data were normalized using the quantile algorithm. The probes that at least 1 out of 2 conditions have flags in 'P' were selected for further data analysis. Differentially expressed genes (DEGs) or long non-coding RNAs (lncRNAs) were identified using fold change (FC) and the P-value from the unpaired t-test. The thresholds for up- and downregulated genes were a fold change of ≥1.5 and P≤0.05. Hierarchical clustering was performed to display the expression patterns of distinguishable genes among samples. The microarray assay data were deposited at the Gene Expression Omnibus (GEO) database and are accessible via the accession no. GSE255143.

The Database for Annotation, Visualization and Integrated Discovery (DAVID) Bioinformatics Resources (http://david.ncifcrf.gov) was used for functional enrichment analysis of GO terms and KEGG pathways. P<0.05 was selected as the cut-off criterion. The GO terms included biological process, cellular component and molecular function. The top 20 GO terms and KEGG pathways were plotted according to their P-values.

The RNA sequencing data of 57 pairs of LSCC and matched adjacent normal mucosa tissues (GSE127165) were downloaded from the GEO database. Of the 1,063 DEGs in the cells in which FSCN1 was knocked down, 641 were expressed in the 57 LSCC tissues. The co-expression between FSCN1 and the 641 DEGs was then calculated using Pearson's correlation analysis, and Pearson's correlation coefficients of <0.4 or >-0.4 were eliminated. Both up- and downregulated DEGs that were positively and negatively correlated with FSCN1, respectively, were eliminated.

The identified FSCN1-regulated genes were analyzed using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) online tool v11.5 (http://www.string-db.org) for interaction networks. The PPI network was generated with a confidence score of ≥0.7 (high confidence). The highly connected clusters from the PPI network were identified using the 'Molecular Complex Detection (MCODE)' plugin tool in Cytoscape v3.7.1 software. The top three high ranking clusters were selected for further analysis.

293T cells were cultured in 60-mm plates and each plate was transfected with 5 μg FSCN1 overexpression vector and 5 μg PTGR1 or DHCR24 or SLC38A2 overexpression vector using Lipofectamine 3000^® (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C with 5% CO₂ for 6 h, according to the manufacturer's instructions. At 48 h following transfection, the cells were collected for use in subsequent co-IP experiments. Whole-cell extracts were collected by lysing 1×10⁷ cells in Pierce IP Lysis buffer (Thermo Fisher Scientific, Inc.) with the addition of a complete protease inhibitor mixture for 30 min on ice, and clarified using centrifugation at 12,000 x g for 15 min at 4°C. For co-IP, protein extracts were incubated at 4°C overnight with mouse monoclonal anti-flag antibody (1:500; cat. no. F1804; Merck KGaA) or control mouse immunoglobulin G (IgG) (1:500; cat.no. A7028; Beyotime Institute of Biotechnology), and precipitated proteins were captured using Pierce Protein A/G magnetic beads (Thermo Fisher Scientific, Inc.). Following three washes with Pierce IP lysis buffer (Thermo Fisher Scientific, Inc.), bound proteins were eluted in 2X SDS loading buffer and examined using western blot analysis. Rabbit HA tag antibody (1:1,000; cat. no. 51064-2-AP; Proteintech Group, Inc.) was used for western blot analysis.

The cells were lysed using Pierce RIPA lysis buffer (Thermo Fisher Scientific, Inc.) containing protease inhibitor cocktail for 30 min on ice. The protein concentration was determined using a BCA Protein Assay kit (Thermo Fisher Scientific, Inc.), and total protein was separated using 10% SDS-PAGE gels, transferred to a PVDF membrane and blocked with 10% w/v non-fat milk powder in TBST at room temperature for 2 h. The membranes were incubated with primary antibodies against Flag (1:1,000; cat. no. F1804; Merck KGaA), FSCN1 (1:1,000; cat. no. 54545S; Cell Signaling Technology, Inc.), and GAPDH (1:1,000; cat.no. HC301-02; TransGen Biotech Co., Ltd.)overnight at 4°C, followed by incubation with appropriate horseradish peroxidase-conjugated secondary antibodies (1:1,000; cat. no. A0216; Beyotime Institute of Biotechnology) at room temperature for 2 h. Immunoreactive bands were visualized using WesternBright^® ECL HRP substrate (Advansta Inc.).

Cell proliferation was determined using the Cell Counting Kit-8 (TransGen Biotech Co., Ltd.), following the manufacturer's instructions. Briefly, 3,000 cells were seeded into each well of a 96-well plate. At 0, 12, 24, 48 and 72 h after seeding, each well was replaced with 100 μl fresh complete medium and 10 μl CCK-8 (TransGen Biotech Co., Ltd.) followed by incubation at 37°C with 5% CO₂ for 1 h. The absorbance of the solution was measured at 450 nm using a Spectra Max i3x Multifunctional microplate detection system (Molecular Devices, LLC). A total of three independent experiments were performed.

Following transfection, the cells were suspended in a serum-free medium. Serum-free DMEM (200 μl) containing 1×10⁵ cells were added to the upper chamber. A total of 500 μl DMEM medium supplemented with 20% fetal bovine serum (Biological Industries) was then added to the lower chamber. At 48 h following incubation in 37°C with 5% CO₂, the cells in the upper chamber were removed with cotton swabs and the lower side of the chamber was gently washed twice with phosphate-buffered saline (PBS), and fixed with 4% paraformaldehyde for 20 min at room temperature. The cells stained with 0.1% crystal violet (Amresco, LLC) for 10 min at room temperature, and images were then captured using an inverted microscope (Leica Microsystems GmbH).

Total RNA was isolated from the cells in which FSCN1 was knocked down or control cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.), following the manufacturer's instructions. For RT-qPCR of messenger RNA (mRNA), cDNA was synthesized using the HiScript II 1st Strand cDNA Synthesis kit (Vazyme Biotech Co., Ltd. according to the following conditions: 25°C for 5 min, 50°C for 15 min, and 85°C for 2 min. ChamQ SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd.) was used for qPCR. The following thermocycling conditions were used: 95°C for 30 sec, followed by 40 cycles at 95°C for 5 sec and 60°C for 10 sec. The relative expression level of target genes was calculated using the 2(−ΔΔCq) method (21). The primer sequences used are listed in Table SI.

Statistical analysis was performed using SPSS software v22.0 (IBM Corp.). The data were analyzed using the two-tailed Student's t-test; P<0.05 was considered to indicate a statistically significant difference.

To investigate genes regulated by FSCN1, the microarray analysis of differentially up- or downregulated genes was performed by knocking down FSCN1 using siRNAs in TU-177 cells. The results of RT-qPCR indicated that FSCN1 was successfully knocked down (Fig. 1A). Following data preprocessing, 1,063 DEGs (462 up- and 601 downregulated) were identified in the TU-177 cells transfected with FSCN1 siRNAs compared with those transfected with the control under the thresholds of FC>1.5 and P<0.05. Based on the volcano plot and hierarchical clustering analysis of these DEGs, the control and FSCN1 knockdown groups were clearly distinguished (Fig. 1B and C). In addition, 436 differentially expressed lncRNAs (118 up- and 318 downregulated) and 198 differentially expressed circular RNAs (circRNAs) (44 up- and 154 downregulated) were obtained between the FSCN1 knockdown and control groups (Fig. 1D). These genes are listed in Table SII. In addition, the expression of the top 10 up- and downregulated genes in TU-177 cells in which FSCN1 was knocked down compared with the control cells based on microarray data are shown in Fig. 1E and F.

To identify the putative functions and pathways associated with the FSCN1-regulated genes, functional enrichment analysis was performed through the DAVID Bioinformatics Resources (22). The upregulated DEGs were significantly enriched with biological functions of 'cilium assembly', 'regulation of transcription from RNA polymerase II promoter' and 'regulation of transcription, DNA-templated' (Fig. 2A). By contrast, the downregulated DEGs were associated with 'nervous system development', 'sterol biosynthetic process' and 'response to radiation' (Fig. 2B). In the cellular components category, 'cilium' and 'cytoplasm' were significantly enriched using the upregulated DEGs, whereas 'integral component of membrane' and 'plasma membrane' were involved with the downregulated DEGs (Fig. 2A and B). For the GO molecular functions category, the upregulated DEGs were enriched with functions, such as 'metal ion binding', 'RNA polymerase II core promoter proximal region sequence-specific DNA binding' and 'RNA polymerase II transcription factor activity, sequence-specific DNA binding' (Fig. 2A). The downregulated DEGs were associated with 'calcium ion binding', 'endopeptidase inhibitor activity' and 'oxidoreductase activity' (Fig. 2B). According to the findings of KEGG pathway analysis, the upregulated DEGs demonstrated significant enrichment in the pathways of 'herpes simplex virus 1 infection', 'focal adhesion' and 'ECM-receptor interaction' (Fig. 2C). By contrast, the downregulated DEGs were involved in 'steroid biosynthesis', 'metabolic pathways' and 'biosynthesis of cofactors' (Fig. 2D). The complete results are presented in Table SIII.

RNA sequencing (RNA-seq) was previously performed in 57 pairs of LSCC and matched adjacent normal mucosa tissues to construct mRNA expression profiles (23). Based on the aforementioned RNA-seq data and 1,063 DEGs in the cells in which FSCN1 was knocked down, a network of genes co-expressed with FSCN1 in LSCC was constructed using the flowchart illustrated in Fig. 3A. Finally, 48 DEGs (10 up- and 38 downregulated DEGs) were found to be co-expressed with FSCN1 in LSCC with Pearson's correlation coefficients of 0.4. A network was constructed and visualized using Cytoscape software (Fig. 3B). In addition, functional enrichment analysis revealed that the 48 genes co-expressed with FSCN1 were significantly associated with the functions of 'defense response to virus', 'response to virus', 'cellular response to interferon-α', 'innate immune response' and 'neutral amino acid transport' (Fig. 3C). KEGG pathway analysis demonstrated that 'biosynthesis pathways of cofactors' and 'metabolic pathways' were significantly enriched by genes co-expressed with FSCN1 (Fig. 3D). The complete results are presented in Table SIV.

To explore the interactions among these FSCN1-regulated genes and identify specific functional complexes, a PPI network was constructed using the STRING database. In the PPI network, the minimum required interaction score was set as >0.700 (high confidence), and disconnected nodes in the network were hidden. A PPI network containing 455 nodes and 570 edges was constructed (Fig. 4A). In addition, three highly connected clusters were identified from the PPI network using the MCODE plugin tool (24) in Cytoscape (Fig. 4B-D). The GO and KEGG enrichment analyses revealed that Cluster 1 was mainly involved with functions of defense response to virus (GO:0051607) and response to virus (GO:0009615). By contrast, Cluster 2 was associated with the function of cholesterol biosynthetic process (GO:0006695) and the KEGG pathway of steroid biosynthesis (hsa00100). The genes in Cluster 3 were mainly associated with cilium assembly (GO:0060271).

Subsequently, RT-qPCR was used to validate the expression of genes involved in the function of defense response to virus and steroid biosynthesis pathway. Compared with the control group, interferon induced protein with tetratricopeptide repeats (IFIT)2, IFIT3, 2′-5′-oligoadenylate synthetase like (defense response to virus; QASL), squalene epoxidase (SQLE), farnesyl-diphosphate farnesyltransferase 1 (FDFT1) and DHCR24 (steroid biosynthesis) were significantly downregulated in the LSCC cells in which FSCN1 was knocked down (Fig. 4E and F), indicating that FSCN1 affected these functions and pathways.

FSCN1-interacting proteins were previously characterized in LSCC cell lines using a mass spectrometry-based proteomics approach, and 238 proteins were identified as interacting partners of FSCN1 in TU-177 cells (25). Following Venn analysis, seven overlapped target genes [DHCR24, SLC38A2, PTGR1, peroxiredoxin 4 (PRDX4), tropomyosin 4 (TMP4), abhydrolase domain containing 16A (ABHD16A), phospholipase and capping protein regulator and myosin 1 linker 1 (CARMIL1)] were identified (Fig. 5A). Of note, it was found that a proven FSCN1-regulated gene (DHCR24; Fig. 4E and F), interacted with FSCN1 in TU-177 cells. SLC38A2, a gene co-expressed with FSCN1 in LSCC (Fig. 3B) interacted with FSCN1 in TU-177 cells. Furthermore, another DEG (PTGR1) in cells in which FSCN1 was knocked down was identified as an FSCN1-interacting protein in Hep-2 cells (25). RT-qPCR was performed to verify the relative expression levels of PTGR1 and SLC38A2 in LSCC cells in which FSCN1 was knocked down. As shown in Fig. 5B, SLC38A2 was significantly downregulated in LSCC cells in which FSCN1 was knocked down, whereas the expression of PTGR1 was not significantly altered.

The present study then attempted to validate the interaction between FSCN1, and PTGR1, DHCR24 and SLC38A2 through co-IP. The three genes were cloned into vectors with a 3xFlag tag, and the FSCN1 gene was cloned into vectors with an HA tag. FSCN1 and Flag-tagged proteins were transiently co-expressed in 293T cells. The expression of all three Flag-tagged proteins and HA-tagged FSCN1 was successfully confirmed using western blot analysis (Fig. 5C). Co-IP was performed on whole cell extracts with Flag-tag antibodies, with matched normal IgG used as a negative control. Out of these three proteins, two were confirmed to bind to FSCN1 (Fig. 5C), and one protein (SLC38A2) did not interact with FSCN1.

As DHCR24 binds to the FSCN1 protein and its expression is regulated by FSCN1, the expression levels of DHCR24 in the FD-LSC-1 and TU-177 LSCC cell lines were first investigated. Compared with the normal cell line (293T), DHCR24 was significantly downregulated in the FD-LSC-1 and TU-177 LSCC cell lines (Fig. 6A). To further investigate the functional role of DHCR24 in LSCC cells, the DHCR24 overexpression vector (OE-DHCR24) and its empty vector (OE-NC) were transfected into FD-LSC-1 and TU-177 cells. The efficiency of DHCR24 protein overexpression was further verified using western blot analysis (Fig. 6B). Compared with the negative control group, DHCR24 overexpression promoted LSCC cell proliferation (Fig. 6C and D). Furthermore, the results of Transwell assay demonstrated that DHCR24 overexpression markedly promoted LSCC cell migration (Fig. 6E). These findings suggested that DHCR24 promotes the proliferation and migration of LSCC cells.