The human bronchial epithelial cell line 16HBE was purchased from Zhongqiaoxinzhou Biotech (Catalog Number: ZQ0001. Shanghai, China). 16HBE cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS, ScienCell, San Diego, CA, USA) at 37°C under 5% CO₂. A stable PLAC9-overexpressing cell line (16HBE-GFP-Plac9) and the corresponding stable control (16HBE-GFP) were custom-made by Genechem Co., Ltd, Shanghai, China.

Cells were randomly pooled into three groups. The three replicates were processed for iTRAQ analysis to identify differentially expressed proteins, as previously described with minor revisions (26, 27). Briefly, each sample was sonicated for 15 min in 500 μL SDT lysis buffer (4% sodium dodecyl sulfate (SDS) and 100mM Tris-HCl, pH 7.6). Samples were then incubated in water for 15 min at 95°C and centrifuged at 14,000 × g for 15 min. The supernatant was collected, and protein concentration was determined using the bicinchoninic acid (BCA) assay. Protein samples were then stored at -80°C until further analyses. 30 μL protein mixture was mixed with 1 M dithiothreitol (DTT) at a final concentration of 100 mM, and then incubated in water bath at 95°C for 5 min. After cooling to room temperature, the sample was mixed with 200 μL UA buffer (8 M urea and 150 mM Tris-HCl, pH 8.5), loaded onto an ultrafiltration filter (30-kDa cutoff, Sartorius, Germany), centrifuged at 12,500 × g for 25 min, and washed twice with UA buffer. Subsequently, 100 μL of iodoacetamide solution (100 mM iodoacetamide in UA buffer) was added to the filter, vortexed for 1 min at 600 rpm, incubated for 30 min at room temperature in the dark, and centrifuged at 12,500 × g for 25 min. The resulting filtrate was discarded. The filters were washed twice with 100 μL of UA buffer (12,500 × g, 15 min). Next, 100 μL of dissolution buffer (Applied Biosystems, Foster City, CA, USA) was added and centrifuged at 12,500 × g for 15 min. This step was repeated twice. Then, 40 μL of trypsin (Promega, Madison, WI, USA) buffer (5 μg trypsin in 40 μL dissolution buffer) was added to the filter. The filter was gently vortexed for 1 min at 600 rpm, incubated at 37°C for 16–18 h, transferred to a new tube, and centrifuged at 12,500 × g for 15 min. The digested peptides were collected with 20 μL of dissolution buffer, and the peptide concentration was measured using a Nanodrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA) at 280 nm.

A total of 100 μg of peptide mixture was labeled with iTRAQ reagents according to the manufacturer’s instructions (Applied Biosystems, Foster City, CA, USA). Triplicate 16HBE-GFP-Plac9 samples were labeled with the reagents 114, 115, and 116, and triplicate 16HBE-GFP samples were labeled with the reagents 117, 118, and 121. The labeled samples were mixed and analyzed using an Agilent 1260 Infinity II HPLC system (Agilent Technologies, Palo Alto, CA, USA). The column was equilibrated with buffer A (10 mM HCOONH₄, 5% (v/v) acetonitrile, pH 10.0), and the samples were separated in buffer B (10 mM HCOONH₄, 85% (v/v) acetonitrile, pH 10.0) gradient as follows: 0% (v/v) for 25 min, 0–7% (v/v) for 5 min, 7–40% (v/v) for 35 min, 40–100% (v/v) for 5 min, and 100% (v/v) for 15 min at a flow rate of 1 mL/min. Approximately 36 samples were collected, lyophilized, dissolved in 0.1% formic acid (FA), and pooled into 10 fractions for further mass spectrometry analysis.

Each fraction was separated on an Easy nLC system (Thermo Fisher Scientific, San Jose, CA, USA). Buffer C consisted of 0.1% (v/v) formic acid in MilliQ water, and buffer D consisted of buffer C with 80% (v/v) acetonitrile. The samples were loaded via an autosampler onto an analytical column (Acclaim™ PepMap™ RSLC 50 μm × 15 cm, nano viper (Thermo Fisher Scientific, San Jose, CA, USA) and separated at a flow rate of 300 nL/min.

Peptide analysis was performed on a Q-Exactive Plus mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA) in positive ion mode for 120 min, with a selected mass range of 350–1,800 mass/charge (m/z). For the survey scan, the first-order mass spectrum resolving power was set to 70,000, the automatic gain control (AGC) target value was set to 3E6, and the first-order maximum ion injection (IT) time was 50 ms. The m/z ratios of polypeptide and polypeptide fragments were obtained according to the methods described in the following sections. Ten-fragment mass spectral (MS2) scans were collected after each full scan. The MS2 activation type was a higher-energy collisional dissociation (HCD). The isolation window was set at 2 m/z. The second-order MS resolution was 17,500 with a 1 µscan. The second-order maximum IT time was 45 ms. The normalized collision energy was 30 eV.

Raw mass spectrometry (MS) data were derived from the RAW files. Database searches and quantitative analyses were performed using Mascot 2.6 and Proteome Discoverer 2.1 (Thermo Fisher Scientific, San Jose, CA, USA). Proteins were deemed to be differentially abundant if measured ratios exceeded a 1.2-fold change between the two sample groups analyzed and if the associated p value was below 0.05. The database used (Uniprot_HomoSapiens_161584_20180123) was downloaded from the UniProt website (http://www.uniprot.org) on 2018-01-23. The local sequence alignment software NCBI Basic Local Alignment Search Tool (BLAST 2.2.28+-win32.exe; http://blast.ncbi.nlm.nih.gov/Blast.cgi) was used to perform sequence alignment between the identified proteins and protein sequences in the UniProt database. The mapping function of the Blast2GO command line (www.geneontology.org; version go_201504.obo) was used to extract GO entries correlated with the sequences aligned for all the differentially expressed proteins identified. The webserver Kyoto Encyclopedia of Genes and Genomes (KEGG) Automatic Annotation Server (KAAS; http://www.genome.jp/tools/kaas/) was used to align the target proteins to the KEGG GENES database (28). The heatmap was generated using the online ClustVis tool (https://biit.cs.ut.ee/clustvis/) (29). The network of protein-protein interactions was mapped using the tool STRING 10.5, available online (http://string-db.org).

mRNA levels were determined using quantitative real-time PCR analysis using a QuantStudio^® 5 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. Briefly, total RNA was extracted from the cells using an RNA extraction kit (Bioteke, Beijing, China), and cDNA was synthesized using a cDNA synthesis kit (Thermo Fisher Scientific, San Jose, CA, USA). Then, 5 ng cDNA was amplified using SYBR Green Real-time PCR Master Mix (Toyobo, Osaka, Japan). The primers used are listed in Table 1 . The PCR cycling conditions were as follows: 95°C for 30 s, 45 cycles of 95°C for 5 s, and 60°C for 30 s. PCR amplification was followed by melting curve analysis using the default program of the QuantStudio^® 5 Real-Time PCR machine (Thermo Fisher Scientific, San Jose, CA, USA). The mRNA expression levels of the genes of interest were calculated using the 2^-ΔΔCt method and normalized using the expression levels of glyceraldehyde 3-phosphate dehydrogenase (GAPDH).

Cells were treated with 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma, St Louis, MO, USA) for the cell proliferation assay, as previously described (25, 30). Briefly, the cells were seeded at 1,000 cells/well in a 96-well tissue culture plate. Then, cells were incubated with 20 µL of MTT reagent for 4 h at 37°C. After adding 200 µL dimethyl sulfoxide (DMSO) to each well, the absorbance was quantified at 490 nm at the time of DMSO administration, and 24, 48, 96, and 120 h later. In the colony formation assay, approximately 800 cells were plated and cultured in 6-well plates, fixed with ethanol, and stained with 0.5% crystal violet for 2 weeks. The colonies were imaged and quantified.

Flow cytometry was performed as described previously (25, 30). Briefly, approximately 10⁶ 16HBE cells were pelleted and fixed with 70% ethanol for 1 h at 4°C. After resuspending with staining buffer (2 mg/mL of propidium iodide (PI) and 10 mg/mL RNase in 1 × PBS), cell cycle distribution was assessed using a BD FACS Calibur Flow Cytometer (BD Biosciences, San Jose, CA, USA) according to the manufacturer’s instructions. For the apoptosis assay, approximately 10⁶ 16HBE cells were harvested, resuspended, and labeled with fluorescein isothiocyanate (FITC)-Annexin V and PI in the dark at room temperature for 15 min. Apoptotic cells were analyzed using a BD FACS Calibur flow cytometer according to the manufacturer’s instructions.

Wound healing and invasion assays were performed as previously reported (31). 16HBE-GFP-Plac9 or 16HBE-GFP clones (4 × 10⁵) per well were seeded in a 12-well plate with complete medium (DMEM containing 10% FBS). When cells achieved 90% confluence, a wound was created by scraping the cell layer with a sterile pipette tip. The floating cells were removed by replacing the culture medium. Digital images of the wounds were captured immediately after replacing the medium, and after 12 and 24 h. The healing rate was calculated by comparing the wound width at different time points to the starting point.

For the cell invasion assay, 2 × 10⁵ 16HBE-GFP-Plac9 or 16HBE-GFP cells per well in 100 µL medium (DMEM supplemented with 1% FBS) were seeded in the top chambers of 24-well transwell plates (8.0 µm pore size; Corning Costar, Corning, NY, USA), and DMEM supplemented with 10% FBS was added to the bottom chamber. Cells were maintained at 37°C for 48 h. The cells on the top side of the filter were wiped out with a cotton swab, while those on the bottom side (those that had migrated) were fixed with 4% polyoxymethylene and stained with Giemsa staining solution (Yeasen, Shanghai, China). Five randomly selected fields per well were photographed, and the number of cells in five random fields per well was counted.

Cells were collected and lysed in protein lysis buffer (Beyotime, Shanghai, China) supplemented with Phenylmethanesulfonyl fluoride (PMSF, Beyotime, Shanghai, China) and protease inhibitor cocktail (Bimake, Houston, TX, USA). Protein samples (20 μg per lane) were separated using 10% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to nitrocellulose (NC) membranes. The membranes were blocked with 5% nonfat dry milk and incubated with the corresponding antibodies (Cell Signaling Technology, Beverly, MA, USA) according to the manufacturer’s instructions. The protein bands were visualized using enhanced chemiluminescence (ECL)™ reagents (Thermo Fisher Scientific, San Jose, CA, USA). β-Actin was used as an internal control.

Differences between groups were analyzed for statistical significance using Student’s t-test. GO or KEGG enrichment analyses were performed using Fisher’s exact test. All experiments were repeated at least three times. Results are expressed as mean ± standard deviation (SD). Statistical significance was set at p < 0.05.

To determine if PLAC9 expression is associated with LC, we screened the Oncomine database (www.oncomine.org). We observed that in three previously published LC cohorts (32–34), the expression of PLAC9 was significantly decreased in lung adenocarcinoma compared to paired adjacent normal tissues ( Figure 1 ), implicating a role of PLAC9 in lung epithelial pathogenesis.

To investigate the role of PLAC9 in LC, we evaluated the mRNA expression levels of PLAC9 in several of lung carcinoma cell lines, including NCI-H1299, A549, H1688, 95-D, MRC-5 as well as in the human bronchial epithelial cell line 16HBE ( Figure 2A ). The levels of PLAC9 mRNA were almost undetectable in human lung fibroblast cells MRC-5, and the rest of the cell lines had very low levels of PLAC9 mRNA. The expression level of PLAC9 mRNA in 16HBE cells was significantly higher than that in the lung cancer cell lines NCI-H1299, A549, H1688, and 95-D. The mRNA expression pattern of PLAC9 in normal lung cells compared with lung carcinoma cells was consistent with the expression of PLAC9 in normal lung tissue compared with lung carcinoma, as shown in Figure 1 .

To further explore the role of PLAC9 in human bronchial epithelial pathogenesis, we constructed a stable PLAC9-overexpressing cell line (16HBE-GFP-Plac9). For this purpose, we transfected the lentiviral Plac9 vector, which expresses both GFP and Plac9 under the control of the constitutive CMV promoter, into 16HBE cells. In addition, we generated a control line stably expressing only GFP, 16HBE-GFP, as previously described (25). Figure 2B shows a representative photograph of the stable cell lines generated (16HBE-GFP-Plac9 and 16HBE-GFP). Further qRT-PCR and western blot analyses showed that the 16HBE-GFP-Plac9 cells had significantly increased mRNA and protein levels of PLAC9 compared to the control 16HBE-GFP cells ( Figures 2C, D ).

To investigate the changes in protein expression caused by PLAC9 overexpression, iTRAQ quantitative proteomic technology was applied to analyze the protein samples from 16HBE-GFP-Plac9 and control 16HBE-GFP cells. The MS proteomics data were deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the iProX partner repository (35) with the dataset identifier PXD019147. A total of 279,431 spectra were generated, of which 125,694 spectra matched known peptides. Ultimately, 69,296 peptide fragments were identified, among which 6,841 unique proteins were detected and quantified. The peptide/protein distribution-based ion score, molecular weight, isoelectric point, peptide length, protein sequence coverage, peptide count, and ratio of expression levels between the two cell lines were analyzed and found to be consistent with a technically successful and reliable iTRAQ experiment ( Figure S1 ). Considering a threshold value of fold change ≥ ± 1.2 and p < 0.05, 714 proteins were found to have significantly different levels that those in PLAC9-overexpressing 16HBE cells ( Table S1 , see also the volcano plot in Figure S2 ). Among them, 405 proteins were upregulated, whereas 309 proteins were downregulated ( Table S1 ). Hierarchical cluster analysis of the proteins that differed in abundance in the three replicates of each cell line showed highly reproducible patterns ( Figure 3A ). The validity of iTRAQ was independently confirmed using western blot analysis of two of the proteins that presented the most significant differences. The proteins analyzed were gap junction protein delta 3 (GJD3), which was the most upregulated protein (18.4-fold increase, p value < 0.001) and leucine zipper like transcription regulator 1 (LZTR1), which was the most downregulated protein (0.21-fold decrease, p value < 0.001) ( Figure 3B ).

To gain insight into the functions of the proteins that were altered by PLAC9 overexpression, the differentially expressed proteins were categorized into three groups (biological process, molecular function, and cellular component) based on GO analysis. The differentially expressed proteins covered a wide range of biological processes, molecular functions, and cellular components, which could be classified into 26, 15, and 18 subcategory groups, respectively ( Figure 4 ). The largest group within the biological process category was that of cellular processes (654/668), followed by single-organism processes (612/668) and metabolic processes (553/668). Binding (628/662) and catalytic (315/662) activity were the most common categories for molecular function. The cellular component functions of these proteins were mainly related to the cell (675/685), organelles (651/685), membranes (463/685), and macromolecular complexes (389/685). The top 30 enriched GO categories are listed in Table 2 , and the total enriched GO categories are listed in Table S2 . These results showed that the predominant functions of the differentially expressed proteins were cellular processes, particularly cell migration.

To identify the biological pathways up- or down-regulated in PLAC9-overexpressing 16HBE cells, we conducted KEGG pathway-based analysis for the differentially expressed proteins and obtained 316 maps. Predictions for the most differentially expressed proteins suggest that they are involved in 20 pathways ( Table 3 ), all of which are listed in Table S3 . Among these pathways, RNA transport, RNA polymerase, mRNA surveillance pathways, all of which regulate gene expression, and the cell adhesion molecule (CAM) pathway, which regulates cell proliferation and migration, are notable. For example, among the regulated pathways, we found that CDK2 (cyclin-dependent kinase 2) was downregulated (0.9-fold change, p value < 0.05), while E-cadherin was upregulated (1.1-fold change, p value < 0.05), which indicated that overexpression of PLAC9 might influence cell proliferation and migration.

To further explore these hypotheses, the protein-protein interaction network was analyzed using the publicly available program STRING. The results are shown in Figure S3 and Table S4 . A total of 714 differentially expressed proteins determined from iTRAQ were analyzed using the molecular interaction tool. A total of 587 interactive proteins were identified. One quarter of the proteins identified were related to cell proliferation, cell cycle, and cell motility ( Figure 5 and Table S5 ). The protein interaction network indicated that PLAC9 might play an important role in several cellular processes, especially cell proliferation, cell cycle, and cell migration, further supporting the potential role of PLAC9 in cell proliferation and motility.

To investigate the role of PLAC9 in cell proliferation, MTT and colony formation assays were performed using 16HBE-GFP-Plac9 and 16HBE-GFP cell lines. The results of MTT assay showed that the proliferative capacity of 16HBE-GFP-Plac9 cells was significantly lower than that of 16HBE-GFP cells ( Figure 6A ). Furthermore, in vitro colony formation assays demonstrated that the frequency of colony formation of 16HBE-GFP-Plac9 cells was significantly lower than that of 16HBE-GFP cells ( Figures 6B, C ). These results indicated that PLAC9 plays an inhibitory role in 16HBE proliferation.

According to the iTRAQ data, the key regulator of the G1-S transition CDK2 was downregulated (0.9-fold, p value < 0.05). Consistently, flow cytometry showed that overexpression of PLAC9 altered the cell cycle distribution of 16HBE cells ( Figures 7A, B ). The number of cells in G1 phase was reduced, while the number of cells in S phase was significantly increased when PLAC9 was overexpressed. Furthermore, western blot analysis showed that the level of proteins promoting cell cycle progression, such as c-Myc, cyclin D3, cyclin E2, CDK2, and CDK4 were decreased in cells overexpressing PLAC9. Meanwhile, the protein levels of cell cycle inhibitors such as p21^Waf/cip1 and myt1 were increased in PLAC9-overexpressing cells ( Figure 7C ). In addition, M-phase entry repressors, including cyclin B1 and p-cdc2, were also increased. In contrast, overexpression of PLAC9 had no effect on apoptosis (data not shown).

As indicated above, the iTRAQ data suggest that PLAC9 affects cell migration. To explore this hypothesis, we performed two cell motility assays (wound healing and cell invasion assay) as previously described elsewhere (36). As shown in Figures 8A, B , the wound-healing assay showed that wound closure after 48h in 16HBE-GFP-Plac9 cells was more pronounced than that in 16HBE-GFP cells, indicating that PLAC9 overexpression strongly enhanced cell migration. Next, we used a Transwell^® chamber to perform the cell invasion assay ( Figures 8C, D ). The number of invading cells was significantly higher in 16HBE-GFP-Plac9 cells than in control cells. The results indicate that overexpression of PLAC9 markedly increased cellular motility.

It is well known that E-cadherin, N-cadherin, and vimentin are key regulators of embryonic development, organ morphogenesis, and tumor growth (37, 38), and are involved in epithelial cell motility via EMT (39–41). In PLAC9-overexpressing 16HBE cells, E-cadherin levels were significantly higher than those in the control, which was consistent with the iTRAQ data ( Figure 8E ), while vimentin was inhibited in PLAC9-overexpressing 16HBE cells. N-cadherin levels were not significantly different between the two cell lines analyzed. Taken together, these results indicated that the overexpression of PLAC9 facilitated cell motility despite increased E-cadherin and reduced vimentin protein levels.