GINS2 Functions as a Key Gene in Lung Adenocarcinoma by WGCNA Co-Expression Network Analysis

Introduction

Lung cancer is one of the most prevalent cancers in the world, with 226,160 new cases and 160,340 mortalities in 2012.¹ Lung adenocarcinoma is a major subtype of lung cancer.² Due to the complex pathomechanism, the survival rate of lung adenocarcinoma is still low.³ Therefore, it is important to explore the underlying molecular mechanisms for lung adenocarcinoma development and progression.

GINS complex subunit 2 (GINS2), a member of GINS family, plays a crucial role in DNA duplication.⁴ In acute promyelocytic leukemia, up-regulation of GINS2 facilitated HL60 cell proliferation.⁵ Down-regulation of GINS2 inhibited breast cancer cells growth by activation of endogenous DNA damage.^6,7 Elsewhere, it was found that deregulated GINS2 expression promoted distant metastasis free survival and endocrine therapy resistance in breast cancer patients.⁸ A previous study has reported that GINS2 mRNA was found to be significantly up-regulated in stage II lung adenocarcinoma.⁹

In this research, we identified a novel oncogene, GINS2, in lung adenocarcinoma based on GSE32863 dataset. Additionally, we found that up-regulation of GINS2 had better diagnostic values in lung adenocarcinoma. Notably, over-expression of GINS2 promoted lung adenocarcinoma cells proliferation and migration and inhibited apoptosis in vitro.

Materials and Methods

Data Collection and DEGs Identification

GSE32863 contained 58 lung adenocarcinoma and 58 adjacent non-tumor lung fresh frozen tissues. The gene expression profile of GSE32863 was downloaded. Raw data was normalized by GEO2R calculation. Adjusted p<0.05 and |log₂FC|>1 were set as the criterion for DEGs. The RNA sequencing and clinical data of lung adenocarcinoma were downloaded from TCGA database (http://cancergenome.nih.gov/). A total of 106 clinical samples from the case-control study were collected to analyze the relationship between GINS2 expression and clinical information. All patients were informed of the study and provided written informed consent. The experiment was approved by the Institutional Review Board of China Medical University (No.67 in 2010) and conducted using methodologies conforming to the standards set by the Declaration of Helsinki.

Construction of WGCNA Co-Expression Network

The WGCNA was performed to construct a co-expression network of DEGs via R (3.6.2) WGCNA package. The soft threshold was determined when scale-free R²>0.85. Then the cluster dendrogram and network heatmap of genes were constructed. In addition, eigengene values of modules were calculated, and a clustering tree was carried out.

KEGG and PPI Network Analyses

In order to explore the biological functions of cancer-associated genes, KEGG enrichment was analyzed via R ggplot2 package. PPI network helps us to explore novel molecular mechanism. STRING (https://string-db.org/) online database is a tool to search the interaction among different proteins. The significantly associated genes with weight value >0.8 were collected to construction PPI network.

Real Time-Polymerase Chain Reaction

Total RNA from cells was isolated by the reagent (Takara 9108, Takara, Japan). Then RNA was retro-transcribed according to the manufacturer`s instructions (Takara RR047A, Takara, Japan). RT-PCR was performed by the SYBR Green PCR kit (Takara, Japan). PCR amplification was performed in triplicate according to Takara RR820A (Takara, Japan) under the conditions of 95°C for 30 and 40 second cycles at 95°C for 5 seconds and 60°C for 34 seconds. The Ct values were calculated via ABI 7500 Fast Real-time PCR System (ABI, USA). Fold changes in GINS2 were measured by the 2^−ΔΔCT method. The sequences of primer probes were as follows:

GINS2: Forward: AAACTCCGCACGAACCT

Reverse: CCACGAGTACCTCATCACG

Western Blot

First, 20 μL proteins were separated by 10% SDS-PAGE (Beyotime, People's Republic of China) and then transferred onto 0.45 μm PVDF (Millipore, USA) membranes. After blocking with skim milk (BD 232100, BD, USA), the membranes were incubated overnight at 4°C with anti-GINS2 (ABcam 37683, Abcam, People's Republic of China) and anti-GAPDH (ABcam 8245, Abcam, People's Republic of China) antibodies. Then membranes were incubated with a horseradish peroxidase-conjugated secondary antibody (ABcam 150117, Abcam, People's Republic of China) for 1 hour. The signal was detected with enhanced chemiluminescence (Thermo Scientific™ 34076, USA).

Cell Culture and Transfection

HBE and A549 were purchased from GeneChem (Shanghai, People's Republic of China). HBE and A549 cells were all cultured in RPMI-1640 (GIBCO, USA). All cell lines were maintained at 37°C in a humidified chamber with 5% CO₂. GINS2 overexpression plasmid (OE-GINS2) and negative control plasmid (NC) were all purchased from Syngeneiech (Beijing, People's Republic of China). 1x10⁵ to 1.5x10⁵ cells were seeded into 6-well plates and transfected plasmids using jetPRIME transfection reagent (Polyplus transfection, USA) according to the manufacturer`s instructions. After being transfected for 48 hours, all cells were collected and used for the next experiments. The sequences of all plasmids were as follows:

OE-GINS2: 5′-GGATCCCGAAGGCAGACGA-3′

NC: 5′-TTCTCCGAACGTGTCACGT-3′

Cell Growth

A total of 1x10⁴ cells were collected and seeded into 96-well plates. The cell growth was measured with MTS solution (Promega G3580, People's Republic of China) according to the instructions. Cells were incubated at 37°C for 2 hours, adding 20 uL MTS solution. The OD was assessed at 0, 12, 24 and 48 hours at 490 nm. Each group was tested in five parallel samples.

Cell Apoptosis

Transfected cells were seeded in a 6-well plate. And then cells were collected by trypsin digestion without EDTA (Thermo Fisher, USA). The apoptotic rates of cells were estimated by Annexin V-APC/7-AAD (KGA1016, KeyGen BioTECH, People's Republic of China) according to the protocols. Cells were stained with 5 μL 7-Amino-Actinomycin D and 5 μL Annexin V-Allophycocyanin at 37°C for 10 minutes. The cells were detected by FCM (Merck Millipore, Germany).

Cell Migration

Transfected cells were starved for 48 hours. Then cells (1x10⁴) were resuspended in 200 μL serum‐free medium, and seeded in the top chamber with 8.0 μm pores in the membrane (Corning, NY, USA). Culture medium with 10% FBS was added to the bottom chamber. After 48 hours at 37°C, the cells in the top chamber were immobilized with paraformaldehyde for 15 minutes and then stained with crystal violet (0.1%) for 10 minutes. After washing with double-distilled water three times, the chamber was stained with 0.1% the crystal violet. After drying, the number of cells migrating was counted by an inverted microscope (ECHO, CA, USA).

Statistical Analysis

The overall survival (OS) were analyzed via R survival package. Student`s test was performed to validate GINS2 expression differences between different groups via R ggplot2 package. Two-way analysis of variance was conducted to analyze the difference in cell growth between the two groups. A two-sided p<0.05 was considered statistically significant.

Results

Identification of DEGs

In order to verify DEGs, we normalized the raw data of GSE32863 via GEO2R calculation (Figure 1A). The heatmap of all genes is shown in Figure 1B. Adjusted p<0.05 and |log₂FC|>1 were set as the criterion of DEGs (Figure 1C), including 628 up-regulated genes and 339 down-regulated genes (Table 1).

Table 1 The Differently Expressed Genes in GSE32863

Figure 1 Identification of DEGs in GSE32863 dataset. (A) Normalization of GSE32863 data. (B) Hatmap of all genes expression in GSE32863. (C) The volcano map of DEGs.

WGCNA Co-Expression Network

Based on GSE32863 up-regulated genes expression, we constructed a WGCNA co-expression network. As shown in Figure 2A, the soft threshold of β=18 was identified (R²=0.9, slope=−1.38). Then, all up-regulated genes were split into eight modules (Figure 2B). Each module had different color and genes (Table 2). Figure 2C plotted the network heatmap of interaction of eight modules. All the modules were significantly independent of each other, and a similar result was also observed in gene expressions in each module. In addition, eigengene module values were calculated in each module and a clustering tree is presented in Figure 2D.

Table 2 Each Number of Each Module in WGCNA

Figure 2 WGCNA co-expression network analysis of GSE32863. (A) Scale independence, mean connectivity and scale-free topology. (B) Cluster dendrogram of all up-regulated genes. (C) interaction relationship of co-expression genes. (D) Clustering tree based on the module eigengenes of modules. (E) KEGG analysis of genes in grey module. (F) PPI analysis of top nine up-regulated genes.

Among all the modules, the grey module had the highest number of genes, containing 559 up-regulated genes. KEGG results indicated that those highly expressed genes were mainly enriched in regulation of growth, cell migration and cell-substrate adhesion associated pathways (Figure 2E). Then we constructed a PPI network to further identify the hub gene. The PPI network of the top nine genes is presented in Figure 2F. Table 3 shows that GINS2 in the grey modules had the highest score (score=0.999). Herein, we explored the biological function of GINS2 in lung adenocarcinoma.

Table 3 Nine Hub Genes in the PPI Network Constructed by STRING

Expression and Clinical Value of GINS2 in Lung Adenocarcinoma

Figure 3A shows that the expressions of GINS2 in patients from the TCGA-LUAD cohort, GSE3286 and clinical samples dataset were significantly higher than that in healthy controls (TCGA-LUAD cohort: p<0.001; GSE32863: p<0.001; clinical samples dataset: p<0.001). Overexpression of GINS2 in the TCGA-LUAD cohort was strongly associated with poor prognosis (Figure 3B: p=0.0075). Next, we performed the ROC curve to establish the prognostic ability of GINS2 in the TCGA-LUAD cohort (Figure 3C), GSE32863 (Figure 3D) and clinical samples dataset (Figure 3E). Table 4 shows that the AUCs of GINS2 in the TCGA-LUAD cohort, GSE32863 and clinical samples dataset were 0.847 (p=0.0023), 0.894 (p<0.001) and 0.851 (p<0.001), respectively. Combined with high sensitivity and specificity, GINS2 could be considered as a candidate prognostic biomarker in lung adenocarcinoma. In addition, GSEA results revealed that high GINS2 expression in the TCGA-LUAD cohort was significantly enriched in the gene sets of “HALLMARK_G2M_CHECKPOINT”, “HALLMARK_E2F_TARGETS”, “HALLMARK_DNA_REPAIR” and “HALLMARK_MYC_TARGETS_V2” (Figure 3F). The results further suggest that GINS2 plays a vital role in the prognosis of lung adenocarcinoma.

Table 4 ROC Analyses Results of GSE32863, TCGA and Clinical Samples

Figure 3 Integrative analyses of GINS2. (A) GINS2 expression in TCGA-LUAD cohort, GSE32863 and clinical samples dataset. (B) Prognostic analysis of GINS2 in TCGA-LUAD cohort. (C–E) ROC analysis of GINS2 in TCGA-LUAD cohort (C), GSE32863 (D) and clinical samples dataset (E). (F) GSEA analysis results of GINS2.

We also explored the relationship between GINS2 expression and clinical information. As Table 5 illustrates, in the clinical samples dataset, GINS2 expression is significantly related to the clinical stage (p<0.001), but not with age, gender or smoking status.

Table 5 The Relationship Between GINS2 Expression and Clinical Features in Our Case-Control Study

Effects of GINS2 on Growth, Migration and Apoptosis of Tumor Cells

The results of RT-PCR (Figure 4A) and Western blot (Figure 4B) showed that the expression of GINS2 in A549 cells was significantly up-regulated. Then we transfected OE-GINS2 and NC plasmids to A549 cells to explore biological functions of GINS2. Expression levels of GINS2 in cells transfected with blank, NC and OE-GINS2 were estimated by PCR (Figure 4C) and Western blot (Figure 4D). GINS2 expression was significantly over-expressed in cells transfected with OE-GINS2 (p<0.001). The result of the MTS assay showed that the growth rate was significantly increased in cells transfected with OE-GINS2 (Figure 4E: p<0.001), suggesting that up-regulation of GINS2 had a significant promotion effect on lung adenocarcinoma cells proliferation. In addition, Figure 4F showed that the migration viability of cells transfected with OE-GINS2 was significantly increased compared with cells transfected with NC and blank (p < 0.001). Figure 4G showed that the apoptosis rate of tumor cells with GINS2 overexpression was significantly decreased. No significant differences between NC and blank groups were observed.

Figure 4 Biological mechanism of GINS2 in lung adenocarcinoma. (A) GINS2 mRNA expression in HBE and A549 cells. (B) GINS2 protein expression in HBE and A549 cells. (C) GINS2 mRNA expression in cells transfected with blank, NC and OE-GINS2. (D) GINS2 protein expression in cells transfected with blank, NC and OE-GINS2. (E) up-regulation of GINS2 promoted tumor cells growth. (F) overexpression of GINS2 promoted tumor cells migration. (G) up-regulation of GINS2 suppressed tumor cells apoptosis. **p<0.01, *** p<0.001.

Discussion

The pathomechanism of lung adenocarcinoma is complicated and multifaceted. Numbers of oncogenes or tumor suppressor genes had been proved in the occurrence and progression of lung adenocarcinoma. In our study, we systematically analyzed the oncogene role of GINS2 in lung adenocarcinoma.

Initially, we analyzed the raw data in GSE32863 and 967 DEGs were identified. Next, we constructed a WGCNA co-expression network to classify DEGs into different modules. WGCNA is a systematic approach to dividing genes with similar expression patterns into the same modules and further exploring correlation between modules.^10,11 In our study, after identifying the key module, pathway prediction results illustrated that those genes were mainly enriched in regulation of growth, regulation of cell proliferation and substrate adhesion pathways, suggesting that those DEGs may participate in the occurrence and development of lung adenocarcinoma. In addition, we constructed a PPI network to further identify the hub gene in the grey module. GINS2, with the highest score, was selected to explore the molecular mechanism in lung adenocarcinoma.

Based on the heterotetramer complex structure, GINS2 plays a vital role in chromatin binding and replication activity.¹² In other human tumors, for example, in thyroid cancer, over-expression of GINS2 activated cancer cells proliferation and inhibited apoptosis by mediating CITED2 and LOXL2;^13,14 In early-stage cervical cancer, GINS2 was significantly over expressed in tumor tissues and cancer cells and had a negative correlation with OS of cervical cancer patients.¹⁵ In our study, we downloaded GINS2 expression profile and clinical characteristics in TCGA-LUAD cohort. Over-expression of GINS2 was observed in the TCGA-LUAD cohort, GSE32863 and clinical samples dataset. Additionally, aberrant expression of GINS2 was significantly associated with poor prognosis in lung adenocarcinoma patients. ROC analysis results showed the higher diagnostic value of GINS2 in lung adenocarcinoma. Besides, we concluded that aberrant GINS2 expression had a significant correlation with lung adenocarcinoma clinical stage. Consistently, a study had illustrated that higher expression of GINS2 had significantly association with TNM stages and clinical stages.¹⁶ Above all, these findings suggested that overexpression of GINS2 could be a novel diagnostic biomarker in lung adenocarcinoma.

The previous study had indicated that GINS2 acted on PI3K/Akt and MEK/ERK pathways to regulate NSCLC cells proliferation, migration and invasion.¹⁶ In our study, GSEA results showed that GINS2 may be enriched in G2M_checkpoint, E2F_targets, DNA-repair and MYC_targets_V2 pathways. Researchers have reported that G2M_checkpoint plays a crucial role in various human tumors.^17,18 For example, G2M_checkpoint participated in regulating cells growth and apoptosis of ovarian tumor,¹⁹ colon cancer,²⁰ hepatocellular carcinoma,²¹ and breast cancer.²² Similarly, E2F_targets,^23–25 DNA-repair^26–28 and MYC_targets_V2^29–31 pathways were found to participate in tumor therapy, drug resistance, immune evasion and progression. These results further supported that GINS2 participated in the initiation and progression of lung adenocarcinoma.

We found GINS2 to be significantly over expressed in lung adenocarcinoma cells in vitro. MTS results demonstrated that GINS2 accelerated proliferation of tumor cells which is consistent with a previous report.¹⁶ In addition, the number of migration cells with higher expression of GINS2 was significantly increased. Consistent with the GSEA results, we indicated that GINS2 promoted the growth and migration abilities of tumor cells in lung adenocarcinoma. The apoptotic rate of cells transfected with GINS2 overexpression plasmids was significantly lower than blank and NC, proving that GINS2 inactivated tumor cells apoptosis. Consistent with the previous study,¹⁶ these results further imply that GINS2 could accelerate tumor cell proliferation and migration and inhibit tumor cells apoptosis in vitro.

In conclusion, we utilized GEO raw data to construct WGCNA co-expression network. A hub gene was identified based on the score of PPI network. High expression of GINS2 had better diagnostic values in lung adenocarcinoma. GSEA results predicted pathways which GINS2 may be involved in the progress and development of lung adenocarcinoma. Moreover, we found that GINS2 promoted the proliferation and migration of tumor cells and suppressed the apoptotic rates of tumor cells, further indicating that GINS2 could function as a novel diagnostic and therapeutic biomarker in lung adenocarcinoma.