Introduction
Cancer is a leading cause of death, with 19.3 million new cancer cases and 10 million cancer deaths worldwide in 2020 [
1]. The burden of cancer morbidity and mortality is rapidly escalating [
2]. Despite the expanding number of treatments, including improvements in surgical modalities, immunotherapy, targeted therapies, and combination therapies [
3], mortality from liver and lung cancers continues to increase yearly [
4]. With the rapid development of next–generation sequencing, pan–cancer research has been widely used to identify tumor molecular markers and signaling pathways for a more comprehensive and in–depth understanding of the molecular mechanisms of tumor development [
5‐
8]. Therefore, searching for novel diagnostic and prognostic biomarkers for cancer can help improve the prognosis of cancer patients.
Centromere protein O (CENPO), a member of the centromere protein (CENP) family, plays an important role in meiosis and mitosis, especially in chromosome segregation and normal cell division processes [
9]. Many CENP family members have been identified, including CENPA, CENPL, CENPI and CENPU, and CENPO, which are associated with poor prognosis [
6,
9‐
12]. CENPO–like proteins are associated with the prevention of premature sister chromatid separation and are associated with the cell cycle [
13]. Recent studies have demonstrated that breast cancer patients with high CENPO expression have a poor prognosis and are independent factors affecting the distant recurrence–free survival (DRFS) of breast cancer patients [
14]. High CENPO expression promotes the proliferation of gastric cancer cells and is positively associated with poor prognosis in gastric cancer [
12]. CENPO is highly expressed in CRC (Colorectal Cancer) tissues and cells as positively associated with the malignancy of CRC degree and positively correlated with tumor progression possibly through EMT and PI3K/AKT signaling pathways [
15]. CENPO upregulation promoted bladder cancer progression [
16]. However, an investigation into the role of CENPO in pan–cancer is still lacking. We found that CENPO was significantly upregulated in multiple cancers and correlated with lung adenocarcinoma (LUAD) staging, overall survival (OS), and DFS. Thus, we chose LUAD to validate our bioinformatics results.
This study assessed CENPO expression in various human tumors using The Cancer Genome Atlas (TCGA) RNA–seq. The prognostic impact of aberrant CENPO expression on pan–cancer was investigated using the GEPIA database, and the impact of CENPO expression on the prognosis of LUAD patients was analyzed in conjunction with clinical data. Functional enrichment and pathway analysis were then performed, and correlations with immune cell infiltration, tumor mutational load (TMB), microsatellite instability (MSI), and gene methylation were investigated to understand further the biological mechanisms of CENPO in the pathogenesis of LUAD. Subsequently, bioinformatics results were validated by in vitro experiments. We preliminarily explored the potential mechanism of action of CENPO in LUAD, and CENPO could be used as an immunotherapeutic target and prognostic marker for pan–cancer, especially LUAD.
Methods
Material and methods
CENPO mRNA expression analysis
The UCSC Xena database GDC–TCGA data were used to analyze the differential mRNA expression of CENPO in pan–cancer and the relationship between CENPO, and cancer stage was analyzed by the “Stage plots” module of R. Lung adenocarcinoma RNA–seq sequencing profiles were obtained from the UCSC Xena database TCGA–LUAD data (517 LUAD cases and 58 normal cases), (
https://xenabrowser.net/datepages/). We first mined the TCGA–LUAD dataset for differentially expressed genes (DEGs) in LUAD samples vs normal samples by the R limma package and then presented DEGs as a volcano map using Volcano map [
17] (the criteria for determining differential gene expression were
P < 0.05 and logFC absolute value > 1). Cases with insufficient or missing data were removed from subsequent data processing.
Diagnostic value and prognostic analysis of CENPO
The diagnostic value of CENPO in pan–cancer was determined by plotting ROC curves from the GDC–TCGA data of the UCSC Xena database. AUC > 0.8 was thought as a high diagnostic value. In addition, the potential prognostic value of CENPO in pan–cancer was analyzed by overall survival (OS) and disease–free survival (DFS) from the GEPIA database (
http://gepia.cancer-pku.cn/). The GEPIA database was divided into high and low expression groups based on the median CENPO expression in 33 cancers. KM survival curve analysis was performed based on the overall survival and disease-free survival of patients in the above groups, respectively, and risk ratios were calculated based on the Cox PH model.
Methylation analysis and tumor mutation load
To screen methylation–regulated genes, we analyzed the correlation between CENPO expression and methylation in pan–cancer using GDC–TCGA data from the UCSC Xena database. Tumor mutational load (TMB) and microsatellite instability (MSI) analyses were performed on 33 tumors in the UCSC Xena database GDC–TCGA data via the “fmsb” package of R4.2.0.3.4. Next, the data were visualized using the radar chart package, and the obtained TMB correlation coefficients were all between 0.8 and -0.8, and the obtained MSI correlation coefficients were all between 0.4 and -0.4.
Correlation of CENPO expression with clinical factors of LUAD and prognostic analysis
The correlation between CENPO expression and clinical characteristics, including gender, age, TNM stage, T stage, N stage, distant metastasis, and history of radiation therapy, were analyzed in 517 patients with LUAD in the TCGA database. To further evaluate the prognostic value of CENPO in patients with LUAD, univariate and multifactorial COX proportional risk regression analyses were performed to investigate the effects of CENPO and various clinical characteristics on OS in patients with LUAD, and HR and 95% confidence intervals were evaluated. In addition, we constructed column line plots based on CENPO expression values and pathological staging to facilitate the application of CENPO in the clinical prognostic assessment of LUAD. Finally, we assessed the predictive accuracy of the column line plots by calibration curves.
Functional enrichment and immune infiltration analysis
Functional enrichment analysis of DEGs was performed by clusterProfiler in R software for Gene Ontology (GO) with the Kyoto Encyclopedia of Genes and Genomes (KEGG) [
18,
19]. GO is commonly used to annotate genes and analyze their biological processes, which helps to reveal the functional mechanisms behind the patterns observed in transcriptomics, genomics, and proteomic
s data, which are classified into cellular components (CC), molecular functions (MF), and biological processes (BP). Furthermore, KEGG is a reference knowledge base for the biological interpretation of large–scale molecular datasets. From this, key points of pathway mechanisms involved in differential genes in living systems can be obtained. Stromal cell score (StromalScore), immune cell score (ImmuneScore), and combined stromal and immune cell score (ESTIMATEScore) were performed by R4.2.0 for each tumor sample of TCGA–LUAD data in the UCSC Xena database, and the expression of CENPO was analyzed for its correlation with the tumor microenvironment correlation, immune cells, immune checkpoints.
Drug sensitivity analysis
The list of cgp2016ExprRma drugs was obtained by the R package, CENPO was divided into high and low expression groups by median, and CENPO expression in LUAD was analyzed with a drug sensitivity analysis (IC50). For filtered conditions, the P–value was less than 0.001.
Cell culture condition
The human normal lung epithelial cell line Beas–2b, LUAD cell lines XWLC–05, H838, A549, SPCA–1, and H1299 (Cell Bank of Chinese Academy of Sciences, Kunming, China) were placed in an incubator at 37 °C containing 5% CO2 using Roswell Park Memorial Institute (RPMI)– 1640 medium (GIBCO) and 10% fetal bovine serum (FBS, GIBCO).
RNA interference and quantitative real–time PCR
GEMA chemically synthesized CENPO–specific small interfering RNAs (CENPO siRNAs), negative control siRNAs, and three RNA interference target sequences were designed (Table
1). Subsequently, cells were inoculated in 6–well plates and transfected when cell fusion reached approximately 50%. CENPO expression plasmids were transfected into cells using Lipofectamine 2000 (Invitrogen) transfection reagent according to the manufacturer's instructions. After 6 h of transfection, the cells were harvested 48 h later, and qRT–PCR verified the transfection efficiency. According to the kit instructions, RNA was extracted from A549 and H1299 cells 48 h after transfection using a Trizol reagent (Invitrogen, Carlsbad, CA, USA). An iScript cDNA Synthesis Kit (Bio–Rad, USA) was used to reverse transcribe the RNA to cDNA, followed by iTaq Universal SYBR (Bio–Rad, USA). Finally, iTaq Universal SYBR (Bio–Rad, USA) was used for qRT–PCR. The relative mRNA expression of CENPO was quantified using the cycling threshold (Ct) value and normalized using the 2
−ΔΔCt method. The primer sequences are summarized in Table
1. GAPDH was used as a reference control.
Table 1
The sequences of primers for qRT-PCR
CENPO For | GGGAGGAAGTACCAGGCAGACC |
CENPO Rev | CAATCTAGCACAGAACGGGAAGGAC |
CENPO-Homo-1 For | GUGCUAUUGCUGUAUAATT |
CENPO-Homo-1 Rev | UUAUACAGCAAUCUUAGCACTT |
CENPO-Homo-2 For | CCGCAUACAUCACCAUUCATT |
CENPO-Homo-2 Rev | UGAAUGGUGAUGUAUCCGGTT |
CENPO-Homo-3 For | CCAGCGUGAAAGCAUGUAUTT |
CENPO-Homo-3 Rev | AUACAUGCUUUCACGCUGGTT |
GAPDH For | GGAGTCCACTGGCGTCTTCA |
GAPDH Rev | GTCATGAGTCCTTCCACGATACC |
Cell counting kit–8 assay and colony formation assay
Cell proliferation was assessed using the CCK–8 assay. A549 and H1299 cells were digested after 48 h transfection and inoculated in 96–well plates at a density of 2,000 cells/well. After 1, 2, 3, 4, and 5 days of incubation, 10 μl CCK8 solution was added sequentially to each well, and the absorbance (OD) was measured at a wavelength of 450 nm using an enzyme marker after 2 h. The 5–day data were calculated, and cell proliferation curves were plotted. The experiment was repeated three times under the same conditions. A549 and H1299 cells were inoculated in 6–well plates at 500 cells/well and incubated in an incubator at 37 °C in humid air containing 5% CO2 for 14 days until the colonies were visible to the naked eye, followed by washing the colonies twice with PBS and staining with 0.25% crystal violet for 1 h. The colonies were counted and statistically analyzed using ImageJ software.
Cell cycle test
Cells were collected 48 h after transfection and each tube contained 1 × 106 cells. Using the cell cycle assay kit from biosharp, each sample was treated with 500 µl of staining buffer, 25 μl of propidium iodide staining solution (20 ×), and 10 µl of RNase A (50 ×) staining for 30 min. Immediately thereafter, the samples were assayed by flow cytometry.
Wound–healing assay and Transwell invasion assay
The si–transfected (siNC, siCENPO) A549 and H1299 cells were inoculated in 6–well plates at a density of 1 × 106 cells/well. The cells were placed under a microscope for image acquisition at 0 h, 24 h, and 48 h. First, the distance (μm) between scratches at different time points was quantified using Image J software (National Institutes of Health) to calculate the healing area (initial scratch width– final scratch width)/initial scratch width × 100%). Next, A549 and H1299 cells were placed in a Transwell chamber (24–well, 8 mm well) (Corning) at a density of 3 × 104 cells/well and incubated for 24 h at 37 °C in a 5% CO2 incubator. The inner chamber contains 100 μl of cell suspension, and the outer chamber was 600 μl of 1640 medium containing 20% FBS. Afterward, the non–invasive cells in the upper chamber were removed Next, the cells attached to the polycarbonate membrane were fixed with 4% pre–cooled paraformaldehyde for 30 min and stained with 0.1% crystal violet for 20 min at room temperature. Afterward, the cells were placed under a 100 × microscope, and images were captured from five randomly selected fields of view.
Statistical analysis
Data were analyzed and plotted using SPSS 25.0 (IBM, Armonk, NY, USA) and GraphPad Prism 8.0 software (GraphPad Software Inc., San Diego, CA, USA), and analyzed using R language. The results represent at least three replicate experiments and are expressed as Mean \(\pm\) SD. Differences between two groups were compared using an unpaired t–test, and data from more than three groups were analyzed using one–way ANOVA. The Wilcoxon or Kruskal–Wallis test was employed to contrast the data from different groups. P < 0.05 was considered statistically significant (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).
Discussion
Mitotic granules are an important component of chromosome segregation during meiosis and mitosis in normal cells. Their abnormal function can lead to mitotic arrest and altered chromosome numbers [
22]. It has been reported that Centromere proteins control chromosome segregation and that abnormalities in Centromere proteins may lead to chromosomal instability in tumors, which in turn lead to tumor progression and drug resistance [
23]. CENPO is associated with the development and progress of various tumors, including LUAD. These data suggest the potential of CENPO as a prognostic biomarker.
In this study, we verified that CENPO was significantly highly expressed in most cancer tissues and correlated with the TNM stage by the TCGA database. ROC curve analysis indicated that CENPO might serve as a pan–cancer diagnostic biomarker. Meanwhile, KM survival analysis showed that increased expression of CENPO was associated with poorer OS of UVM, KIRC, SKCM, SARC, LIHC, LUAD, MESO, ACC, KICH, LGG, and PCPG, whereas the increased expression of CENPO was associated with KICH, LUAD, LUSC, SARC, SKCM, UVM, ACC, BLCA LGG, LIHC, and PAAD poorer DFS. These findings suggest that CENPO may play a significant role in tumorigenesis and prognosis, and that abnormal CENPO may be a potential prognostic predictive molecule for cancer. Further analysis revealed that CENPO expression in LUAD was significantly correlated with gender, TNM stage, N stage, T stage, and whether radiotherapy was administered. Columnar plots were constructed to further evaluate the role of CENPO in LUAD prognosis.
Epigenetic mechanisms regulate the expression of immunosynaptic proteins by tumor cells, thereby suppressing immune function [
24]. DNA methylation is a key epigenetic mechanism of immune regulation, and cancer cells often use epigenetic dysregulation to silence tumor suppressors or activate oncogenes [
25]. Our results showed that CENPO expression was negatively correlated with methylation in LUAD, suggesting that CENPO may be involved in epigenetic modifications to regulate mitophagy assembly and mitophagy function [
26]. The tumor microenvironment is a complex survival environment for tumor cells, mainly characterized by immune cells, mesenchymal cells, blood vessels, and an extracellular matrix [
27]. In recent years, there has been increasing evidence that the tumor microenvironment and the level of immune cell infiltration are closely related to the growth and progression of tumor cells [
28‐
31]. Although previous studies have demonstrated the prognostic importance of CENPO in various types of cancer, there are few reports on whether aberrant CENPO expression influences immune cell infiltration. Immune cells of the tumor microenvironment are associated with the efficacy of immunotherapy, including CD4 + T cells [
32] CD8 + T cells [
33], NK cells [
34], and macrophage cells [
35]. Our results show for the first time that CENPO expression in LUAD is significant positive associated with the infiltration of memory–activated CD4 T cells, M0 macrophages, M1 macrophages, T cells CD8, Neutrophils, naive B cells, and NK cell resting; our results also show a significant negative correlation with the resting Mast cells, resting Dendritic cells, memory B cells, Monocytes, and memory resting CD4 T cells. Macrophages have been reported to promote tumor progression through tissue remodeling and angiogenesis [
36]. High infiltration of M1 macrophages significantly reduced survival time in prostate cancer patients [
29]. The M1 macrophage content was significantly higher in the metastatic group than in the non–metastatic group of medulloblastoma [
37]. Our results also showed that increased CENPO expression was associated with poor LUAD survival.
The level of immune cell infiltration in TME is critical for tumor prognosis [
38]. An immune–based assay called “Immunoscore” was defined to quantify T–cell infiltration in situ and proved superior to the TNM classification of colorectal cancer patients [
39]. Furthermore, ImmuneScore has been reported to predict that survival in CRC patients is superior to microsatellite instability [
40]. Our results showed that StromalScore, ImmuneScore, and ESTIMATEScore were significantly lower in the CENPO high expression group, which may indicate that increased CENPO expression is associated with disturbed immune status in LUAD patients. Immune checkpoints are involved in physiological immune responses, and immune checkpoints and their ligands are upregulated in the tumor microenvironment of various malignancies to induce antitumor immune responses [
41]. The clinical application of immune checkpoint inhibitors has improved the prognosis of many patients with advanced tumors [
42]; we showed that CENPO was positively correlated with immune checkpoint CD276, TNFSF4, TNFSF9, CD247, and LAG3. Furthermore, it was negatively correlated with immune checkpoints LGALS9, CD44, TNFRSF14, CD40LG, and HHLA2. These results suggest that CENPO plays an important immunomodulatory role and may serve as a potential immunotherapeutic relevant biomarker. Previous studies reported that the CENP family is associated with immune function [
6,
43], and our bioinformatics results also demonstrated that CENPO is associated with immune regulation. We then demonstrated by in vitro experiments that CENPO was highly expressed in A549 and H1299 cell lines and that the knockdown of CENPO significantly inhibited the proliferative activity, migration, and invasive ability of A549 and H1299 cells, suggesting that CENPO may affect LUAD metastasis. Studies have reported that the knockdown of CENPO inhibited GC cell growth, induced apoptosis, and decreased the expression of ATM, cyclin D1, and c–Jun [
12]. Knockdown of CENPO in CRC downregulated the expression of N–cadherin, Vimentin, Snail, and CCND1, inhibited cell proliferation, and attenuated migration and invasion [
15]. This is consistent with our Bioinformatics Analysis, the functional enrichment results showed significant enrichment of cell cycle, P53 pathway, and extracellular matrix. Thus, our study suggests that the knockdown of CENPO significantly decrease the expression of CDK2 and CyclinB1, and induce G0/G1 arrest in A549 and H1299 cells. We speculate that CENPO may promote LUAD progression by regulating the cell cycle. These results suggest that CENPO may be a potential molecular target for anti–tumor therapy in LUAD, and the specific mechanism of CENPO in regulating the development of LUAD will be further investigated in the future.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.