Identification of EFNA family as new potential prognostic biomarkers correlated with immune cell infiltration in hepatocellular carcinoma

Introduction

Background

Primary liver cancer ranks sixth globally in morbidity, and also the third leading cause of cancer-related deaths in 2022, with approximately 865,269 newly diagnosed cases and 757,948 deaths annually (1). Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer, representing 75–85% of cases (2-4). Despite significant progress in diagnostic techniques and treatment strategies, the 5-year survival rate for advanced HCC remains below 5% (5,6). Thus, there is an urgent need to identify new diagnostic markers for early detection and potential drug targets to enhance prognosis and offer personalized treatments for HCC patients.

Receptor tyrosine kinases (RTKs) are often deregulated in human cancers, resulting in increased RTK signaling that influences several cellular processes related to cancer progression (7). The hepatocyte Eph receptor is the largest RTK subfamily, capable of producing erythropoietin and interacting with its ephrin (EFN) ligand, which plays roles in normal physiology and disease pathogenesis (8). Increasing evidence indicates that the Eph-EFN system plays a role in human carcinogenesis. It is involved in abnormal cell proliferation, tumor angiogenesis, invasion, and the maintenance of cancer stem cells (9,10). Members of the Eph/EFN family are widely and abnormally expressed in various human cancers, including endometrial cancer (11), prostate cancer (12,13), esophageal cancer (14), colorectal cancer (15), and liver cancer (16). Overexpression of the Eph/EFN system leads to tumorigenesis by facilitating angiogenesis and metastasis through signal recognition in the tumor microenvironment (17). Therefore, the Eph/EFN system is acknowledged as a promising target for cancer therapy (18).

EFNs act as ligands for Eph receptors and are classified into two families based on structure and binding properties: EFNAs (EFNA1–5), which attach to the cell membrane via a GPI anchor, and EFNBs (EFNB1–3), which interact with the transmembrane protein domain (19). By regulating the interaction between the Eph receptor and the EFN ligand, these domains facilitate the formation of a homodimer, which subsequently induces tyrosine phosphorylation.

Many studies have shown that some EFNA family members are abnormally expressed and may act as prognostic biomarkers in different cancers. For instance, EFNA1 overexpression in HCC tissue and cell lines is associated with poor clinical outcomes in patients with alpha-fetoprotein (AFP)-producing HCC (19,20). Multivariate Cox regression analysis revealed that high levels of EFNA1 in HCC are independent prognostic factors associated with poorer survival outcomes (21). Feng et al. found that EFNA2 plays a crucial role in the initiation and progression of HCC by enhancing cancer cell survival (22). In high-grade serous carcinoma (HGSC), both transcription and protein levels of EFNA5 increase significantly with disease progression (23).

Despite studies demonstrating abnormal EFNA expression in patients with HCC and its associations with clinicopathological characteristics and clinical outcomes, a comprehensive bioinformatics analysis elucidating the potential functional roles of EFNA family genes in HCC initiation and progression is still lacking. As a core component of modern biological and biomedical research, genomic investigations have been revolutionized by the rapid advances in microarray and RNA sequencing (RNA-seq) technologies (24). The present study aims to systematically investigate the expression profiles of all EFNA family genes in HCC tissues relative to normal liver tissues, clarify their correlations with key clinicopathological features of HCC, and evaluate their prognostic value for HCC patients by mining and integrating the latest publicly available multi-omics datasets. We also intend to explore the potential biological functions and signaling pathways of EFNA family genes in HCC development via bioinformatics enrichment analyses, and further reveal their associations with tumor immune cell infiltration in the HCC microenvironment. Collectively, this study will identify potential EFNA family members that can serve as prognostic biomarkers for HCC, and the findings are expected to provide novel insights into the molecular mechanisms underlying HCC progression and offer promising candidate targets for the precision therapy of HCC patients. The workflow of the study design is illustrated in Figure S1. We present this article in accordance with the REMARK reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-aw-2532/rc).

Methods

GEPIA

GEPIA (http://gepia.cancer-pku.cn/) is a web resource providing information on RNA sequence expression based on The Cancer Genome Atlas (TCGA) and the Genotype-Tissue (GTEx) projects (25). The “Single Gene Analysis” section was used to perform expression profiles analysis, survival analysis, and Pearson correlation analysis. The cutoff value for the survival analysis was determined by the quartile method. We applied the “multiple gene comparison” section of GEPIA to conduct multiple gene comparison analyses of the EFNA family, employing the Liver Hepatocellular Carcinoma (LIHC) dataset. The P value cutoff was set at 0.05.

UALCAN

UALCAN (http://ualcan.path.uab.edu/analysis.html) is an interactive database that combines clinical information with RNA-seq data from cancer studies, including TCGA and the MET500 cohort (26). It provides information about the transcriptional levels of target genes in carcinomas compared to normal tissues. Additionally, it presents information on the correlation between messenger RNA (mRNA) expression levels and clinicopathologic characteristics. In this study, we used UALCAN to investigate the mRNA expressions of EFNAs and their association with clinicopathologic characteristics in LIHC. We assessed the differences in expression between groups using a two-sample Student’s t-test, setting the P value cutoff at 0.01.

TCGA database

We analyzed clinical data and raw RNA-seq read counts of EFNA family members from TCGA-LIHC and corresponding normal tissue data from GTEx. We acquired RNA-seq data in TPM format and performed log₂ transformation for differential gene expression analysis between samples (27). In this study, we downloaded clinicopathological parameters and mRNA expression data for 374 HCC specimens from the TCGA database (https://portal.gdc.cancer.gov/projects/ TCGA-LIHC) and analyzed them using R software. Baseline data showed the clinical characteristics of study data profiles in Table S1. We calculated P values, hazard ratios (HR), and 95% confidence intervals (CI) for the Kaplan-Meier survival curves using log-rank tests and univariate Cox regression. We used R version 3.6.3 to perform Cox regression analysis and assess the correlation between EFNA mRNA expression and the survival of HCC patients. Univariate Cox regression was used to estimate the influence of EFNAs mRNA expression and clinical parameters on the survival rate of HCC patients; those with P≤0.10 were included in the subsequent analysis. Additionally, we conducted multivariate Cox regression to analyze the association between EFNA expression and patient survival, adjusting for other parameters such as Child-Pugh stage and histologic grade. We estimated the association between EFNA family expression and the abundance of immune cells, including activated dendritic cells (ADC), B cells, and immature dendritic cells (iDC) in LIHC, using Spearman’s correlation with TCGA LIHC project level 3 HTSeq - RNAseq FPKM format data and clinical data.

The Gene Expression Omnibus (GEO)

The GSE14520 (GPL3921) gene expression microarray dataset was collected from the GEO (https://www.ncbi.nlm.nih.gov/geo/) via the GEO 1010 query package, which contained 225 tumor samples and 220 normal controls (28).

The International Cancer Genome Consortium (ICGC) database

RNA-seq and corresponding clinical information of 212 tumors and 177 adjacent normal tissues, were obtained from the ICGC-LIR-JP cohort (https://icgc.org/).

Human Protein Atlas (HPA)

HPA (https://www.proteinatlas.org) is a biological research platform, comprising immunohistochemical expression information for 20 types of cancers, with 12 individual subtypes per cancer (29). In our report, we used immunohistochemistry images retrieved from HPA to validate the protein expression of EFNAs genes in human HCC and normal tissues.

Kaplan-Meier plotter

Kaplan-Meier plotter (http://kmplot.com/analysis/) is a publicly available database that can give information on the association of 54,000 genes with survival in 21 cancers (30). Kaplan-Meier plotter was used to analyze the prognostic value of EFNA members for overall survival (OS) of HCC. A P value below 0.05 indicated a statistically significant difference.

cBioPortal

cBioPortal (www.cbioportal.org) is a user-friendly platform that offers visualization and access to large-scale genomics datasets of cancer (31). The genetic alterations and co-expression of EFNA genes in patients with LIHC were queried from the TCGA database via cBioPortal. We set the mRNA expression z-score threshold as 2.0 between the unaltered and altered patients.

LinkedOmics database

The LinkedOmics database (http://www.linkedomics.org/login.php) is a visual platform and is used to explore the gene expression profile (32). We used LinkedOmics to determine the EFNA coexpression genes by using Pearson’s correlation coefficient and showed the results via heat maps and volcano plots.

Enrichment analysis (Metascape)

Metascape (http://metascape.org) is a well-maintained platform for biological pathway enrichment analysis (33). EFNA family genes and the frequently altered neighboring genes identified from cBioPortal were uploaded to the “Custom Analysis” module. Gene Ontology (GO) biological process/molecular function/cellular component terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were tested. Enriched terms were ranked by P value, and terms with P<0.05 (Benjamini-Hochberg adjusted when available in Metascape outputs) were considered significant. For network analysis, Metascape generated a protein-protein interaction (PPI) network based on curated interaction databases and identified densely connected components using the MCODE algorithm; top clusters and representative enriched terms were reported.

Tumor Immune Estimation Resource (TIMER)

TIMER (https://cistrome.shinyapps.io/timer) is a comprehensive resource that could provide analyses with the dataset of 10,897 samples among diverse cancers in the TCGA database (34). EFNA family expression scatter plots and their relationship with the levels of immune cell infiltration in LIHC were assessed using Spearman’s rank correlation with TCGA_LIHC datasets. A bilateral Wilcoxon rank-sum test was conducted to detect the difference in the infiltration abundance affected by the somatic copy number alterations (SCNA) category. Statistical significance was identified as P<0.01.

Cancer Cell Line Encyclopedia (CCLE)

The CCLE database (https://sites.broadinstitute.org/ccle/) provides data on the genomes of over 1,100 cell lines from 44 different tumors, including a total of 56 cancer cell lines (35). The CCLE database data are mainly obtained by high-throughput sequencing, which contains five main dataset types: copy number, mRNA expression (Affymetrix), inverse phase protein array, reduced representation bisulfite sequencing, and mRNA expression (RNA-seq). The expression of EFNA genes in liver tumor cell lines was evaluated using the CCLE.

Clinical samples

HCC tissues and the corresponding adjacent normal tissues were collected from patients undergoing surgery at Fujian Medical University Union Hospital between April 2013 and April 2015. Patients who received neoadjuvant chemoradiotherapy were excluded. The tissues were stored in liquid nitrogen until use. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Fujian Medical University Union Hospital (No. 2020WSJK031) and informed consent was obtained from all individual participants or their legally appointed representatives participating in our study.

Immunofluorescence staining

HCC tissues and adjacent normal tissue were collected after surgery, then fixed in 10% neutral buffered formalin and embedded in paraffin. Tissue samples were then sectioned at 4 µm thickness. Sections were deparaffinized, rehydrated, processed for antigen retrieval, blocked, and incubated with primary antibody (EFNA1–5, 1:500 dilution) at 4 ℃ overnight, followed by incubation corresponding fluorescence-conjugated secondary antibody (1:500 dilution) for 1h at room temperature. Then slides were mounted with 4’,6-diamidino-2-phenylindole (DAPI), sealed, and photographed using an inverted microscope (Leica; Japan).

Statistical analysis

Statistical analyses were performed as described within each database module and using R (v3.6.3) for Cox regression. Group comparisons were conducted using Student’s t-test or Wilcoxon rank-sum tests where appropriate. Receiver operating characteristic (ROC) curves were generated to estimate diagnostic performance [area under the curve (AUC) with 95% confidence interval (CI)]. Survival differences were evaluated by log-rank tests, and HRs with 95% CIs were reported. Correlation analyses used Pearson or Spearman coefficients as specified. For immunofluorescence, staining included appropriate negative controls and all images were acquired using identical exposure settings; the current study provides qualitative validation and does not claim quantitative replicate-based effect estimates.

Results

Different expression patterns of EFNA family members in HCC

To explore the different clinical values of EFNA family genes in HCC patients, we used the TCGA, the GEO, ICGC, and the HPA database to analyze the mRNA and protein expression patterns. The TCGA, the GEO, along the ICGC database implied that EFNA1/3/4 mRNA expression level was considerably upregulated, while EFNA5 was decreased in HCC tissues (Figure 1A-1C). EFNA2 was downregulated in the TCGA database and the GEO database, while upregulated in the ICGC database. Additionally, we evaluated the relative EFNAs mRNA expression levels in HCC samples and observed that the highest expression among all the EFNAs we evaluated was EFNA1 (Figure 1D).

Figure 1 Different expression patterns of EFNAs family members in HCC patients. (A) The transcriptional of EFNAs in HCC tissues compared with normal tissues in TCGA. ***, P<0.001. (B) The transcriptional of EFNAs in HCC compared with normal (GEO). ***, P<0.001. (C) The transcriptional of EFNAs in HCC compared with normal (ICGC). ***, P<0.001. (D) The relative mRNA expression level of EFNAs in HCC compared with normal tissues in GEPIA. Color intensity represents the fold change expression of the genes in the tissue. (E) The protein levels of different EFNA family members in HCC tissues compared with normal liver tissues (HPA) (magnification ×100; immunohistochemistry stained; https://www.proteinatlas.org). GEO, Gene Expression Omnibus; GTEx, Genotype-Tissue Expression; HCC, hepatocellular carcinoma; HPA, human protein atlas; ICGC, International Cancer Genome Consortium; LIHC, Liver Hepatocellular Carcinoma; mRNA, messenger RNA; TCGA, The Cancer Genome Atlas; TPM, transcripts per million.

After detecting the mRNA expressions of EFNA family genes in HCC samples, we use HPA to further investigate the protein expressions of EFNA family members in HCC. The results indicated that the protein expressions of EFNA1/3/4 were considerably upregulated in HCC samples compared to normal liver tissues, while the protein expression of EFNA5 in HCC samples was lower than that in normal tissues (Figure 1E).

In conclusion, our findings indicate that EFNA1, EFNA3, and EFNA4 were significantly overexpressed at both transcriptional and protein levels in HCC patients across different databases. Meanwhile, EFNA5 was found to be downregulated, and EFNA2 displayed inconsistent expression levels in HCC.

Diagnostic value of EFNA family genes in HCC

We made use of the ROC curve to assess the specificity and sensitivity of EFNA family members to distinguish between HCC patients and healthy people. As shown in Figure 2A, EFNA1 (AUC =0.885, 95% CI: 0.849–0.922), EFNA3 (AUC =0.928, 95% CI: 0.903–0.953), and EFNA4 (AUC =0.963, 95% CI: 0.946–0.980) had great diagnostic capability. EFNA2 (AUC =0.631, 95% CI: 0.565–0.696) and EFNA5 (AUC =0.643, 95% CI: 0.593–0.692) were of moderate diagnostic capability (Figure 2A). Moreover, the diagnostic sensitivity, specificity, and accuracy of EFNA1/3/4 for HCC were evaluated and compared with AFP. We found that EFNA1/3/4 had higher diagnostic capability than AFP (Figure 2B). We also analyzed the diagnostic capability of EFNA1/3/4 and AFP in different tumor (T) stages, which showed that EFNA1/3/4 had higher diagnostic capability than AFP. EFNA1/3/4 was comparable to AFP as a serum marker for the diagnosis of HCC, combination of AFP and EFNA1/3/4 can elevate the sensitivity of diagnosis.

Figure 2 Diagnostic value of EFNA family genes in HCC. (A) ROC curves of the EFNA genes. (B) ROC curves of the EFNA genes in different T stage vs. normal. AFP, alpha-fetoprotein; AUC, area under the curve; CI, confidence interval; FPR, false positive rate; HCC, hepatocellular carcinoma; ROC, receiver operating characteristic; T, tumor; TPR, true positive rate.

Relationship between expression of EFNA family genes and clinicopathological characteristics in HCC

After the mRNA and protein expression profile was found in HCC patients, we utilized the GEPIA to evaluate the expression of different EFNA family genes expression with the respective pathological stages of HCC. Figure 3A uncovered that EFNA3 (P=3.74e−07), EFNA4 (P=0.008), and EFNA5 (P=0.03) expression were closely linked to tumor stage.

Figure 3 Relationship between the mRNA expression levels of EFNA family members and clinicopathological characteristics in HCC. (A) Associations between EFNA family members’ transcriptional and HCC patients’ tumor stage in GEPIA. (B) Associations of EFNA family members’ expression and tumor stages of HCC in UALCAN. (C) Associations of EFNA family members expression with tumor grades of HCC in UALCAN. (D) Associations of EFNA family members expression with HCC patients’ tumor nodal metastasis status in UALCAN. (E) Associations of EFNA family members’ mRNA expression with TP53 mutation status of HCC in UALCAN. *, P<0.05; **, P<0.01; ***, P<0.001. HCC, hepatocellular carcinoma; LIHC, Liver Hepatocellular Carcinoma; mRNA, messenger RNA; TCGA, The Cancer Genome Atlas.

We further analyzed mRNA expression levels of different EFNA family genes and their association with clinicopathological characteristics (containing tumor stage, tumor grade, lymph node metastasis, as well as TP53 mutation) of HCC patients using UALCAN. Figure 3B revealed that the expressions of EFNA1/2/3/4 were drastically related to the tumor stage of HCC, and overexpression of EFNA1/2/3/4 was found in the advanced tumor stage. The expression of EFNA1/2/3/4 in HCC stage III patients was the highest, which was higher than that in stage IV. However, the sample size was too small (only 6 cases in HCC stage IV), which may account for the statistics. In contrast, we found that patients with the T3–4 stage of HCC harbored an increased EFNA5 compared with patients at the T1–2 stages. Likewise, as displayed in Figure 3C, EFNA1/3/4 expression was notably related to tumor grade and tended to increase with higher tumor grade. However, the mRNA expression of EFNA5 in HCC patients with G3–G4 was higher compared to G1–G2. EFNA1 was significantly linked with lymph node metastasis, and the expression of EFNA1 increased with the lymph node metastasis stage (Figure 3D). In addition, we also explored the association between the TP53 mutations and EFNAs in HCC patients, and it revealed that EFNA3/4 expression in TP53 mutation patients was extensively higher compared to TP53 non-mutation patients (Figure 3E).

Collectively, these results indicated that EFNA1/3/4 expression levels were elevated in HCC, which were positively related to tumor stage and pathological tumor grade. In particular, EFNA3/4/5 tended to be elevated in advanced cancer stages and tumor grade. Patients of HCC in advanced tumor nodal metastasis status probably expressed higher mRNA of EFNA1. EFNA3 and EFNA4 were remarkably correlated with TP53 mutation in HCC.

The prognostic value of EFNAs mRNA expression in HCC

In Figure 4A, we used GEPIA to further understand the clinical outcomes of differential EFNA expression. Elevated transcription levels of EFNA1 (HR =1.9, P=0.01), EFNA3 (HR =2.5, P<0.001), EFNA4 (HR =2.6, P<0.001), and EFNA5 (HR =1.7, P=0.048) were correlated with shorter OS in HCC. We also accessed the value of EFNA family genes in the disease-free survival (DFS) of HCC and observed that low transcription of EFNA3 (HR =1.9, P=0.004) was remarkably related to longer DFS in HCC (Figure 4B).

Figure 4 The prognostic value of EFNAs mRNA expression in HCC. (A) The impact of different EFNAs mRNA expression on the OS curve of HCC patients (GEPIA). (B) The impact of different EFNAs mRNA expression on disease-free survival curve of HCC patients (GEPIA). (C) The impact of different EFNAs mRNA expression on OS of HCC patients (Kaplan-Meier Plotter). (D) The impact of different EFNAs mRNA expression on relapse-free survival curve of HCC patients (Kaplan-Meier Plotter). CI, confidence interval; HCC, hepatocellular carcinoma; HR, hazard ratio; LIHC, Liver Hepatocellular Carcinoma; mRNA, messenger RNA; OS, overall survival; TCGA, The Cancer Genome Atlas.

In addition, the Kaplan-Meier plotters were employed to evaluate the prognostic value of EFNAs in HCC. EFNA1 (HR =1.57, P=0.02), EFNA3 (HR =1.93, P<0.001), EFNA4 (HR =2.31, P=1e−05), and EFNA5 (HR =1.77, P=0.001) upregulation were closely related to shorter OS in HCC (Figure 4C). The relapse-free survival (RFS) curves showed lower mRNA expression levels of EFNA2 (HR =0.63, P=0.01), higher levels of EFNA3 (HR =1.87, P<0.001), and EFNA4 (HR =1.92, P=0.001) were significantly related to unfavored RFS (Figure 4D).

Analysis of TCGA data indicated similar results that upregulation of EFNA1 (HR =1.55, P=0.02), EFNA3 (HR =1.91, P=0.001), EFNA4 (HR =2.26, P<0.001), and EFNA5 were closely correlated with the shorter OS in HCC (Figure S2A). The upregulation of EFNA1 (HR =1.71, P=0.03), EFNA3 (HR =2.05, P=0.002), EFNA4 (HR =2.45, P<0.001), and EFNA5 (HR =1.65, P=0.03) mRNA levels were considerably correlated with unfavorable disease-specific survival (DSS) (Figure S2B). Lower EFNA3 (HR =1.58, P=0.01) and EFNA4 (HR =1.68, P=0.003) expressions were related to better progress-free interval survival (PFI) (Figure S2C).

Since chronic viral infection and alcohol consumption are common risk factors for HCC (36), we investigated whether EFNA family overexpression had any additional effect on the patient’s survival. EFNA3/5 had a significantly worsening effect overall in patients without alcohol consumption. The HR value was 1.64 and 1.73, respectively (Figure 5A). EFNA1/3/4/5 had worsening effects overall in patients with alcohol consumption. The HR value was 2.29, 3.32,6.53 and 2.04, respectively (Figure 5B). Conversely, patients with alcohol consumption had a favorite OS in the EFNA2 overexpression group (Figure 5B). Patients without viral hepatitis rendered a worse OS status in the EFNA2 upregulation group (Figure 5C). In contrast, patients with viral hepatitis had a favorite OS in the EFNA2 overexpression group (Figure 5D). EFNA3/4 upregulation had worsening effects overall in patients either with or without viral hepatitis (Figure 5C,5D). When combining alcohol consumption and viral hepatitis, EFNAs had not significantly worsening effect overall in patients without both alcohol consumption and hepatitis (Figure 5E). EFNA3/4/5 had a significantly worsening effect overall in patients with alcohol consumption but without viral hepatitis (Figure 5F). EFNA2 had a favorite OS in patients without alcohol consumption but with viral hepatitis (Figure 5G). Conversely, patients without alcohol consumption but with viral hepatitis showed a worse OS status in EFNA3/4 overexpression.

Figure 5 EFNA family expression, alcoholic and viral effects the OS in HCC. (A,B) The survival status was analyzed on the Kaplan-Meier Plotter platform in patients with or without alcohol consumption. Case numbers in each group were listed at the bottom of the figure. (C,D) The survival status was analyzed in patients with or without viral hepatitis. (E,F) The survival status was analyzed in without viral hepatitis patients with or without alcohol consumption. (G) The survival status was analyzed in without alcohol consumption patients with viral hepatitis. CI, confidence interval; HCC, hepatocellular carcinoma; HR, hazard ratio; OS, overall survival.

Independent prognostic value of EFNAs mRNA expression and OS in HCC patients

After finding that EFNA1/3/4/5 mRNA expression was remarkably correlated with the clinical outcomes of HCC, we further attempted to evaluate the independent prognostic value of EFNA1/3/4/5 expression in terms of OS in HCC. We used open-source software to download clinical data and EFNAs mRNA expression of 374 HCC patients in the TCGA database (Table S2) for Cox regression analysis. Univariate analysis showed high T stage (HR =2.126, 95% CI: 1.481–3.052, P<0.001), metastasis (M) stage (HR =4.077, 95% CI: 1.281–12.973, P=0.02), pathology (HR =2.090, 95% CI: 1.429–3.055, P<0.001), high EFNA3 (HR =1.609, 95% CI: 1.136–2.280, P=0.007), EFNA4 (HR =1.726, 95% CI: 1.215–2.450, P=0.002), EFNA5 (HR =1.748, 95% CI: 1.227–2.488, P=0.002) expression indicated shorter OS in patients with HCC (Table 1). Multivariate analysis showed that upregulation of EFNA1 (HR =1.494, 95% CI: 1.024–2.178, P=0.04), EFNA4 (HR =1.699, 95% CI: 1.168–2.473, P=0.006), EFNA5 (HR =1.669, 95% CI: 1.150–2.422, P=0.007) were independently related to poor OS of HCC patients (Tables S2-S6, Figure S3).

Together, these data indicated that elevated expressions of EFNA1/4/5 were HCC patients’ independent prognostic factors for poor OS.

The genetic alteration and interaction analyses of EFNA family members in HCC

Next, we investigated genetic alteration in the EFNAs gene and its association with OS in HCC patients. We revealed a high alteration rate of EFNAs in HCC tissues (Figure 6A). Among the 360 sequenced HCC patients, 179 exhibited gene alterations, resulting in a mutation rate of 49.72%. The alteration rates for EFNA4, EFNA1, and EFNA3 were 31%, 29%, and 24%, respectively. These rates ranked these genes as having the highest frequency of mutations. Additionally, the Kaplan-Meier plot and log-rank test indicated a significant association between EFNA gene alterations and poor OS in HCC patients (Figure 6B, P=4.987e−3). These results suggested that EFNA gene alterations may significantly affect the prognosis of HCC.

Figure 6 EFNA family gene alteration analysis and interaction analyses in HCC. (A) Genetic alteration of different EFNA family genes in HCC (cBioPortal). (B) Association between genetic alterations in EFNAs with OS. (C) Correlation between EFNA family members (GEPIA). (D) Correlation between EFNA family genes and TP53 (GEPIA). Cor, correlation; HCC, hepatocellular carcinoma; OS, overall survival; TPM, transcripts per million.

Besides, the correlations of EFNAs with each other in HCC patients were evaluated via the GEPIA online web, with Pearson’s correction included. There were co-expression associations between the following EFNA proteins: EFNA1 positively with EFNA2/3/4, EFNA2 with EFNA1/3/4, EFNA3 with EFNA1/2/4/5, EFNA4 positively with EFNA1/2/3/5, EFNA5 positively with EFNA3/4 (Figure 6C). Furthermore, the expression of EFNA1/2/3/4/5 was positively co-expression with TP53 (Figure 6D).

Correlated significant genes with EFNAs in HCC

The LinkedOmics database investigated genes correlated significantly with the EFNA family members. The top 50 correlated genes were shown in the volcano (Figure 7A) and heatmap plot (Figure 7B,7C). We found that the most negatively correlated genes with EFNA1 included ARL6IP5, MYO9A, and LOC653653 while the positively correlated genes with EFNA1 included BRP44, TM4SF5, and ELF3. The most negatively correlated genes with EFNA2 included FKBP1A, SNRPD2, and RBPJ while the positively correlated genes with EFNA2 included MYO18A, ANKRD56, and NFIC. The most negatively correlated genes with EFNA3 included FAT4, DPF3, and MPDZ while the positively correlated genes with EFNA3 included EFNA4, JMJD6, and LOC92659. The most negatively correlated genes with EFNA4 included PPAP2B, RNF125, and SNRK while the positively correlated genes with EFNA4 included VPS72, PYGO2, and PRCC. The most negatively correlated genes with EFNA5 included METTL7A, ETNK2, and CAT while the positively correlated genes with EFNA5 included PDLIM7, PKM2, and C3orf52.

Figure 7 The top 50 positively correlated genes and top 50 negatively correlated genes of the EFNA family genes in patients with HCC. (A) A volcano plot of the EFNA family and its co-expressed genes in HCC. Red dots: significantly positively correlated genes. Green dots: significantly negatively correlated genes. Black dots: non-significantly correlated genes. Top correlated genes are labeled. (B) The top 50 genes with a positive correlation with EFNA family genes are visualized in a heatmap. (C) The top 50 genes with a negative correlation with EFNA family genes are visualized in a heatmap. HCC, hepatocellular carcinoma.

Enrichment of EFNA family members and their 50 frequently altered adjacent genes ontology in HCC

Using cBioPortal, we investigated 50 neighboring genes closely related to EFNA gene alterations. Some genes were positively associated with EFNA family members, whereas others were negatively associated with the proteins. We performed GO analysis, KEGG analysis, and PPI enrichment analysis using Metscape. The functions enrichment and predicted pathways of the EFNA gene and similar genes are displayed in Figure 8A: biological processes such as GO:0022411 (cellular component disassembly), GO:0001649 (osteoblast differentiation), GO:0042060 (wound healing), GO:0034446 (substrate adhesion-dependent cell spreading), and GO: 0051091 (positive regulation of DNA-binding transcription factor activity) are significantly regulated by mutations of EFNAs in HCC. Cellular components, including GO:0005743 (mitochondrial inner membrane), GO:0005758 (mitochondrial intermembrane space), GO:0071438 (invadopodium membrane), GO:0042641 (actomyosin), and GO:1904813 (ficolin-1-rich granule lumen) were prominently associated with the EFNAs alterations in HCC. Furthermore, alterations in EFNAs significantly impact various molecular functions, including GO:0019904 (protein domain-specific binding), GO:0005178 (integrin-binding), GO:0004714 (transmembrane receptor protein tyrosine kinase activity), GO:0042803 (protein homodimerization activity), and GO:0008134 (transcription factor binding).

Figure 8 Enrichment of EFNA family members and their 50 frequently altered adjacent genes ontology in HCC patients. (A) GO analysis and KEGG analysis of EFNA genes and similar genes. (B) Network of enriched terms of EFNA genes and adjacent similar genes; PPI enrichment analysis of these genes and MCODE components identified in the gene lists (Metascape). BP, biological process; CC, cellular component; GO, Gene Ontology; HCC, hepatocellular carcinoma; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function; mRNA, messenger RNA; PPI, protein-protein interaction.

In KEGG analysis, hsa04151 (PI3K-Akt signaling pathway), hsa04931 (insulin resistance), hsa05225 (HCC), and ko04064 (NF-kappa B signaling pathway) were correlated to the EFNAs mutations in HCC (Figure 8A). Figure 8B displays the PPI network of involved genes including adherens junction, positive regulation of DNA-binding transcription factor, PI3K-Akt signaling pathway, and positive regulation of transferase activity.

Immune cell infiltrations analysis of EFNA family members in HCC

We comprehensively studied the relationship between differential EFNA gene expression and immune infiltration in HCC, using TIMER database. EFNA1 expression was positively related with B cells [correlation (Cor) =0.13, P=1.61e−2], CD4⁺ T cells (Cor =0.134, P=1.31e−2), neutrophil cells (Cor =0.124, P=2.13e−2), and dendritic cells (Cor =0.189, P=4.63e−4). EFNA2 expression was inversely associated with CD8⁺ cell infiltration (Cor =−0.13, P=1.08e−2), while positively associated with CD4⁺ T cell infiltration (Cor =0.318, P=1.68e−9), and neutrophil cell infiltration (Cor =0.128, P=1.70e−2). EFNA3 was positively correlated with B cell infiltration (Cor =0.176, P=1.06e−3), CD8⁺ T cell infiltration (Cor =0.153, P=4.6e−3), CD4⁺ T cell infiltration (Cor =0.18, P=8.2e−4), macrophage cell infiltration (Cor =0.25, P=3.08e−6), neutrophil cell infiltration (Cor =0.258, P=1.23e−6), and dendritic cell infiltration (Cor =0.25, P=3.03e−6). Similarly, EFNA4 was positively linked with B cell infiltration (Cor =0.312, P=3.15e−9), CD8⁺ T cell infiltration (Cor =0.183, P=6.53e−4), CD4⁺ cell infiltration (Cor =0.267, P=4.88e−7), macrophage cell infiltration (Cor =0.302, P=1.31e−8), neutrophil cell infiltration (Cor =0.237, P=8.54e−6), and Dendritic cell infiltration (Cor =0.291, P=4.69e−8). EFNA5 was positively linked with B cell infiltration (Cor =0.238, P=8.42e−6), CD8⁺ T cell infiltration (Cor =0.224, P=3.03e−5), CD4⁺ T cell infiltration (Cor =0.418, P=5.46e−18), macrophage cell infiltration (Cor =0.385, P=1.59e−13), neutrophil cells (Cor =0.359, P=6.5e−12), and dendritic cells (Cor =0.301, P=1.45e−8) (Figure 9A).

Figure 9 Immune cell infiltrations analysis of EFNA family members in HCC. (A) Associations of EFNA family genes expression and immune cell infiltration (TIMER). (B) Correlations between EFNA family genes and immune cell infiltration (TCGA). Cor, correlation; HCC, hepatocellular carcinoma; LIHC, Liver Hepatocellular Carcinoma; TCGA, The Cancer Genome Atlas; TIMER, Tumor Immune Estimation Resource; TPM, transcripts per million.

The TCGA LIHC project, utilizing level 3 HTSeq-FPKM format RNAseq data and clinical data, revealed that members of the EFNA family, including EFNA1, EFNA3, and EFNA4, negatively correlate with the infiltration of most immune cells. In contrast, EFNA5 shows a positive association with immune cell infiltration (Figure 9B).

These findings collectively indicate that EFNA family members may interact with immune cell infiltration, which could influence the outcomes for patients with HCC.

Validation of EFNA family members in liver cancer lines and clinical samples

The expression of EFNA genes in different liver cancer cell lines was investigated via the Cell database, and we found that EFNA1/4 had a high expression level in almost all liver cancer cell lines (Figure S4). To validate the findings in the above databases and further reveal which EFNA members play a crucial role in the progression of HCC, we used immunofluorescence staining to investigate the protein expression of EFNAs in HCC tissues. Immunofluorescence staining results showed that the expression of EFNA1/3/4 was much higher in HCC tissues than that in the adjacent normal tissues, EFNA2/5 was downregulated in HCC (Figure 10).

Figure 10 Validation of EFNA family members in clinical samples. IF staining for EFNA genes in HCC tissues and normal liver tissues. (A) EFNA1 was significantly higher in HCC tissues than that in precancerous tissues. (B) EFNA2 was significantly lower in HCC tissues than that in precancerous tissues. (C) EFNA3 was significantly higher in HCC tissues than that in precancerous tissues. (D) EFNA4 was significantly higher in HCC tissues than that in precancerous tissues. (E) EFNA5 was significantly lower in HCC tissues than that in precancerous tissues. Scale bars =50 μm. DAPI, 4’,6-diamidino-2-phenylindole; HCC, hepatocellular carcinoma; IF, immunofluorescence.

Discussion

The Eph-EFN system plays a crucial role in development and tissue homeostasis. Previous literature has extensively documented its abnormal expression in various human malignancies. While the role of EFNAs in the carcinogenesis and progression of human malignancies is partially understood, a comprehensive bioinformatics analysis of HCC has not yet been conducted. In this study, we evaluated the expression profile, prognostic values, and tumor immunity of various EFNA family genes in HCC using updated public resources for the first time. Our primary findings showed that EFNA1, EFNA3, and EFNA4 were significantly upregulated in HCC, while EFNA5 was markedly downregulated. EFNA2 exhibited variable expression across different databases. Additionally, the protein expression levels of EFNA1, EFNA3, EFNA4, and EFNA5 in the HPA database corresponded with their mRNA levels. The expression levels of EFNA2, EFNA3, EFNA4, and EFNA5 were higher in stage II and stage III compared to stage I. High expression levels of EFNA1, EFNA3, EFNA4, and EFNA5 were associated with poor OS. Furthermore, genes correlated with the EFNA family were enriched in pathways related to adherens junctions, DNA-binding transcription factors, the PI3K-Akt signaling pathway, and transferase activity. The expressions of EFNA1, EFNA3, EFNA4, and EFNA5 are significantly linked to tumor purity. Additionally, all EFNA expressions are associated with varying levels of immune cell infiltration. We hope that our data will enhance the discovery of biomarkers, optimize treatment strategies, and improve clinical outcomes for patients with HCC.

EFNA1, the first member of the EFNA family discovered in cancer cells, is a protein induced by TNF (37). The sentence structure is complex; it can be broken down for better flow. Most studies indicate that EFNA1 is highly expressed in renal, colorectal, and gastric cancers. This high expression is significantly associated with poor clinical outcomes across various tumors (19,38,39). In our study, data from TCGA, GEO, ICGC, and HPA confirmed that EFNA1 mRNA and protein expression levels are higher in HCC compared to normal liver tissues, consistent with previous findings. Wada found that EFNA1 expression is higher in cirrhotic tissue than in normal tissue, with the highest levels observed in HCC (21). Further studies showed that EFNA1 influenced the expression of genes associated with tumor cell growth, angiogenesis, as well as metastasis (matrix metalloproteinases and integrin) in HCC (20). In our present research, EFNA1 expression was markedly related to advanced tumor grade as well as nodal metastasis status. Kaplan-Meier Plotter survival analysis revealed that high expression of EFNA1 indicated unfavorable OS and disease-free survival (DFS) in HCC. EFNA1 was related to poor OS in patients with alcohol consumption. Additionally, EFNA1 serves as an independent prognostic factor linked to poor OS in HCC patients, indicating its potential carcinogenic role in this disease.

EFNA2 expression was significantly higher in HCC, especially in patients with portal vein invasion (22). Additionally, in vitro experiments showed that EFNA2 activates the Akt/nuclear factor-kappa B (NF-κB) pathway in HCC cells, which helps the cancer cells survive (22). In our report, EFNA2 expression in HCC was higher in the TCGA and GEO databases, while the ICGC data did not support this finding. We concluded that the inconsistency may stem from the different histopathological types and tumor stages present in the samples. Our results further proved that EFNA2 was upregulated in patients with advanced tumor stage in HCC. However, lower EFNA2 expression predicted poor RFS in HCC patients, which seemed inconsistent with EFNA2 as a carcinogen. EFNA2 did not affect OS in patients without alcohol consumption and viral hepatitis. However, lower EFNA2 expression predicted poor recurrence-free survival (RFS) in HCC patients, which appears inconsistent with EFNA2’s role as a carcinogen.

The overexpression of EFNA3 was closely associated with the high metastatic potential of breast cancer cells (40). Barderas et al. also identified a difference in EFNA3 expression between highly metastatic colon cancer cells and non-metastatic colon cancer cells (41). Furthermore, data from clinical samples also indicated a significant correlation between elevated EFNA3 expressions and metastasis risk (40). Mechanistically, EFNA3 enhances the metastatic potential of cancer cells by repelling the vascular endothelium and improving their ability to enter and exit blood vessels (40). In our current study, we found that EFNA3 levels were elevated at both the mRNA and protein levels in HCC tissues compared to normal liver samples. EFNA3 expression was remarkably linked to advanced tumor stage and grade in HCC. Furthermore, the overexpression of EFNA3 was significantly associated with poor clinical outcomes in HCC, strongly indicating its role as an oncogene. Subgroup analysis shows EFNA3 had no affection on the OS in patients without alcohol consumption and viral hepatitis. While overexpression of EFNA3 had worse OS in patients either with alcohol consumption or viral hepatitis, which indicated that EFNA3 participated in the liver carcinogenic process associated with alcohol consumption and viral hepatitis.

Recent reports show that EFNA4 is overexpressed in several types of cancer, including osteosarcoma (42) and oral squamous cell carcinoma (OSCC) (43). Elevated EFNA4 was positively associated with NANOG and the octamer-binding transcription factor 4 (OCT4). The co-expression of EFNA4 with either NANOG or OCT4 was linked to poorer RFS in OSCC (43). However, no studies have been reported on its expression and prognosis value in HCC. In our study, we found that EFNA4 expression was significantly upregulated in HCC samples. Additionally, EFNA4 levels were closely related to cancer stage, tumor grade, and TP53 mutation status. Increased EFNA4 expression in HCC was strongly associated with poorer OS, RFS, DSS, and progression-free interval (PFI). Multivariate Cox regression analysis further verified that EFNA4 was an independent predictor for shorter OS in HCC, which indicated that EFNA4 contributed to the tumorigenesis of HCC.

Through differential gene expression analysis, we found that EFNA5 expression is decreased in HCC compared to matched normal samples from various online databases. However, our analysis of EFNA5 expression in HCC samples revealed that EFNA5 is up-regulated in later tumor stages and poorer tumor grades. Recent studies indicate that, among histopathological ovarian cancer (OC) subtypes, EFNA5 expression is particularly elevated in the aggressive HGSC, aligning with our findings. To explore the potential role of EFNA5 in HCC, we conducted survival analysis and both univariate and multivariate Cox regression analyses using data from the TCGA database. We found that the high EFNA5 expression could predict an unfavorable prognosis in HCC. These conflicting results raise questions about the role of EFNA5 in the initiation and progression of HCC. By reviewing existing literature, we noticed that EFNA5 was reported to be a putative cancer suppressor gene, which was decreased in glioma (44), colorectal cancer (45), and leukemia (46). The tumor suppressor signal triggered by the Eph2-EFNA complex usually participates in the progression of invasive cancer (47). On the contrary, it is reported that EFNA5 is overexpressed in OC (48), pancreatic ductal adenocarcinomas (49), and pancreatic cancer (50), which is related to poor prognosis. Therefore, the clinical significance of EFNA5 seems to depend on the type of cancer and tumor microenvironment. The regulatory mechanisms of EFNA5 in HCC are unknown. According to our findings, we speculate that EFNA5 may not participate in the occurrence of HCC, but it is closely associated with the progression of HCC. Further studies are required to verify whether EFNA5 takes part in the tumorigenesis of HCC.

Additionally, genetic analysis revealed significant alterations in EFNA family genes among HCC patients, with the most common alteration being high mRNA expression. A connection exists between different EFNA family members, suggesting that these proteins may have either antagonistic or synergistic effects on HCC tumorigenesis. We identified the 50 genes most closely associated with each EFNA family member using cBioPortal. These genes were further annotated based on GO enrichment analysis as well as KEGG pathway enrichment analysis. The results showed that these genes were primarily enriched in DNA-binding transcription factors and the PI3K-Akt signaling pathways. Additionally, they were associated with HCC pathways. EFNA1 stimulates pLE cell growth and migration by activating PI3K and MAPK signaling pathways (51). EFNA3 activates the epithelial-mesenchymal transition (EMT) by the PI3K/AKT signaling pathway, which facilitates tumor growth, invasion, metastasis, as well as drug resistance (52). The above studies have demonstrated EFNA family could promote cancer development via the PI3K-Akt signaling pathway. Another significant finding from our research is that EFNA family members are closely associated with the infiltration of various immune cell types in HCC. This suggests that the EFNA family may also influence the immune microenvironment in HCC. In particular, EFNA1/3/4 was negatively related to most types of immune cell infiltration, while EFNA5 was positively associated with most immune cell infiltration. These findings indicated that EFNA proteins may influence the abundance of most immune cells in HCC tissues, thus playing a crucial role in HCC progression.

Although this study systematically revealed the expression patterns, prognostic value, and association with the tumor immune microenvironment of the EFNA family in HCC through the integration of multi-omics data, certain limitations must be acknowledged. First, our conclusions rely primarily on bioinformatics analysis of public databases, lacking validation from independent experimental cohorts, which may affect generalizability to specific clinical subgroups. Second, while revealing strong statistical associations, this approach cannot directly elucidate the precise molecular mechanisms by which specific EFNAs (e.g., upregulated EFNA1/3/4 or downregulated EFNA5) drive HCC progression via pathways like PI3K-Akt or ferroptosis. Finally, cancer development involves complex regulatory networks including non-coding RNAs and epigenetic modifications (53,54). Our analysis has not deeply explored the cross-talk between EFNAs and these layers, such as HCC-related lncRNAs (55) or nucleic acid-targeting strategies (56). Future work should prioritize: (I) mechanistic interrogation of EFNA1/3/4 and EFNA5 in HCC cell-state transitions (migration, EMT, angiogenesis) and tumor-immune communication; (II) multi-omics and single-cell/spatial analyses to resolve cell-type-specific EFNA signaling; and (III) prospective studies to evaluate EFNAs as prognostic biomarkers and as potential markers for immune stratification. Similar to other proposed molecular biomarkers, the translational value of EFNA candidates will ultimately depend on rigorous experimental validation and clinically meaningful performance in independent cohorts (55). More broadly, advances in immunotherapeutic and nucleic-acid-based strategies for liver disease further motivate continued exploration of molecular determinants that link etiology, tumor biology, and immune response (56). In terms of sustainability, repurposing drugs as GLP-1 based therapy or developing new imidazo[2,1-b] quinazoline derivative or thymidylate synthase conformer-selective inhibitors as prophylactic strategies may have a positive impact on cancer prevention and management (57-59).

Conclusions

In conclusion, our comprehensive investigation of the expression profiles and prognostic values of EFNA family member genes in HCC revealed that EFNA1, EFNA3, and EFNA4 act as tumor promoters, while EFNA5 is a poor prognostic predictor. Thus, EFNA1, EFNA3, EFNA4, and EFNA5 are promising candidates for prognostic biomarkers and therapeutic targets in HCC.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-aw-2532/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-aw-2532/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-aw-2532/prf

Funding: This study was supported in part by the General Project of Fujian Natural Science Foundation (Nos. 2023J01651 and 2025J01513), the Young and Middle-aged Talents Training Project of Fujian Provincial Heath commission (No. 2023GGA018), Startup Fund for Scientific Research from Fujian Medical University (No. 2020QH1154), and the Key Clinical Specialty Discipline Construction Program of Fujian, China (grant No. Min Wei Ke Jiao 2012 No.149).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-aw-2532/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Fujian Medical University Union Hospital (2020WSJK031) and informed consent was obtained from all individual participants or their legally appointed representatives participating in our study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Rumgay H, Ferlay J, de Martel C, et al. Global, regional and national burden of primary liver cancer by subtype. Eur J Cancer 2022;161:108-18. [Crossref] [PubMed]
3. Youness RA, Hassan HA, Abaza T, et al. A Comprehensive Insight and In Silico Analysis of CircRNAs in Hepatocellular Carcinoma: A Step toward ncRNA-Based Precision Medicine. Cells 2024;13:1245. [Crossref] [PubMed]
4. Hamdy NM, Sallam AM, Elazazy O, et al. LincRNA-miR interactions in hepatocellular carcinoma: comprehensive review and in silico analysis: a step toward ncRNA precision. Naunyn Schmiedebergs Arch Pharmacol 2025;398:14785-812. [Crossref] [PubMed]
5. Forner A, Reig M, Bruix J. Hepatocellular carcinoma. Lancet 2018;391:1301-14. [Crossref] [PubMed]
6. Yang X, Yang C, Zhang S, et al. Precision treatment in advanced hepatocellular carcinoma. Cancer Cell 2024;42:180-97. [Crossref] [PubMed]
7. Cheong TC, Jang A, Wang Q, et al. Mechanistic patterns and clinical implications of oncogenic tyrosine kinase fusions in human cancers. Nat Commun 2024;15:5110. [Crossref] [PubMed]
8. Pasquale EB. Eph receptors and ephrins in cancer progression. Nat Rev Cancer 2024;24:5-27. [Crossref] [PubMed]
9. Guo X, Yang Y, Tang J, et al. Ephs in cancer progression: complexity and context-dependent nature in signaling, angiogenesis and immunity. Cell Commun Signal 2024;22:299. [Crossref] [PubMed]
10. Yang Y, Ding T, Cong Y, et al. Interferon-induced transmembrane protein-1 competitively blocks Ephrin receptor A2-mediated Epstein-Barr virus entry into epithelial cells. Nat Microbiol 2024;9:1256-70. [Crossref] [PubMed]
11. Dasari SK, Joseph R, Umamaheswaran S, et al. Combination of EphA2- and Wee1-Targeted Therapies in Endometrial Cancer. Int J Mol Sci 2023;24:3915. [Crossref] [PubMed]
12. Zhao Y, Cai C, Zhang M, et al. Ephrin-A2 promotes prostate cancer metastasis by enhancing angiogenesis and promoting EMT. J Cancer Res Clin Oncol 2021;147:2013-23. [Crossref] [PubMed]
13. Offenhäuser C, Dave KA, Beckett KJ, et al. EphA2 regulates vascular permeability and prostate cancer metastasis via modulation of cell junction protein phosphorylation. Oncogene 2025;44:208-27. [Crossref] [PubMed]
14. Ren Y, Ju Q, Zhang J, et al. MiR-302a-3p reduces cisplatin resistance of esophageal squamous cell carcinoma cells by targeting EphA2. J Chemother 2024;36:72-81. [Crossref] [PubMed]
15. Sakuraba S, Koizumi A, Iwasawa T, et al. Serum EphA2 as a Promising Biomarker for the Early Detection and Diagnosis of Colorectal Cancer. Biomolecules 2024;14:1504. [Crossref] [PubMed]
16. Zhong X, Zhu Z, Du Y, et al. EFNA4-enhanced deubiquitination of SLC7A11 inhibits ferroptosis in hepatocellular carcinoma. Apoptosis 2025;30:349-63. [Crossref] [PubMed]
17. Psilopatis I, Souferi-Chronopoulou E, Vrettou K, et al. EPH/Ephrin-Targeting Treatment in Breast Cancer: A New Chapter in Breast Cancer Therapy. Int J Mol Sci 2022;23:15275. [Crossref] [PubMed]
18. Psilopatis I, Karniadakis I, Danos KS, et al. May EPH/Ephrin Targeting Revolutionize Lung Cancer Treatment? Int J Mol Sci 2022;24:93. [Crossref] [PubMed]
19. Hao Y, Li G. Role of EFNA1 in tumorigenesis and prospects for cancer therapy. Biomed Pharmacother 2020;130:110567. [Crossref] [PubMed]
20. Iida H, Honda M, Kawai HF, et al. Ephrin-A1 expression contributes to the malignant characteristics of {alpha}-fetoprotein producing hepatocellular carcinoma. Gut 2005;54:843-51. [Crossref] [PubMed]
21. Wada H, Yamamoto H, Kim C, et al. Association between ephrin-A1 mRNA expression and poor prognosis after hepatectomy to treat hepatocellular carcinoma. Int J Oncol 2014;45:1051-8. [Crossref] [PubMed]
22. Feng YX, Zhao JS, Li JJ, et al. Liver cancer: EphrinA2 promotes tumorigenicity through Rac1/Akt/NF-kappaB signaling pathway. Hepatology 2010;51:535-44. [Crossref] [PubMed]
23. Jukonen J, Moyano-Galceran L, Höpfner K, et al. Aggressive and recurrent ovarian cancers upregulate ephrinA5, a non-canonical effector of EphA2 signaling duality. Sci Rep 2021;11:8856. [Crossref] [PubMed]
24. Sealfon SC, Chu TT. RNA and DNA microarrays. Methods Mol Biol 2011;671:3-34. [Crossref] [PubMed]
25. Tang Z, Li C, Kang B, et al. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. Nucleic Acids Res 2017;45:W98-W102. [Crossref] [PubMed]
26. Chandrashekar DS, Bashel B, Balasubramanya SAH, et al. UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses. Neoplasia 2017;19:649-58. [Crossref] [PubMed]
27. Vivian J, Rao AA, Nothaft FA, et al. Toil enables reproducible, open source, big biomedical data analyses. Nat Biotechnol 2017;35:314-6. [Crossref] [PubMed]
28. Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics 2007;23:1846-7. [Crossref] [PubMed]
29. Asplund A, Edqvist PH, Schwenk JM, et al. Antibodies for profiling the human proteome-The Human Protein Atlas as a resource for cancer research. Proteomics 2012;12:2067-77. [Crossref] [PubMed]
30. Nagy Á, Lánczky A, Menyhárt O, et al. Validation of miRNA prognostic power in hepatocellular carcinoma using expression data of independent datasets. Sci Rep 2018;8:9227. [Crossref] [PubMed]
31. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012;2:401-4. [Crossref] [PubMed]
32. Zhou G, Soufan O, Ewald J, et al. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res 2019;47:W234-41. [Crossref] [PubMed]
33. Zhou Y, Zhou B, Pache L, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 2019;10:1523. [Crossref] [PubMed]
34. Li T, Fan J, Wang B, et al. TIMER: A Web Server for Comprehensive Analysis of Tumor-Infiltrating Immune Cells. Cancer Res 2017;77:e108-10. [Crossref] [PubMed]
35. Barretina J, Caponigro G, Stransky N, et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 2012;483:603-7. [Crossref] [PubMed]
36. Thylur RP, Roy SK, Shrivastava A, et al. Assessment of risk factors, and racial and ethnic differences in hepatocellular carcinoma. JGH Open 2020;4:351-9. [Crossref] [PubMed]
37. Bartley TD, Hunt RW, Welcher AA, et al. B61 is a ligand for the ECK receptor protein-tyrosine kinase. Nature 1994;368:558-60. [Crossref] [PubMed]
38. Chu LY, Huang BL, Huang XC, et al. EFNA1 in gastrointestinal cancer: Expression, regulation and clinical significance. World J Gastrointest Oncol 2022;14:973-88. [Crossref] [PubMed]
39. Cui Y, Chang Y, Ma X, et al. Ephrin A1 Stimulates CCL2 Secretion to Facilitate Premetastatic Niche Formation and Promote Gastric Cancer Liver Metastasis. Cancer Res 2025;85:263-76. [Crossref] [PubMed]
40. Gómez-Maldonado L, Tiana M, Roche O, et al. EFNA3 long noncoding RNAs induced by hypoxia promote metastatic dissemination. Oncogene 2015;34:2609-20. [Crossref] [PubMed]
41. Barderas R, Mendes M, Torres S, et al. In-depth characterization of the secretome of colorectal cancer metastatic cells identifies key proteins in cell adhesion, migration, and invasion. Mol Cell Proteomics 2013;12:1602-20. [Crossref] [PubMed]
42. Abdou AG, Abd el-Wahed MM, Asaad NY, et al. Ephrin A4 expression in osteosarcoma, impact on prognosis, and patient outcome. Indian J Cancer 2010;47:46-52. [Crossref] [PubMed]
43. Chen YL, Yen YC, Jang CW, et al. Ephrin A4-ephrin receptor A10 signaling promotes cell migration and spheroid formation by upregulating NANOG expression in oral squamous cell carcinoma cells. Sci Rep 2021;11:644. [Crossref] [PubMed]
44. Li JJ, Liu DP, Liu GT, et al. EphrinA5 acts as a tumor suppressor in glioma by negative regulation of epidermal growth factor receptor. Oncogene 2009;28:1759-68. [Crossref] [PubMed]
45. Li S, Hou X, Wu C, et al. MiR-645 promotes invasiveness, metastasis and tumor growth in colorectal cancer by targeting EFNA5. Biomed Pharmacother 2020;125:109889. [Crossref] [PubMed]
46. Kuang SQ, Bai H, Fang ZH, et al. Aberrant DNA methylation and epigenetic inactivation of Eph receptor tyrosine kinases and ephrin ligands in acute lymphoblastic leukemia. Blood 2010;115:2412-9. [Crossref] [PubMed]
47. Pasquale EB. Eph receptors and ephrins in cancer: bidirectional signalling and beyond. Nat Rev Cancer 2010;10:165-80. [Crossref] [PubMed]
48. Bao M, Zhang L, Hu Y. Novel gene signatures for prognosis prediction in ovarian cancer. J Cell Mol Med 2020;24:9972-84. [Crossref] [PubMed]
49. Cao D, Hustinx SR, Sui G, et al. Identification of novel highly expressed genes in pancreatic ductal adenocarcinomas through a bioinformatics analysis of expressed sequence tags. Cancer Biol Ther 2004;3:1081-9; discussion 1090-1. [Crossref] [PubMed]
50. Xie J, Xing S, Shen BY, et al. PIWIL1 interacting RNA piR-017061 inhibits pancreatic cancer growth via regulating EFNA5. Hum Cell 2021;34:550-63. [Crossref] [PubMed]
51. Lim W, Bae H, Bazer FW, et al. Functional Roles of Eph A-Ephrin A1 System in Endometrial Luminal Epithelial Cells During Early Pregnancy. J Cell Physiol 2017;232:1527-38. [Crossref] [PubMed]
52. Wang L, Song Y, Wang H, et al. MiR-210-3p-EphrinA3-PI3K/AKT axis regulates the progression of oral cancer. J Cell Mol Med 2020;24:4011-22. [Crossref] [PubMed]
53. Karimkhanilouyi S, Ghorbian S. Nucleic acid vaccines for hepatitis B and C virus. Infect Genet Evol 2019;75:103968. [Crossref] [PubMed]
54. Abam F, Ghorbian S. The dual role of LncRNAs in hepatocellular carcinoma: Friend and foe. Gastroenterology & Endoscopy 2024;2:186-95. [Crossref]
55. Moghimi A, Bani Hosseinian N, Mahdipour M, et al. Deciphering the Molecular Complexity of Hepatocellular Carcinoma: Unveiling Novel Biomarkers and Therapeutic Targets Through Advanced Bioinformatics Analysis. Cancer Rep (Hoboken) 2024;7:e2152. [Crossref] [PubMed]
56. Aalijahan H, Ghorbian S. Clinical Application of Long Non-Coding RNA-UCA1 as a Candidate Gene in Progression of Esophageal Cancer. Pathol Oncol Res 2020;26:1441-6. [Crossref] [PubMed]
57. Mostafa AM, Hamdy NM, Abdel-Rahman SZ, et al. Effect of vildagliptin and pravastatin combination on cholesterol efflux in adipocytes. IUBMB Life 2016;68:535-43. [Crossref] [PubMed]
58. Khodair AI, El-Hallouty SM, Cagle-White B, et al. Camptothecin structure simplification elaborated new imidazo[2,1-b]quinazoline derivative as a human topoisomerase I inhibitor with efficacy against bone cancer cells and colon adenocarcinoma. Eur J Med Chem 2024;265:116049. [Crossref] [PubMed]
59. El-Mesallamy HO, El Magdoub HM, Chapman JM, et al. Biomolecular study of human thymidylate synthase conformer-selective inhibitors: New chemotherapeutic approach. PLoS One 2018;13:e0193810. [Crossref] [PubMed]