EXOSC10 is a novel hepatocellular carcinoma prognostic biomarker: a comprehensive bioinformatics analysis and experiment verification

Sample information

The mRNA expression data and related clinical data of HCC were downloaded from The Cancer Genome Atlas (TCGA) database (https://www.cancer.gov/ccg/research/genome-sequencing/tcga/) (Cancer Genome Atlas Research Network et al., 2013), which included patient ID, age, survival time, survival status, tumor grade, stage, and stage of TNM. And the missing and uncertain clinical data were deleted. If there were multiple samples from the same patient, the average value of mRNA expression of the same gene was obtained in the samples, and the deadline was March, 2021. Data extraction and merging were performed using Perl (version 5.3.0). Finally, we obtained 374 HCC patient samples and 50 normal control samples from the TCGA database for a follow-up study.

The TCGA database is open to the public, and all data have been agreed to be used for analysis and have obtained moral recognition. This study is based on open source data, strictly abides by the release guidelines and access policies of the database, and is not bound by other ethics.

EXOSC10 mRNA expression analysis

The analysis of EXOSC10 mRNA expression was based on the TCGA database using the package R “Limma”, and the “ggplot2” package was used to draw a box plot for visualization.

The Shiny Methylation Analysis Resource Tool (SMART) database (http://www.bioinfo-zs.com/smartapp/) was used to investigate the distribution of methylation sites in EXOSC10 and pan-cancer analysis of EXOSC10 methylation (Li, Ge & Lu, 2019). The expression of EXOSC10 in pan-cancer was identified in the TIMER database (http://timer.cistrome.org) (Li et al., 2017).

Survival and clinicopathological analysis in HCC patients

The correlation between EXOSC10 expression and patient prognosis, such as overall survival (OS), and progression-free survival (PFS), was analyzed in HCC and shown by forest plots, nomogram, and Kaplan–Meier by R package “survival” and “RMS”. The multivariate or univariate Cox regression models were performed using R package “survival”. Logistic regression analysis was used to explore the correlation of EXOSC10 expression with clinical factors of HCC using the R package “Limma”.

GO annotation enrichment and KEGG pathway enrichment analysis

The samples in TCGA were divided into high and low EXOSC10 expression groups according to the median level of EXOSC10 mRNA. Differential expression genes (DEGs) between high and low EXOSC10 expression groups were identified using R package “Limma”, with absolute logarithmic fold change (FC) >1 and FDR < 0.05 were considered the cutoff criterion. Gene ontology (GO) and KEGG pathway analysis were performed using DEGs in high and low expression groups of EXOSC10 using the R package “clusterProfiler”. P < 0.05 was considered statistically significant for GO annotation enrichment analysis and KEGG pathway enrichment analysis.

GSEA (gene set enrichment analysis)

Potential EXOSC10-regulated pathways were analyzed by GSEA. We used GSEA software to investigate the tumor hallmarks between high and low EXOSC10 gene expression. P < 0.05 and a false discovery rate (FDR) < 0.25 were considered statistically significant.

Correlation analysis of EXOSC10 with immune cell infiltration

TIMER (https://cistrome.shinyapps.io/timer/) was used to analyze the correlation of EXOSC10 expression with six kinds of tumor-infiltrating immune cells (B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells). Furthermore, we analyzed the association of the immune response of immune cells with EXOSC10 expression in HCC by CIBERSORT (Newman et al., 2015). P < 0.05 was considered statistically significant.

Furthermore, we applied the “Estimation of Stromal and Immune cells in Malignant Tumors using Expression data” (ESTIMATE) (Yoshihara et al., 2013) algorithm to assess the relationship between EXOSC10 expression and immune/stromal/estimate scores in each HCC samples.

Correlation analysis of EXOSC10 with immune checkpoints and TMB analysis

The correlation of EXOSC10 expression with immune checkpoints in HCC was evaluated using R package “Limma” and “corrplot” based on TCGA. Results with P < 0.001 were considered statistically significant. In addition, the correlation of EXOSC10 expression with tumor mutation burden (TMB) was calculated by R package “Limma”.

Immunohistochemistry (IHC)

We collected 20 pairs of HCC samples and adjacent non-tumor tissues from October 2018 to April 2019 in the Affiliated Hospital of Youjiang Medical College for Nationalities. Immunohistochemistry (IHC) was performed on the 20 samples to detect EXOSC10 protein expression by DAB staining.

All procedures were performed according to the Ethical Guidelines for Human Genome/Gene Research and approved by the Ethics Committee of Affiliated Hospital of Youjiang Medical College for Nationalities (2022090501). Furthermore, we received verbal informed consent from participants of our study.

Cell culture

The liver cancer cell lines (HepG2, MHCC97H, and Huh-7) and normal human liver cell line LO₂were purchased from Cell Bank of Type Culture, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences. Cells were cultured in DMEM, supplemented with 10% FBS (Fetal Bovine Serum) and 1% antibiotic/antimycotic solution. Cells were maintained at 37 °C in an atmosphere of 5% CO₂.

RNA extraction and quantitative real-time PCR

Total RNA was extracted using AxyPrep Multisource Total RNA Miniprep Kit (Cat No: Scipu003342; Axygen, Suzhou, China), and for mRNA quantification, RNA was reverse transcribed to cDNA using the Revert Aid First Strand cDNA Synthesis Kit (Cat No: K1621; Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s protocol. Quantitative real-time PCR was conducted using the FastStart Universal SYBR Green Master (Cat No: FSUSGMMRO; Roche, USA). The 2^−ΔΔCt method was used for quantitative. Folding changes of target genes are standardized through internal control.

The following primers were used:

EXOSC10(forward: 5′-GGAACCGTAAGGCAGCAGAAT-3′; reverse: 5′-TCTCGAAACTTGAGCTGAGGT-3′);

GAPDH (forward: 5′-GGACCTGACCTGCCGTCTAG-3′; reverse: 5′-GTAGCCCAGGATGCCCTTGA-3′).

Western blotting

Western blot was carried out according to the manufacturer’s instructions. The RIPA buffer with 1% phenylmethanesulfonyl fluoride was used to lyse cells, and extracted proteins were separated by 10% SDS-PAGE. Furthermore, the proteins were transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with skim milk for 2 h at room temperature, incubated with primary antibodies overnight at 4 °C, and were washed by Tris–HCl solution + Tween-20 (TBST) three times for 10 min. Subsequently, they were incubated with appropriate secondary antibodies for 1 h at room temperature. Finally, the blots were detected with enhanced chemiluminescence (ECL Plus). The density of protein bands was assessed by Image J software, and the protein levels were normalized to GAPDH. The dilution ratio of the antibody used is as follows: EXOSC10 (Cat No. 16731-1-AP, 1:3000; Proteintech, Wuhan, China), p53 (Cat No. AF0879, 1:1000; Affinity, Jiangsu, China), p-p53 (Cat No. AF3073, 1:1000; Affinity, Jiangsu, China), p21 (Cat No. AF6290, 1:1000; Affinity, Jiangsu, China), Bax (Cat No. 50599-2-Ig, 1:1000; Proteintech, Wuhan, China), Bcl-2 (Cat No.68103-1-Ig, 1:1000; Proteintech, Wuhan, China), PARP (Cat No. AF7023, 1:1000; Affinity, Jiangsu, China), N-cadherin (Cat No.22018-1-AP, 1:5000; Proteintech, Wuhan, China), E-cadherin (Cat No.20874-1-AP, 1:5000; Proteintech, Wuhan, China), GAPDH (Cat No.60004-1-Ig, 1:20000; Proteintech, Wuhan, China).

Transfection

EXOSC10 siRNA and negative control (NC) siRNA were purchased from GenePharma (Shanghai, China). The siRNAs and plasmids were transfected into cells by lipofectamine 3000 (Cat No. L3000015; Invitrogen Life Technologies, Carlsbad, CA, USA).

For siRNA knockdown, Huh-7 cells were seeded in 6-well plates. After 24 h, siRNA transfections were performed using Opti-MEM and Lipofectamine3000 transfection reagent according to the manufacturer’s instructions, transfected for 48 h, then harvested for further assays. The transfection effect was detected by Western blotting and quantitative real-time PCR, respectively. The following primers were used: Negative Control (NC) (5′–3′, S′-UUCUCCGAACGUGUCACGUTT; AS′-ACGUGACACGUUCGGAGAATT); siEXOSC10-2201 (5′–3′, S′-GCAGCGAAGUUC GAUCCAUTT; AS′-AUGGAUCGAACUUCGCUGCTT).

Cell Counting Kit-8 (CCK-8) assay

The growth inhibition rate of Huh-7 cells was assessed using the CCK-8 assay (Cat No. CK04; Dojindo Laboratories, Japan) assay. Cells (5 × 10³ cells per well) were seeded in 96-well plates with three replicate wells. After transfected for 24 h, 48 h, 72 h, and 96 h, a 10 µl volume of CCK-8 reagent was added to each well. Then measure the absorbance at 450 nm after incubation with the CCK-8 solution at 37 °C for 1 h. The inhibition rate (%) = (the OD values of NC − the OD values of siEXOSC10) / (the OD values of NC − the OD values of blank) ×100%.

Wound healing assay

Huh-7 cells were seeded in 6-well plates, after 48 h of transfection, cell confluence reached 90%. Then the 10 µl pipette tip was used to gently scratch the cell monolayer, washed twice with PBS, and removed the detached cells. The remaining cells were cultured in DMEM medium without FBS. The scratch areas were photographed at 0 and 24 h, respectively.

Migration assay

Twenty-four-well transwell plates were used to detect cell migration. Collect the cells and add them separately to the top compartments of the 24 well transwell plate. Meanwhile, add DMEM medium containing 10% FBS to the bottom of the transwell chamber. After 24 h, cells passing through the insert were stained with crystal violet and counted under a phase contrast microscope. Randomly select five fields and calculate the average number of cells inserted.

Colony formation assay

After 48 h of transfection, cells were collected and 1,000 cells were seeded in each well of a 6-well plate and cultured in DMEM medium containing 10% FBS for 2 weeks. The medium was changed after one week and then changed twice a week. After 2 weeks, cells were washed twice with phosphate buffer saline (PBS) and fixed in one mL of 4% paraformaldehyde fixation and methanol in each well for 20 min, respectively. After aspiration of methanol, the cells were washed twice with PBS again, stained with crystal violet staining solution for 20 min, photographed and counted.

Mitochondrial membrane potential measurement

The JC-1 kit (Cat No. C2006; Beyotime Biotechnology, Shanghai, China) was used to detect the membrane potential. NC or transfected cells were seeded in a 6-well plate and stained with JC-1 by incubation in culture solution for 20 min, and then each group of cells was observed with a fluorescence microscope.

Hoechst33342 accumulation assay

Hoechst 33342 dye stains (Cat No. C1025; Beyotime Biotechnology, Shanghai, China) were used in apoptosis detection. NC or transfected cells were seeded in a 6-well plate and stained with 1 ml of Hoechst 33342 dye by incubation in the culture solution for 30 min, then washed twice with PBS, observed each group of cells with a fluorescence microscope.

Flow cytometry

The Cell Meter™ FITC-Annexin V Binding Apoptosis Assay Kit (Cat No. 22839; AAT Bioquest, Newport Beach, CA, USA) was used to investigate the effect of EXOSC10 silencing on Huh-7 cell apoptosis. NC or transfected cells were cultured in a 6-well plate with an FBS-free medium for 48 h. Cells were resuspended in 200 µL assay buffer, 2 µL Annexin V-FITC, and 2 µL 100 × propidium iodide. The cells were also incubated in the dark at room temperature for 30 min. Finally, we analyze the cells with a flow cytometer.

Statistical analysis

Statistical analyzes were performed using R software v4.0.4 and SPSS 25.0. The one-way ANOVA or two-tailed unpaired Student’s t-test was performed to identify significant differences in the data. *P < 0.05, **P < 0.01, ***P < 0.001, N.S.: no significant difference (P  ≥ 0.05).

Expression of EXOSC10 in HCC patients

The mRNA level of EXOSC10 in HCC patients was evaluated according to TCGA LIHC data sets. As shown in Fig. 1A, HCC samples tended to express higher levels of EXOSC10 mRNA than normal samples. Additionally, the level of EXOSC10 mRNA was also significantly increased in 50 HCC samples compared to their adjacent liver tissues (Fig. 1B).

We evaluated EXOSC10 expression data by all patient characteristics derived from TCGA. In histological grade, we found that the expression of EXOSC10 in G3 was significantly increased in G1 (P = 0:0066) and G2 (P = 0:0065). The expression of EXOSC10 in stage I decreased significantly compared to that in stage II (P < 0.001) and stage III (P <  0.001). The expression of EXOSC10 in stage IV decreased compared to that in stage II (P = 0.033) and stage III (P = 0.018). In the T stage, the expression of EXOSC10 in T1 was significantly decreased than that in T2 (P < 0.001) and T3 (P < 0.001). However, EXOSC10 expression did not show any correlation with the M and N stages, respectively. These results suggested that the expression level of EXOSC10 was related to histological grade, T stage, and pathological stage (Figs. 1C–1I).

The correlation between EXOSC10 expression and prognosis in patients with HCC

We analyzed the correlation between EXOSC10 expression and the prognosis of patients with HCC. According to the Kaplan–Meier survival curve, HCC patients with high expression of EXOSC10 had poor overall survival (OS) and progression-free survival (PFS) (Figs. 2A, 2B). We established a time-dependent survival ROC curve based on the expression of EXOSC10 in the sample to predict 1-, 3-, and 5-year survival rates (AUC = 0.665, 0.620, and 0.565, respectively) (Fig. 2C). The results showed that the use of EXOSC10 expression to predict the prognosis of patients with HCC was reasonably accurate.

Furthermore, univariate Cox regression analysis showed that EXOSC10 expression significantly affects the prognosis of patients with HCC. Multivariate Cox regression analysis showed that EXOSC10 can be independent of prognostic factors of clinical characteristics in patients with HCC patients (Figs. 2D, 2E). To predict the OS of patients with HCC, we established a nomogram based on the expression of EXOSC10 and patient characteristics in the TCGA dataset (Fig. 2F). The calibration plot for the probability of 1-, 3- and 5-year OS showed good agreement between the observed OS and nomogram-predicted OS (Fig. 2G). The above studies collectively demonstrate that EXOSC10 expression is significantly correlated with the prognosis of HCC patients.

Potential pathways regulated by EXOSC10 in HCC

HCC patients were divided into high and low EXOSC10 expression groups according to the median expression level of EXOSC10. To explore the potential pathway-regulated EXOSC10, we identified 2,724 significant DEGs between high and low expression groups, of which 41 genes were down-regulated and 2,683 genes were up-regulated (Text S1). The 50 most significant genes are shown in Fig. S1. Go enrichment analysis revealed that these DEGs were primarily enriched in biological processes such as nuclear division, organelle fission, and chromosome segregation; cellular components such as condensed chromosomes, synaptic membrane, microtubules; molecular functions such as channel activity, motor activity of microtubules and passive transmembrane transporter activity. The most enriched KEGG pathways included cell cycle, neuroactive ligand–receptor interaction, and ECM receptor interaction (Fig. 3A). While GSEA analysis indicated that several tumor hallmarks were enriched in the high EXOSC10 expression group, such as E2F target, MTORC1 signaling pathway, PI3K-AKT-MTOR signaling pathway, p53 signaling pathway, TGF beta signaling pathway, and WNT beta-catenin signaling pathway (Fig. 3B). The results may give some insight into the cellular biological effects related to EXOSC10.

The correlation of EXOSC10 expression and tumor immune infiltration

The relationship between EXOSC10 expression and immune cell infiltration adjusted by purity (B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells) was investigated using TIMER. As shown in Fig. 4A, EXOSC10 has a significant positive correlation with tumor purity (r = 0.092, P = 8.91e−02), B cells (r = 0.263, P = 7.515e−07), CD8+ T cells (r = 0.275, P = 2.26e−07), CD4+ T cells (r = 0.34, P = 9.33e−11), macrophages (r = 0.434, P = 4.00e−17), neutrophils (r = 0.451, P = 1.08e−18), and dendritic cells (r = 0 .421, P = 4.61e−16) in HCC. Next, we performed a correlation analysis between EXOSC10 and the tumor microenvironment, immune/stromal/estimate scores in HCC. Compared to the low EXOSC10 expression group, the high EXOSC10 expression group had a significant decrease in the stromal score and the estimate score (Fig. 4B). However, the potential correlation of EXOSC10 expression and tumor immune infiltration needs to be further tested experimentally. In addition, we compared the fraction of immunocytes with the high or low expression groups of EXOSC10. Unfortunately, only B cell memory with high expression of EXOSC10 was significantly higher than those with low expression of EXOSC10 (Fig. 4C). Finally, we evaluated the correlation between EXOSC10 and immune cells (Figs. 4D–4F). And we found that EXOSC10 had a significant positive correlation with B cell memory (r = 0.28, P = 0.042) and a negative correlation with B cell naiveness (r = −0.3, P = 0.031).

Validation of EXOSC10 expression levels in HCC patients and liver cancer cell lines

Furthermore, we measured the expression of EXOSC10 mRNA and protein in three liver cancer cell lines (HepG2, MHCC97H and Huh-7) and the normal liver cell line LO₂ by quantitative real-time PCR and Western blotting, respectively. These results confirmed that EXOSC10 expression levels were significantly higher in liver cancer cell lines than in control LO₂cells (P < 0.05), consistent with the results of bioinformatic analysis (Figs. 5A–5B). These results suggested that up-regulation of EXOSC10 may be closely associated with the occurrence and development of HCC.

We also found that the expression of the EXOSC10 protein was up-regulated in HCC tissue compared to normal liver tissue by analyzing 20 pairs of HCC samples and adjacent non-tumor tissues (Fig. 5C). As shown in Fig. 5C, the positive area of EXOSC10 was higher in HCC than in normal liver tissue, and the expression of the EXOSC10 protein was located primarily in the cytosol and nuclear lumen.

Silencing EXOSC10 inhibited HCC cell proliferation and migration

To explore the function of EXOSC10 in liver cancer cells, we chose the Huh-7 liver cancer cell line to study. We transfected Huh-7 cells with EXOSC10 siRNA, resulting in a significant decrease in the level of expression of EXOSC10 mRNA and protein, suggesting effective elimination of EXOSC10 (Fig. 6A).

Then we evaluated the function of EXOSC10 in the growth and proliferation of HCC cells. As shown in Fig. 6B, the knockdown of EXOSC10 inhibited the proliferation of HCC cells. Interestingly, silencing of EXOSC10 decreased the colony formation of HCC cells (Fig. 6C). Furthermore, we investigated whether EXOSC10 affects the migration ability of HCC cells. Wound healing and transwell migration assays demonstrated that silencing the expression of EXOSC10 decreased the migration of Huh-7 cells (Figs. 6D–6E).

Silencing EXOSC10 reduced mitochondrial membrane potential and promoted cell apoptosis in Huh-7 cells

Next, we investigate whether EXOSC10 affects cell apoptosis in Huh-7 cells. First, we explore the detection of cell apoptosis using Hoechst 33342 dye. The results showed that apoptosis in the siEXOSC10 group was significantly higher than in the NC group (Fig. 7A). We found that the mitochondrial membrane potential in the NC group was significantly higher than in the siEXOSC10 group (Fig. 7B). Similarly, we confirmed that the apoptosis rate in the NC group was significantly lower than in the siEXOSC10 group by flow cytometer (Fig. 7C).

EXOSC10 regulated the p53 signaling pathway in Huh-7 cells

We found that PARP, N-cadherin, and Bcl-2 protein levels in the NC group were significantly higher than those of the siEXOSC10 group; protein levels of Bax, p21, p53, p-p53, and E-cadherin were significantly lower than those of the siEXOSC10 group (Fig. 8).