TCGA-LUAD cohort data and corresponding clinical information of 535 LUAD patients were downloaded from The Cancer Genome Atlas (TCGA) website (https://portal.gdc.cancer.gov/repository). LUAD patients were classified into low- and high-AC02278.4 expression groups according to the median AC02278.4 expression value. AC02278.4 expression data from GSE81089 datasets were downloaded from the Gene Expression Omnibus (GEO) database and validated for expression analyses. The gene expression profiles were normalized using the scale method provided in the “limma” R package. Data analysis was performed with the R (version 3.6.3) and ggplot2 [3.3.3] packages. The expression data were normalized to transcripts per kilobase million (TPM) values before further analysis. Besides, the receiver operating characteristic (ROC) curve was used to evaluate the diagnostic value of AC02278.4 using the pROC R package and ggplot2 R package.

Based on the multivariate Cox analysis results, a nomogram was established to predict the prognosis of LUAD patients. According to the prognosis model, each patient’s risk score was calculated as the total score of each parameter, which could predict the prognosis of LUAD patients (18).

The gene set enrichment analysis (GSEA) software was utilized to analyze the potential signaling pathway and molecular function in LUAD (19, 20). A customized Perl script was used to perform GSEA between high-AC02278.4 and low-AC02278.4 groups. According to the default statistical methods, an adjusted p-value <0.05 was considered significant.

A GSVA R package was used to examine the LUAD immune infiltration of 24 tumor-infiltrating immune cells in tumor samples through single-sample gene set enrichment analysis (ssGSEA) (21, 22). The correlation between AC02278.4 and infiltration levels of immune cells was analyzed by Spearman’s correlation, and these immune cells with the different expression groups of AC02278.4 were analyzed by rank-sum test.

BEAS-2B cell line was purchased from Cell Bank of Kunming Institute of Zoology and cultured in BEGM media (Lonza, Basel, Switzerland; CC-3170). Lung cancer cell lines, including A549, H1299, and SPC-A1, were purchased from Cobioer (Nanjing, China) with short tandem repeat (STR) documents, and were cultured in RPMI-1640 medium (Corning, Manassas, VA, USA) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin.

For shRNA knockdown experiments, independent shRNAs targeting a different region of AC02278.4 RNA were constructed using a pLKO.1 vector (Addgene, Cambridge, MA, USA), and the oligo sequences were provided as follows. Lentiviruses were generated according to the manufacturer’s protocol as previously documented (5), and indicated cells were infected by viruses twice with 48- and 72-h viral supernatants containing 4 μg/ml of polybrene, and stable cell lines were established by appropriate puromycin selection. The two independent AC02278.47 targeting sequences are as follows: shRNA#1, 5′-GGCACTTCGTGGCTGAACCGA-3′; shRNA#2, 5′-GGGGAACAATGGCTTCAGCAG-3′.

For BrdU incorporation assay, indicated cells were cultured in 8-well chamber slides for 24 h, pretreated with or without SC79 (Beyotime, Shanghai, China; SF2730) for another 24 h and then treated with 10 µM of BrdU (Abcam, Cambridge, UK; ab142567) for 20 min. Subsequently, indicated cells were fixed with 4% paraformaldehyde (PFA) at room temperature for 20 min and then incubated with BrdU primary antibody (Abcam, ab6326) followed by secondary antibody detection. The cell nuclei were stained with DAPI as previously documented (5).

Cell migration assay was performed as previously described (15). Briefly, indicated cells were seeded into 6-well plates (9 × 10⁵/cell) and incubated for 1 day, and then a straight line was scraped with pipette tips. Detached cells were removed. Photographs were taken at the indicated time, and the relative traveled distance was measured. For the trans-well migration assay, 2.5 × 10⁴ cells/well in 100 μl of serum-free medium were plated in a 24-well plate chamber insert, and the lower chamber was filled with 10% FBS. After incubation for 24 h, cells were fixed with 4% PFA, washed, and then stained with 0.5% crystal violet for further imaging.

In the real-time RT-PCR assay, cells were lysed by RNAiso Plus (Takara Bio, Beijing, China; Cat. 108-95-2). Total RNAs were extracted according to the manufacturer’s protocol and then reverse transcribed by using RT reagent Kit (Takara Bio, Beijing, China). The primer used in this study is as follows: β-actin-F: AAGTGTGACGTGGACATCCGC, β-actin-R: CCGGACTCGTCATACTCCTGCT, AC02278.4-F: TCAAGGCATCATGTGTCATT, AC02278.4-R: ACCTGCTGAAGCCATTGTTCC.

For the datasets from TCGA database, statistical analyses were performed using R. The Wilcoxon rank-sum test and chi-square test were used to estimate the association between AC02278.4 and clinical pathologic characteristics. The Kaplan–Meier method was used to calculated LUAD patient survival rates. Univariate and multivariate Cox analyses were performed to assess the correlation between clinical features and overall survival (OS), disease-free survival, and progression-free survival (PFS). For the data regarding the function of AC02278.4, Graph Pad Prism 7.0 was used for statistical analyses. Student’s t-test evaluated the statistical significance between groups. The significance of the data between the two experimental groups was determined by Student’s t-test, and multiple group comparisons were analyzed by one-way ANOVA. p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***), were significant.

We simultaneously analyzed the expression profiles of AC02278.4 (ENSG00000248538) based on TCGA database. The results confirmed that AC02278.4 was highly expressed in multiple cancer types compared to the normal samples ( Figure 1A ). To examine the expression of AC02278.4 in LUAD and normal samples, we analyzed the expression of AC02278.4 in 535 tumor tissues and 59 normal prostate tissues of TCGA data, and we uncovered that AC02278.4 was upregulated in LUAD tissues ( Figure 1B ). There were 59 pairs of LUAD cancer samples and matched adjacent normal samples in TCGA data. The expression level of AC02278.4 was also higher in LUAD samples than matched adjacent normal samples ( Figure 1C ). Moreover, we found that AC02278.4 increased in lung cancer tissues by analyzing the GEO dataset ( Figure 1D ). To validate the expression of AC02278.4 in LUAD, we first employed qRT-PCR assay to detect the expression of AC02278.4 in 20 pairs of LUAD tissues and matched adjacent normal tissues. It was identified that AC02278.4 was significantly upregulated in LUAD tissue than adjacent normal tissue ( Figure 1E ).

To examine the clinical relevance of AC02278.4 in LUAD, 535 LUAD patients with clinical parameters were classified into two subgroups based on the mean value of relative AC02278.4 expression. We then explored the correlations between AC02278.4 expression and clinical parameters, including pathologic stage, TNM stage, and residual tumor. Regarding the tumor pathological stage, a significant increase in AC02278.4 expression was observed in LUAD patients in stages 1, 2, 3, and 4 ( Figure 2A ). Based on the cancer stage, AC02278.4 expression was higher in patients with LUAD classified as T stage (T1, T2, and T3), N stage (N0, N1, and N2), and M stage (M0 and M1) ( Figures 2B – D ). We also found that AC02278.4 expression was significantly correlated with the residual tumor stage ( Figure 2E ). Furthermore, ROC analysis showed that the AC02278.4 could be used to differentiate LUAD patients from normal control with a specificity (AUC = 0.882) ( Figure 2F ). Additionally, Kaplan–Meier analysis showed that LUAD patients with higher AC02278.4 expression were associated with poor OS, disease-free survival, and PFS ( Figures 2G – I ). To further validate the correlation between AC02278.4 expression and OS, we examined the prognosis of AC02278.4 in lung cancer by clinical samples from Yunnan Cancer Hospital (N = 181). The results also showed that high AC02278.4 expression had a worse OS than low AC02278.4 expression ( Figure 2J ).

We conducted a univariate Cox regression analysis in the TCGA-LAUD cohort to determine whether AC02278.4 expression level and pathologic stage might be valuable prognostic biomarkers. In the univariate Cox regression analysis, high expression of AC02278.4, pathologic stage, and TNM stage were associated with OS in LUAD patients. To ascertain whether AC02278.4 expression level could be an independent prognostic factor for patients with LUAD, a multivariate Cox regression analysis was performed. We confirmed that increased AC02278.4 expression was a significant independent prognostic factor in the TCGA-LAUD cohort that directly correlated with adverse clinical outcomes, along with T stage ( Table 1 ).

The multivariate analysis result confirmed that AC02278.4 is an independent prognostic factor in LUAD. We then constructed a prediction model for OS, disease-free survival, and PFS by integrating AC02278.4 expression and pathologic stage. We established a nomogram to integrate AC02278.4 as a LUAD biomarker; higher total points on the nomogram for OS, progression-free interval (PFI), and disease-specific survival (DSS) indicated a worse prognosis ( Figures 3A–E ). A higher point on the nomogram represented a worse prognostic factor.

Based on the calibration curve of nomograms for OS, DSS, and PFS, the predictions conformed well to observations in all patients, and the test showed no deviation from the perfect fit. The nomogram had a C-index of 0.657 and contained 1,000 bootstrap replicates [95% CI: (0.634–0.679)] for OS. We also found DSS [C-index: 0.594, CI: (0.573–0.615)] and PFS (C-index: 0.645, CI: 0.617–0.673). It was found that the bias-corrected line in the calibration plot was close to the ideal curve, indicating a strong correlation between predicted values and observed values ( Figures 3D–F ). In summary, these results indicated that the nomogram can well predict short- or long-term survival of LUAD patients. We also used the GEO dataset to validate the above result, as is shown in Figures 3G, H ; using the GSE31552 dataset, we found that this nomogram could well predict the OS of LUAD patients.

GSEA was used to identify potential signaling pathways that are activated by high AC02278.4 expression. As shown in Figures 4A, B , there were eight significant signaling pathways related to the high AC02278.4 expression phenotype, including the focal adhesion, epithelial cell signaling in Helicobacter pylori infection, Wnt signaling pathway, VEGF signaling pathway, toll-like receptor signaling pathway, T-cell receptor signaling pathway, B-cell receptor signaling pathway, and cytokine–cytokine receptor interaction ( Figures 4A, B ).

Using the ssGSEA method, we explored the correlation between AC02278.4 expression and infiltrating immune cells in LUAD. These results confirmed that AC02278.4 expression was negatively associated with infiltration levels of Th17 cells, B cells, CD8 T cells, eosinophils, cytotoxic cells, Tem, Th1 cells, macrophages, T cells, mast cells, Tcm, DC, TFH, T helper cells, and iDC (p < 0.001) and was positively correlated with that of Th2 cells, Tgd, and NK CD56bright cells ( Figure 5A ). Furthermore, we found that patients with AC02278.4 high-expression group showed a reduction in the numbers of infiltrating aDC, B cells, CD8 T cells, cytotoxic cells, DC, eosinophils, iDC, macrophages, T helper cells, T cells, mast cells, Tcm, Tem, TFH, and Th1 cells ( Figure 5B ).

The above studies indicated that AC02278.4 expression was distinctly upregulated in LUAD tissues, and AC02278.4 might influence the progression in LUAD. To further investigate the biological role of AC02278.4 in LUAD, we first confirmed that the expression of AC02278.4 was significantly upregulated in H1299, A549, and SPC-A1 lung cancer cell lines compared to the human bronchial epithelial cells (BEAS2B) ( Figure 6A ). Furthermore, specific siRNA for AC02278.4 was used to construct A549 and SPC-A1 cells with stable knockdown of AC02278.4 expression. The knockdown efficiencies in transformed cell lines were detected by qRT-PCR analysis ( Figures 6B, C ). It was confirmed that knockdown of AC02278.4 reduced the proliferative capacity of A549 and SPC-A1 cells ( Figures 6D, E ) upon BrdU assays. Moreover, transwell assay and wound healing revealed that the migration and invasion abilities of A549 and SPC-A1 cells were significantly inhibited through downregulating AC02278.4 expression level ( Figures 6F – I ). Using various bioinformatics tools, we also constructed an mRNA–miRNA–lncRNA network; further study should be conducted to verify this result ( Figure 6J ). Briefly, it was suggested that AC02278.4 plays the role of an oncogene in lung cancer cells.