DNA methylation-mediated high expression of CCDC50 correlates with poor prognosis and hepatocellular carcinoma progression

Analysis of CCDC50 expression in pan-cancer

GEPIA (http://gepia.cancer-pku.cn/) is a web-based tool that quickly outputs customizable data results, and can be interactively used for analysing gene expression data of the TCGA clinical data and RNA-seq [13]. We analysed the CCDC50 expression across TCGA tumors, and the GTEx data and matched TCGA normal were contained as controls. UALCAN database is a powerful web-portal [14], herein, we detached CCDC50 protein expression via UALCAN web-portal.

The prognostic and clinical information of CCDC50 in HCC

We used the GEPIA [13] and PrognoScan [15] to examine the prognostic OS, and DSS of CCDC50 in HCC. The gene mutation features of CCDC50 in HCC were analysed via the cBioPortal [16]. We chose the three datasets: “MSK, Clin Cancer Res 2018 [17], MERiC/Basel, Nat Commun. 2022 [18], and TCGA. We explored the TIMER2.0 to estimate the immunological roles of CCDC50 in HCC [19].

Cell culture, siRNA and qRT-PCR

LIHC cell lines and LO2 cells lines were obtained from ATCC and cultured in RPMI-1640 medium supplemented with 10% FBS. CCDC50 siRNA kits (si-CCDC50#1: 5’- CCGUGCUUAUGCAGAUAGUTT-3’; si-CCDC50#2: 5’- GCAGCAAAUUCCAAGUCAATT-3’) and negative control siRNAs (si-NC: 5’-UUCUCCGAACGUGUC ACG UTT-3’) were purchased from GenePharma. Transfections were performed as previously described [20]. The qRT-PCR assay was conducted as previously documented [20]. The following primer sequences were used in this study: CCDC50: 5’-GACGACGCATTCAGGAGAAGA-3’, 5’-ACTATCTGCATAAGCACGGGTT-3’; β-actin: 5’-GTCTTCCCCTCCATCGTG-3’’, 5’-AGGGTGAGGATGCCTCTCTT-3’.

Western blot

Western blot assays were fulfilled as described previously [20]. The following primary antibodies were used in this finding: CCDC50 (A17836; Abclonal) and β-actin (sc-47778; Santa Cruz). All the assays were independently repeated thrice.

Cell proliferation assay

CCK8 assays were fulfilled as described previously [20]. For CCK8 assay, Hep3B and Huh7 cells were cultured in a 96-well plate supplemented with 200 μL of RPMI-1640 medium.

Correlation between CCDC50 and cancer drug sensitivity

GSCA [21] is an integrated platform combining clinical and small molecular drugs information, it can help us easily analyse CCDC50 expression with drug sensitivity GDSC and CTRP [22, 23].

Data availability statement

The data from this article can be obtained from the public database The Cancer Genome Atlas (https://portal.gdc.cancer.gov/).

CCDC50 was differentially expressed in multiple cancer including HCC

Firstly, we combined the TCGA and GTEx databases, and confirmed that the expression of CCDC50 was significantly lower in 14 cancer tissues than normal tissues. In addition, CCDC50 was up-regulated in 12 types of cancer including HCC (Figure 1A). As the main undertaker of life activities, the change of protein expression level is directly related to cancer progression [24]. To determine the protein levels of CCDC50 in different types of cancer, we analysed the UALCAN database and found that the protein expression of CCDC50 was low in 4 types of cancer, whereas it was high in 5 types of cancer including HCC (Figure 1B).

To explore CCDC50 expression levels in HCC, we used HPA and GEO datasets to validate its expression in HCC. We showed that CCDC50 was upregulated in HCC (Figure 2A, 2B). Furthermore, we showed that CCDC50 was correlated with adverse clinical features, such as grade, pathological stage and OS event, not related to NM stage (Figure 2C–2G). GEO datasets also confirmed that high CCDC50 expression had worse OS in HCC cancer patients (Figure 2H, 2I).

Prognosis value of CCDC50 in HCC

As CCDC50 expression was up-regulated in HCC tissues, we explored the prognostic value of CCDC50 in HCC. The KM survival curve results showed that increased expression of CCDC50 related to poor OS and DSS in HCC (Figure 3A, 3B). We comprehensively analysed the relationship between CCDC50 expression and OS in HCC. High CCDC50 expression had worse OS in diverse subgroups of HCC, including residual tumour (Figure 3C), gender (Figure 3D), age (Figure 3E), race (Figure 3F), histologic grade (Figure 3G), weight (Figure 2H), TNM stage (Figure 3H–3K), and tumour status (Figure 3L).

To examine whether CCDC50 can be used in HCC prognosis, we constructed a nomogram to predict the OS, and DSS in HCC patients and found that pathological stage and CCDC50 expression act as prognostic factors (Figure 4A, 4B and Tables 1–3). The calibration curves indicated that the nomogram can reliably predict the 1-, 3-, and 5-year OS, and DSS in HCC (Figure 4C, 4D). In summary, the model constructed above can optimize the prediction of different survival rates for liver cancer patients, and has high performance in prediction.

Gene mutation landscape of CCDC50 in HCC

The mutational data of CCDC50 from the cBioPortal that showed genetic alterations in the CCDC50 gene were observed in only 1.4% of the HCC patients (Figure 5A), and 1 missense site was found between amino acids ‘0’ and ‘306’ (Figure 5B). These results indicate that genetic alterations of CCDC50 may not be the main factor affecting its prognostic ability.

DNA methylation modulates CCDC50 expression in HCC

DNA methylation and m6A methylation play crucial roles in cancer progression. Therefore, it is crucial to reveal whether CCDC50 is regulated by DNA methylation and then abnormally expressed in hepatocellular carcinoma. First, we confirmed that CCDC50 was positively correlated with m6A regulatory factors expression in HCC (Figure 6A). Furthermore, we found that the level of DNA methylation of CCDC50 was downregulated in liver cancer tissue and was negatively correlated with liver cancer metastasis (Figure 6B, 6C). Finally, we found that the level of DNA methylation of CCDC50 was negatively related to the expression of CCDC50 in HCC (Figure 6D). However, the level of DNA methylation of CCDC50 did not affect the prognosis of HCC patients in TCGA LIHC datasets (Figure 6E, 6F). Interestingly, the level of DNA methylation of CCDC50 was positively correlated with the Tr1 and iTreg cell infiltration level (Figure 6G). These results partially suggest that DNA methylation may affect the infiltration levels of diverse immune cells by regulating the expression of CCDC50 in HCC.

Biological functions of CCDC50 in LIHC

Using the “clusterProfiler” R package, we performed functional annotations of CCDC50-associated differentially expressed genes (DEGs) in HCC patients, and 423 DEGs (mRNA and lncRNA) were captured, including 331 upregulated and 92 downregulated genes (Figure 7A–7D). The GO enrichment results showed that the DEGs were mainly involved in the extracellular space, acute-phase response, cell differentiation, chemoattractant activity, and epidermal growth factor receptor interactions (Figure 7E). KEGG results showed that the DEGs were mainly involved in the gastric cancer, pancreatic secretion, gastric acid secretion, pertussis, glycosaminoglycan biosynthesis-keratan sulphate, axon guidance, phenylalanine metabolism, nitrogen metabolism and protein digestion and absorption (Figure 7F).

GSEA results suggested that CCDC50 mainly participated in TOLL like receptor, T cell receptor, MAPK and wnt signaling pathway, chemokine signaling pathway (Figure 8A–8D). These results confirmed that CCDC50 regulated immune-related pathways and participate in the malignant progression of liver cancer.

Immune cell infiltration of CCDC50 in HCC

Immune cells have a critical role in cancer progression [25]. We used the TIMER database to reveal the relationship between CCDC50 expression and the infiltration levels of 24 immune cell types in HCC. The expression of CCDC50 significantly correlated with these 8 major immune cells in HCC (Figure 9A). CCDC50 expression levels showed a negative correlation with pDC (Figure 9B), Cytotoxic cells (Figure 9C), DC (Figure 9D), and Tgd (Figure 9E), and positive association with Eosinophils (Figure 9F), Tcm (Figure 9G), Th2 cells (Figure 7H), and T helper cells (Figure 9I). Due to the heterogeneity of tumor cells and immune cells, the expression of CCDC50 in hepatocellular carcinoma may be significantly different from the correlation between different immune cells. It is this difference that leads to tumor immune escape or the formation of tumor immune tolerance.

Knockdown of CCDC50 suppresses the malignant phenotype of LIHC

In vitro cell biology functional experiment validation of the functional role of CCDC50 in HCC was performed. Compared to LO2, a human normal hepatocyte, CCDC50 was highly expressed in LIHC cell lines (Figure 10A) and inhibited by siRNA in Hep3B and Huh7 cells. QRT-PCR and western blot were employed to examined the knockdown efficacy (Figure 10B, 10C). As expected, CCDC50 knockdown inhibited the proliferation and migration abilities of Hep3B and Huh7 cells (Figure 10D, 10E).

Cancer drug sensitivity analysis of CCDC50

In the GDSC database, the expression of CCDC50 positively related to the following drugs: EKB-569, WZ3105, KIN001-102, AT-7519, GSK690693, BMS345541, AT-7519, GSK1070916, NPK76-II-72-1 and TAK-715. However, it was negatively correlated with the following drugs: PLX4720, dabrafenib, (5Z)-7-Oxozeaenol, bleomycin, SB590885, CHIR-99021, selumetinib, TGX221, PD-0325901, AG-014699 and piperlongumine (r < −0.13, p < 0.0001) (Figure 11A). In the CTRP database, the expression of CCDC50 positively correlated with the following drugs: SR8278, GSK-J4, austocystin D, afatinib, linifanib and BRD-K41597374 (r > 0.16, p < 0.0001). However, it was negatively associated with vemurafenib, MLN2480, fluvastatin and lovastatin (r < −0.13, p < 0.0001) (Figure 11B). Due to the heterogeneity of tumor cells, there are certain differences in the correlation between CCDC50 and different drugs in different tumors, which may have a certain impact on different tumor treatments. These results show that CCDC50 is significantly related to drug sensitivity in diverse cancer cell lines, and it has the potential to be a promising cancer therapeutic target.