Overexpression of chaperonin containing TCP1 subunit 7 has diagnostic and prognostic value for hepatocellular carcinoma

Chaperonin containing TCP1 subunit 7 (CCT7) regulates the expression of many tumor-related proteins. We investigated the diagnostic and prognostic value of CCT7 expression for hepatocellular carcinoma (HCC). In datasets from The Cancer Genome Atlas and the Gene Expression Omnibus, CCT7 mRNA levels were greater in HCC tissues than adjacent normal tissues, and these results were validated using immunohistochemistry. In patients with early-stage disease and low alpha-fetoprotein expression, CCT7 expression was still higher in HCC tissues than normal tissues. Receiver operating characteristic curve analyses indicated that CCT7 expression had better diagnostic value than alpha-fetoprotein for HCC patients with early-stage disease and low alpha-fetoprotein expression. The positive predictive value of CCT7 expression was higher than that of alpha-fetoprotein expression. Higher CCT7 mRNA and protein levels were independent risk factors for poorer overall and recurrence-free survival in HCC patients. Greater methylation of the CpG site cg19515186 was associated with better overall survival in HCC patients. Genes co-expressed with CCT7 were upregulated in HCC and associated with poorer overall survival. Gene Ontology, Kyoto Encyclopedia of Genes and Genomes and Gene Set Enrichment Analyses demonstrated that CCT7 expression correlated with spliceosome signaling. These findings demonstrate that CCT7 has diagnostic and prognostic value for HCC.

Prognostic value analysis using CCT7 mRNA expression and clinicopathological data in the TCGA database.
We downloaded the gene expression RNAseq and clinical characteristics dataset from the TCGA database to study the relationship between the expression of CCT7 and the clinical outcome of HCC patients. According to the mRNA expression level, 372 HCC patients were divided into high and low CCT7 mRNA expression groups to evaluate the prognostic value of CTT7. We de ned overall survival (OS) as the time interval between surgery and death or between surgery and the last observation point. We de ned recurrence-free survival (RFS) as the time interval between the date of surgery and the date of diagnosis of any type of recurrence [23].
Prognostic value analysis using CCT7 protein expression and clinicopathological data in HCC cohort with 118 patients.
To further investigated the association between CCT7 protein expression with the clinical outcomes, we performed immunohistochemistry (IHC) staining using 118 HCC tissues and paired adjacent normal liver tissues which were collected from patients who underwent hepatectomy from February 2013 to November 2014 in 900 Hospital of the Joint Logistics Team. The shortest follow-up time is 5-years. We obtain follow-up data through re-examinations, telephone calls, and the Social Security Death Index. Assessment of liver function and tumor stages using Child-Pugh classi cation and the 2010 International Union Against Cancer Tumor-Node-Metastasis (TNM) classi cation system, respectively [24,25]. The inclusion criteria were: only one lesion or multiple lesions con ned to one liver lobe; patients without any distant metastasis; did not receive chemotherapy or TACE or immunotherapy before surgery; postoperative pathology con rmed hepatocellular carcinoma. This study has been performed in accordance with the principles of the Declaration of Helsinki and approved by the Human Research Ethics Committee of 900 Hospital of the Joint Logistics Team (Fuzhou, China). All participants were given written informed consent before surgery and the collection of the specimens.
118 HCC specimens were cut into 4 μm sections and xed in the special microscope glass slide. Then, the slides were depara nized and dehydrated using gradient concentrations of malondialdehyde and ethanol. Next, the slices were immersed in a boiled solution of Tris/EDTA (pH 9.0) for 20 minutes for antigen retrieval. The slides were immersed in 3% H2O2 for 10 minutes to inhibit endogenous peroxidase. Then, the slides were incubated with the primary antibody (1:250; 15994-1-AP, Proteintech, Wuhan, China) and the secondary antibody (1:50,000; KIT-5010; anti-rabbit/mouse IgG; Maixin Biotechnology Development Co., Ltd., Fuzhou, China) in turn, and washed three times with PBS. Finally, the sections were stained with 3,3'-diaminobenzidine and substrate chromogen (Dako) for 2 minutes at room temperature and then counterstained with hematoxylin for 40 seconds. The slides incubated with only a second antibody without primary antibody were as negative control. The IHC staining was assessed by two separate pathologists who are not informed of any patient information. A 5-point scale system was used to assess the protein expression of CCT7, as 0, 1, 2, 3, and 4 represented no positive cells, <25% positive cells, 26-50% positive cells,51-75% positive cells, and >75% positive cells, respectively. DNA methylation and genetic alteration of CCT7 in HCC.
We downloaded gene expression and DNA methylation data from gene expression RNAseq and Illumina Human Methylation 450 datasets in the TCGA database (https://xenabrowser.net/datapages/) [26], respectively. Then, we analyzed the correlation between CCT7 mRNA and DNA methylation and identi ed CpG sites that affect mRNA expression in the MethSurv database, which was A web tool to perform multivariable survival analysis using DNA methylation data [27]. Next, we investigated genetic alteration and mutation hotspots of CCT7 in the Liver Hepatocellular Carcinoma (TCGA, Firehose Legacy) dataset from the cBioPortal database (http://www.cbioportal.org/) [28]. Then, we analyzed the correlation between the genetic alteration of CCT7 and the prognosis of HCC patients.
GO and KEGG enrichment analysis and PPI network construction.
We analyzed the correlated gene of CCT7 using the liver hepatocellular carcinoma (LIHC) dataset from the cBioportal database and the LinkedOmics database, respectively [29]. Then, the overlapping correlated gene with Spearman's correlation greater than 0.6 in the cBioportal and the LinkedOmics database was screened as the co-expressed genes of CTT7. Next, the Functional Annotation Tool in the DAVID database was utilized to perform GO and KEGG enrichment analysis on the overlapping co-expressed genes of CTT7 to explore the gene role which may regulate the tumorigenesis and progression in HCC [30]. We constructed the protein-protein interaction (PPI) network on overlapping co-expressed genes of CCT7 in the STRING database and visualized it in the Cytoscape software (Version 3.7.2) [31]. In the process of GO and KEGG enrichment analysis, P<0.05 and false discovery rate (FDR)<0.25 were considered statistically signi cant.

GSEA enrichment analysis.
We downloaded the normalized gene expression dataset in the TCGA database and divided 373 HCC samples into high and low expressed groups based on the median of CTT7 expression as the critical point. Then, we performed KEGG enrichment analysis utilizing the GSEA software (Version 4.1.2). In this process, "c2.cp.kegg.v7.0.symbols.gmt" was selected as a functional gene set, and the number of permutations was set as 1000. We set all other parameters as the default settings. The pathway and gene enrichment with a normal p-value<0.05 and FDR q-value<0.25 were considered signi cantly enriched.
Statistical analysis. statistical analysis and generated the statistical gures using GraphPad Prism 6.0(GraphPad Software, Inc., San Diego, CA, USA). The association between the mRNA expression of CCT7 with the clinicopathological characters was analyzed using the two-tail Student t-test, Fisher's Exact test, or Wilcox test. Pearson's chi-square test was utilized to compare the categorical variables. Kaplan-Meier method with the log-rank test was utilized to estimate the survival curve of OS, RFS. Univariate and multivariate analysis with cox's proportional regression model was utilized to predict the risk factors or independent risk factors. Receiver operating characteristic (ROC) curve with area under the curve (AUC) was utilized to estimate the diagnostic value of mRNA expression of CCT7 and DNA methylation of cg19515186 for HCC. P<0.05 was de ned as a statistically signi cant difference unless otherwise stated.

Results
CCT7 mRNA is signi cantly upregulated in HCC and associated with poor survival in the TCGA database.
We analyzed the mRNA expression of CCT7 in the TCGA database and visualized it in the UALCAN database. The result showed a signi cant upregulation of CCT7 in the HCC compared with normal liver samples ( Figure 1A). In addition, the mRNA expression of CCT7 in HCC samples incrementally upregulated with increasing cancer stages ( Figure 1B) and tumor grade ( Figure 1C). Besides, the TNMplot database showed that CCT7 mRNA expression in metastatic samples was higher than tumor samples as well as normal samples ( Figure 1D).
High mRNA expression of CCT7 was correlated with poor survival and clinical outcomes in HCC.
High protein expression of CCT7 correlated with poor survival and clinical outcomes in a cohort with 118 HCC patients.
We found the protein expression of CTT7 in HCC tissues ( Figure 1H-I) signi cantly higher than normal liver tissues ( Figure 1G) in the Human Protein Atlas database, and this nding was supported by our IHC staining (Figure 2A-B). We divided 118 HCC patients into high CCT7 protein group (n=57) and low CCT7 protein group (n=61) based on the IHC score. High protein expression of CCT7 correlated to TNM staging (P=0.043), serum AFP level (P<0.001), tumor differentiation (P=0.010), vascular invasion (P=0.029), and recurrence (P=0.005) of HCC patients (Table 3). In addition, the Univariate Cox Regression analysis revealed that the tumor size ( (Table 4). In addition, the group of high CCT7 protein expression exhibited a worse OS time ( Figure 2C) and RFS time ( Figure 2D). CCT7 is a diagnostic biomarker for HCC.
We analyzed the mRNA expression of CCT7 in the GSE76427 ( Figure 3A), GSE54236 ( Figure 3C), and GSE136247 ( Figure 3E) datasets from the GEO database and the result showed that its mRNA level signi cantly upregulated in HCC tissues compared with the non-HCC tissues (all P<0.001). The ROC curve exhibited a well diagnostic signi cance with the AUC value of 0.847 ( Figure 3B), 0.673 ( Figure 3D), and 0.793 ( Figure 3F), respectively. Furthermore, the heat map showed the mRNA expression of CCT7 in HCC tissues was 90% higher than paired adjacent normal liver tissues in the GSE76427 dataset ( Figure 3G). CCT7 has higher positive predictive value than AFP for HCC diagnosis.
We compared the diagnostic e ciency of CCT7 and AFP mRNA expression for HCC in GEO and TCGA databases. The mRNA expression of CCT7 in HCC was signi cantly higher than normal liver tissues in the GSE25097 ( Figure 4A), GSE63898 ( Figure 4D), and the TCGA LIHC datasets ( Figure 4G). Furthermore, the ROC curve showed that the AUC value of CCT7 was also signi cantly higher than that of AFP in the GSE25097 (0.719 vs 0.677, Figure 4B), GSE63898 (0.803 vs 0.567, Figure 4E), and TCGA LIHC datasets (0.743 vs 0.616, Figure 4H). The best cut-off values for the diagnosis of CTT7 and AFP were identi ed based on the sensitivity and speci city of the ROC curve. We found that the CCT7 has a higher positive predictive value (PPV) than that of the AFP both in the GSE25097 (54.9% vs 44.1%, Figure 4C), GSE63898 (64.5% vs 28.1%, Figure 4F), and TCGA LIHC datasets (55.8% vs 41.3%, Figure 4I), even though they have the statistically similar negative predictive value (NPV). All these results revealed that CCT7 has a higher sensitivity for the diagnosis of HCC than AFP.
CCT7 has a better diagnostic value for HCC patients with low AFP expression.
As we know that AFP is increased in no more than 70% of patients with HCC. we evaluated the diagnostic value of CCT7 in HCC patients with low AFP expression using GSE25097 and GSE63898 datasets in GEO database. The AFP expression in cirrhosis and HCC patients was similar in the two datasets ( Figure 5A, D). CCT7 mRNA expression in HCC was signi cantly higher than the cirrhosis patients ( Figure 5B, E). The ROC analysis revealed that the AFP expression has no diagnostic value in both the GSE25097 and GSE63898 datasets with the AUC value of 0.588 and 0.535 (P>0.05). The CCT7 mRNA expression in two datasets has a signi cant diagnostic value with the AUC value of 0.724 (P<0.001, Figure 5C) and 0.803 (P<0.001, Figure 5F). These results demonstrated that CCT can be used as an accurate diagnostic biomarker for low AFP expression HCC patients.
CCT7 is a better diagnostic value than AFP for early-stage HCC patients.
We evaluated the diagnostic e ciency of CCT mRNA for early-stage HCC patients using the TCGA database. The ROC curve analysis exhibited that the AUC value of CCT for stage 1 HCC patients was signi cantly higher than that of AFP ( Figure 5G). In addition, we found that CCT7 mRNA expression in stage 1 HCC patients has a higher PPV than that of AFP in the TCGA database (50.3% vs 42.1%), and NPV of CCT7 and AFP were similar (92.0% vs 94.0%, Figure 5H). Moreover, we investigated the correlations of CCT7 with other diagnostic biomarkers (AFP, r=0.240; ACE, r=0.066; GPC3, r=-0.150; GPT, r=-0.126, Figure 5I-L) which were identi ed by previous research [32][33][34]. The results showed that CCT7 expression may be an independent diagnostic biomarker for patients with HCC.
Dysregulation of CCT7 expression was associated with DNA methylation status in patients with HCC.
The CCT7 mRNA expression in HCC was frequently upregulated, this dysregulation associated with copy number alteration ( Figure 6A). The analysis of Illumina Human Methylation 450 datasets in the TCGA database demonstrated that CCT7 mRNA level was negatively associated with DNA methylation status ( Figure 6B). We identi ed three CCT7-related methylated CpG sites (cg15777261, cg07135469, and cg19515186) in HCC using the MethSurv database ( Figure 6C). A CpG site (cg19515186) was found associated with survival time of HCC (P<0.001, HR (95%CI):0.49 (0.34-0.72); Figure 6D). In addition, the correlation analysis revealed that the methylation status of cg19515186 was negatively associated with the mRNA expression of CCT7 ( Figure 6E). Furthermore, the ROC curve analysis exhibited a signi cant diagnostic value of the methylation status of cg19515186 for HCC in the TCGA database (AUC=0.821, P<0.001, Figure 6F). Finally, the survival analysis revealed that the high methylation status of cg19515186 was associated with better OS ( Figure 6G). These results demonstrated that dysregulation of CCT7 expression was associated with DNA methylation status in patients with HCC.
Genetic alteration of CCT7 associated with poor survival in patients with HCC.
Analysis of co-expressed genes of CTT7 in patients with HCC.
We screen out the correlated gene of CCT7 using the liver hepatocellular carcinoma (LIHC) dataset in the LinkedOmics and the cBioPortal database, respectively. 11319 negatively and 8670 positively correlated genes with CCT in the LinkedOmics database were exhibited in the volcano plot and heat map ( Figure 7I-K). Then, 45 overlapping genes from two database with Spearman's value greater than 0.55 were identi ed as co-expressed gene of CCT7 ( Figure 7L). Next, we investigated the signi cant interactions of CCT 7 with 45 co-expressed genes with the highest con dence score (greater than 0.9) in the STRING database. A PPI network with 42 nodes and 288 edges was constructed and visualized in Cytoscape software. We noticed that 8 proteins (CCT2, CCT3, CCT4, CCT5, NOP56, RPL8, RPL27, and RUVBL1) directly interacted with CTT7 in the PPI network ( Figure 7M). In addition, the mRNA expression of these eight genes incrementally upregulated with normal, tumor, and metastatic tissues in HCC analyzed in the TNMplot database ( Figure 8A). Furthermore, we performed survival analysis in the GEPIA database and found that high expression of these eight genes was associated with worse OS in patients with HCC ( Figure 8B). Finally, the signi cantly positive correlations between CCT7 with these eight genes were validated in the TCGA database ( Figure 8C).
High CCT7 expression correlated with Spliceosome signaling pathway.
Previous research has suggested that CCT7 promoted the progression of endometrial cancer through Spliceosome signaling pathway [35]. In addition, Spliceosome pathway plays a signi cant role in the tumorigenesis and progression of several tumors [36][37][38][39]. We performed GO and KEGG analysis on 45 coexpressed genes of CCT7 in the DAVID database. AS shown in Figure 9A, co-expressed genes were most enriched in mRNA splicing via spliceosome, rRNA processing, protein folding, translation etc by GO-BP analysis. We also performed enrichment analysis for cellular component (CC) and molecular functions (MF) (Additional Figure 1). The Ribosome, Spliceosome, Purine metabolism, and Ribosome biogenesis in eukaryotes were greatly enriched in KEGG analysis ( Figure 9B). We next analyzed the KEGG pathway of genes enriched that are most relevant to the survival of HCC in the GEPIA database. KEGG analysis revealed that the Spliceosome pathway signi cantly correlated with the prognosis of patients with HCC ( Figure 9C). The genes that are most relevant to the survival of HCC were listed in Additional Table. 1. We further investigated the potential pathway that CCT7 regulate the tumorigenesis and development of HCC by performing GSEA. The co-expressed genes with CCT7 in HCC tissues are shown in Figure 9D. As shown in Figure 9E, we exhibited the top eight differentially regulated pathways. We noticed that the normalized enrichment score (NES) for the Spliceosome signaling pathway was 2.09, which demonstrated that Spliceosome pathway positively correlates with HCC tumorigenesis and progression. These results revealed that mRNA expression of CCT7 positively correlates with the Spliceosome signaling pathway in HCC.

Discussion
A large number of studies have demonstrated that several members of the CCT subunit play a signi cant role in tumor proliferation and migration in various cancers [10,40,41]. Huang et al reported that CCT8 was upregulated in HCC and promoted proliferation [42]. In addition, Zhang et al. revealed that CCT3 promotes HCC cell proliferation by facilitating mitotic and suppressing apoptosis eventually [43].
Moreover, CCT4, CCT6A and CCT6B have diagnostic and prognostic values for HCC [16]. CCT7 was highly expressed in HCC through the method of bioinformatics analysis[15], but its clinical diagnostic, prognostic value and gene function were rarely analyzed. This study is the rst systematic investigation of diagnostic value, clinical signi cance, and the gene function of CCT7 in HCC.
We analyzed the mRNA expression of CCT7 in the UALCAN database and revealed that its expression in HCC samples was signi cantly higher than in normal tissues. Besides, its expression exhibited incrementally upregulated with increasing cancer stages and tumor grade. Moreover, its expression in metastatic samples was higher than tumor samples. High mRNA expression correlated with poor OS. We further analyzed the association between CCT7 mRNA expression with clinical outcomes in the TCGA database. The results suggested CCT7 expression correlated with worse clinicopathological features and was an independent risk factor for worse OS and RFS. Next, we analyzed the protein expression of CCT7 in the Human Protein Atlas database and an HCC cohort with 118 patients. The results showed CCT7 protein was signi cantly upregulated in HCC tissues compared with the adjacent normal tissues. In addition, high protein of CCT7 was also associated with poor clinicopathological features and was an independent risk factor for worse OS and RFS of HCC. These results demonstrated that CCT7 was a diagnostic and prognostic biomarker for HCC.
We evaluated the diagnostic e ciency of CCT7 using independent datasets in the GEO and TCGA database. The ROC curve analysis showed that CCT7 has a higher AUC value than AFP, the current golden biomarker of diagnosis for HCC [44,45]. In addition, the positive predictive value of CCT7 was better than AFP. We also found that CCT7 was highly expressed in HCC patients with low APF expression, suggesting that CCT7 has a better diagnostic signi cance for these patients. For HCC patients with early-stage, CCT7 also exhibited a better AUC value (0.715 vs 0.599) and positive predictive value (50.3% vs 42.1%) than HCC in the TCGA database. These results demonstrated that CCT7 was an effective biomarker for HCC diagnosis, especially for low AFP expressed patients and early-stage patients. Moreover, Spearman's correlations analysis suggested that CCT7 expression may be an independent diagnostic biomarker for patients with HCC.
It is commonly known that DNA methylation was abnormal in all forms of cancer and numerous studies have demonstrated a broad range of genes silenced by DNA methylation in different cancer types [46][47][48]. Therefore, it is reasonable to speculate that dysregulation of CCT7 expression was associated with DNA methylation status in patients with HCC. We found that upregulation of CCT7 mRNA expression was correlated with hypomethylation status of the CpG site of cg19515186. In addition, the hypomethylation status of cg19515186 was associated with better OS in HCC patients. These results demonstrated that dysregulation of CCT7 expression was associated with DNA methylation status in patients with HCC.
We then explored the genetic alteration of CCT7 in HCC patients to gain more insight into the gene function. Our results showed that 41% of queried patients exhibited the CCT7 alteration in the cBioPortal database. Furthermore, its alteration predicted a worse prognosis that has poor overall survival, diseasefree survival, progression-free survival and disease-speci c survival. Therefore, it is reasonable to assume that dysregulation of CCT7 expression was associated with genetic alteration in HCC.
Next, we investigated the role of CCT7 in tumorigenesis and progression of HCC by performing GO, KEGG, and GSEA analysis. The results demonstrated that CCT7 was involved in the signaling pathway of Spliceosome both exhibited by GO, KEGG and GSEA. Interestingly, Spliceosome was one of the pathways of genes enriched that were most relevant to the survival of HCC in the GEPIA database. These results revealed that CCT7 play a role of oncogene which involved in tumorigenesis and progression through the Spliceosome signaling pathway.

Conclusion
In summary, this present study demonstrated that CCT7 mRNA and protein expression were signi cantly upregulated in HCC compared with adjacent normal liver tissues. High CCT7 expression was associated with poor clinical outcomes and prognosis. CCT7 was an effective biomarker for HCC diagnosis and prognosis, especially for low AFP expressed patients and early-stage patients. Upregulation of CCT7 was associated with the hypomethylation status of the CpG sites of cg19515186, and that demethylation