2.1 Data collection and analysis
The tumor immune estimation resource, version 2 (TIMER2.0) (http://timer.cistrome.org/) was used to analyze the differential expression of IGSF6 between the tumor and adjacent normal tissues across all TCGA tumors. To measure IGSF6 expression in LUAD, data from tumor tissues and normal tissues were obtained from the TCGA database and determined by gene expression profiling interactive analysis 2 (GEPIA2) (http://gepia2.cancer-pku.cn/#index).
2.2 IHC staining
IHC images of IGSF6 protein levels in normal tissues and LUAD tissues were downloaded from the HPA (http://www.proteinatlas.org/).
2.3 Patients and samples
Paired tumor tissues and adjacent normal tissues were collected from LUAD patients (n = 9) who underwent primary surgical resection. The present study was approved by the respective Ethics Committee of Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University. Written informed consent was obtained from all the subjects following the Declaration of Helsinki.
2.4 RNA isolation and qRT-PCR
Total RNA was extracted from cells with TRIzol reagent (Invitrogen, California) following the manufacturer’s instructions. Random primers and a ReverTra Ace® qPCR RT Kit (Toyobo, Osaka, Japan) were used to synthesize cDNA. Bio-Rad SYBR Green Supermix (Bio-Rad, Hercules) was used to perform quantitative real-time PCR in triplicate. The primer sequences were as follows: human β-ACTIN, sense 5-GAGTGTGGAGACCATCAAGGA-3, antisense 5-TGTATTGCTTTGCGTTGGAC-3; human IGSF6, sense 5-GCAATCTCGGCTCACTACAACCTC-3, antisense 5-CGTGGTGGTGCGTACCTGTAATC-3. The 2−ΔCT calculation method was used to determine the relative target gene levels.
2.5 Western blot
Protein extracted from tissues was prepared as described previously [21]. Protein (300µg) was separated by 12% SDS-PAGE and then transferred onto Immobilon polyvinylidene fluoride membranes (BioRad, Hercules, CA). The membranes were blocked with 30 mL 5% BSA before probing overnight at 4℃ with rabbit-anti-human IGSF6 polyclonal antibody (1:1000) (ThermoFisher Scientific, Dallas, TX) or rabbit-anti-human β-ACTIN antibody (1:1000) (CST, Danvers, MA) and then with a secondary HRP-conjugated goat anti-rabbit IgG (diluted at 1:8000) (CST, Danvers, MA), followed by chemiluminescent detection (Champion Chemical, Whittier, CA). Full-length blots are included in the Supplementary Information file.
2.6 UALCAN analysis
The UALCAN (http://ualcan.path.uab.edu/) website provides an extensive and interactive study of bioinformatics applying RNA-seq and clinical data of 33 malignancies from TCGA. The database can compare gene expression and promoter methylation levels in tumors to those in healthy samples, and in different tumor stages or subtypes, as well as other clinicopathological features. In the present study, IGSF6 expression levels and promoter methylation levels from major clinical features such as tissue types (normal/primary tumor), LUAD stages (stages 1, 2, 3, and 4), and lymph node stages (N0, 1, 2, and 3) were analyzed by using UALCAN.
2.7 Survival Prognosis Analysis
Kaplan-Meier survival curves were used to assess the OS, DSS, and PFI of IGSF6 in LUAD. Survival curves comparing the IGSF6-mutated and IGSF6-unmutated groups in LUAD were obtained from cBioPortal (https://www.cbioportal.org/). Cutoff-high (50%) and cutoff-low (50%) values were used as the expression thresholds for splitting the high-expression and low-expression cohorts. The hazard ratio (HR) with 95% confidence intervals was also analyzed, as well as the log-rank p-value. Statistical significance was defined as p < 0.05.
2.8 IGSF6-related gene enrichment analysis
Pearson correlation analysis of IGSF6 mRNA and other mRNAs in LUAD was performed using TCGA LUAD data. The 300 genes that were the most relevant to IGSF6 were selected for enrichment analysis to determine the potential function of IGSF6. GO and KEGG analyses were conducted by running the R software clusterProfiler package. Different gene expression profiles between the high- and low-risk subgroups were identified (|log2FC|>1, FDR < 0.05). The Benjamini-Hochberg (BH) method was used to adjust the p-value. Detailed parameters were shown as the following: ont = all, qvalue-Cutoff = 0.05, and pvalue-Cutoff = 0.05. GSEA was performed using the gseKEGG and gsePathway functions in clusterProfiler with the following parameters: maxGSSize = 1000, minGSSize = 10, nPerm = 1000, and pvalue-Cutoff = 0.05.
2.9 Immune cell infiltration
TISIDB (http://cis.hku.hk/TISIDB/index.php) is an online platform that combines various heterogeneous data sources to study tumor and immune system interactions. This database may help researchers understand how tumors and immune cells interact, as well as forecast immunotherapy responses and identify new immunotherapy targets. In the present study, TISIDB was utilized to investigate the association of IGSF6 with tumor-infiltrating lymphocytes (TILs) in LUAD. In addition, Timer2.0 was used to analyze the correlation between IGSF6 and suppressive immune infiltrates including M2 macrophages, Tregs, and MDSCs.
2.10 Prediction of IGSF6 localization in cells
Analysis of cell type which is enriched with IGSF6 protein in the lung, as well as prediction of IGSF6 localization in cells, were obtained from the HPA database.
2.11 FCM
Collagenase II (Sigma-Aldrich, St. Louis, MO) was used to digest tumor tissues and adjacent tissues derived from LUAD patients to obtain a single-cell suspension. Mononuclear cells were isolated by density gradient centrifugation over a Ficoll cushion. Cells were collected and stained with PE-anti-human-CD68 (eBioscience, San Diego, CA), APC-anti-human-CD80 (eBioscience, San Diego, CA), and FITC-anti-human-IGSF6 (Epigentek, Farmingdale, NY) mAbs. Stained cells were collected and then analyzed via FCM (FACSAria, BD Biosciences).
2.12 Statistical Analysis
Statistical analysis of defining group differences is determined by a two-tailed t-test comparing two groups, one-way or two ways ANOVA multiple tests comparing three or more groups. ROC curve was used to analyze the sensitivity and specificity of IGSF6 in the diagnosis of LUAD by estimating AUC. The Wilcoxon test was used to compare the IGSF6 expression in wild-type (WT) and IGSF6-mutated groups of LUAD. In survival curves, HR and p-values were described using the log-rank test. Spearman’s correlation coefficient was calculated to analyze the association of IGSF6 expression with immune infiltrates. The Pearson correlation coefficient was calculated to analyze the connection between IGSF6 expression and other molecules. The strength of correlation was judged to be a very weak correlation if r < 0.2, weak if r < 0.4, moderate if r < 0.6, strong if r < 0.8, and very strong if r < 1.0. Statistical significance was defined as p < 0.05.