lncRNA-AC079061.1/VIPR1 axis may suppress the development of hepatocellular carcinoma: a bioinformatics analysis and experimental validation

The whole workflow chart of the present study was depicted in Fig. 1. The RNA-seq in TPM (transcripts per million reads) format, miRNA-seq in RPM (reads per million mapped reads) format, and clinical data of hepatocellular carcinoma cases were downloaded from The Cancer Genome Atlas (TCGA) portal (http://​cancergenome.​nih.​gov). The whole cohort consisted of 374 HCC samples and 50 normal samples. The GTEX data were obtained from UCSC XENA (https://​xenabrowser.​net/​datapages/​) database. The gene annotation files were downloaded from the GENCODE database and used to extract the lncRNA and mRNA expression matrix, respectively. The immune gene list was obtained from the ImmPort Portal database (https://​www.​immport.​org/​home). The targets of lncRNA were predicted from the LncBase v.2 (http://​carolina.​imis.​athena-innovation.​gr/​diana_​tools/​web/​index.​php?​r=​lncbasev2%2Findex), and the targets of miRNAs were predicted from the miRDB (http://​mirdb.​org/​). The copy-number alteration data was downloaded from the cBioPortal database (http://​www.​cbioportal.​org/​), and the immunohistochemistry data were obtained from the Human Protein Atlas (HPA) database (https://​www.​proteinatlas.​org/​).

The RNA-seq were divided into high and low groups according to the EZH2 expression levels. The difference analysis was performed using the “DESeq2” [15]. For better and effective screening of interaction networks, lncRNAs with |logFC|> 2, adjusted p-value < 0.05, miRNAs with |logFC|> 0.5, adjusted p-value < 0.05, and mRNAs with |logFC|> 1, adjusted p-value < 0.05 were considered as differentially expressed genes. All the expression data were log2 transformed.

Since the expression data of RNA-seq and miRNA-seq did not conform to a normal distribution, we utilized the Spearman correlation to analyze the association between lncRNAs, miRNAs, and mRNAs. The results of the Spearman correlation were visualized using the “ggplot2” R package.

To assess the potential mechanisms of the lncRNA-miRNA-mRNA axis, the KEGG and GO enrichment analyses were performed by using the “clusterProfiler” R package [16]. Biological process terms, cellular component terms, molecular function terms, and pathways with adjusted p-value < 0.05 were considered as potential mechanisms for the onset of HCC.

The markers of 24 immune cells were derived from an immunity article, and the classification and description of specific cells can be found in this article [17]. The immune infiltration analysis was performed through the ssGSEA algorithm in the “GSVA” R package [18]. The Pearson correlation analysis was used to evaluate the correlation between immune cells and differentially expressed genes.

The methylation of hub gene in HCC and normal tissues was evaluated through the MEXPRESS (https://​mexpress.​be/​) and MethSurv database (https://​biit.​cs.​ut.​ee/​methsurv/​). The search terms were VIPR1 and liver hepatocellular carcinoma.

According to the clinical data of HCC, the effects of hub genes on patients’ prognoses were assessed through the Kaplan–Meier curve. The univariate Cox regression analysis was performed to screen for factors to construct the nomogram model. Nomograms were used to evaluate the role of hub genes in HCC through the “rms” and “survival” R packages. P-value < 0.05 was considered significantly.

Human liver cells L0-2, HCC cells (PLC/PRF/5, HepG2, Hep3B) (Cell Bank of the Chinese Academy of Sciences, Shanghai, China), MHCC97H and HCCLM3 (Liver Cancer Institute, Fudan University, Shanghai, China), and Huh7 (Japanese Cancer Research Resources Bank) were cultured in Dulbecco's modified Eagle medium (Gibco) with 10% fetal bovine serum (Gibco) and 1% penicillin–streptomycin (Invitrogen). Cell cultures were done in a thermostatic incubator at 37 °C with a humidified atmosphere of 95% air and 5% CO₂.

The lncRNA-AC079061.1 small interfering RNA (siRNA) and negative control were purchased from RiboBio Company (Guangzhou, China). The transfection of the lncRNA-AC079061.1 siRNA and negative control was performed with Lipofectamine™ 3000 (#L3000008, ThermoFisher, China) according to the manufacturer’s instructions.

The total RNA including miRNA was extracted from cells or tissues using the TRIzol Reagent (Invitrogen), while cDNA was synthesized from RNA using the Reverse Transcription Kit (Takara). Subsequently, cDNA was amplified using the Maxinma SYBR Green qPCR Master Mix (Thermo Scientific). The quantification of target genes was done with the 2^−ΔΔCt method using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) for normalization. The primers for hsa-miR-765 and U6 small nuclear RNA were obtained from RiboBio Company (Guangzhou, China). The sequences were covered by a patent. Analyses of miRNA expression were normalized to the expression of internal control U6 using the 2 − ΔΔCT method. Melting curve analysis was carried out to assess the specificity of PCR products. The primers have been used for real time PCR, lncRNA-AC079061.1 Sequence 5′-3′: Forward, CCCTCAGGCATCCACTCTACCC; Reverse, TCCAACGCCACCCACTTCAAAC; VIPR1 Sequence 5′-3′: Forward, ACAAGGCAGCGAGTTTGGATGAG; Reverse, GTGCAGTGGAGCTTCCTGAACAG.

MHCC97H cells and HCCLM3 cells were seeded into 96-well plates and co-transfected with a mixture of 60 ng luciferase, 6 ng pRL-CMV Renilla luciferase reporter, and miR-765 mimic or the negative control using the Lipofectamine 2000 transfection reagent (RiboBio, Guangzhou, China), according to the manufacturer’s instructions.). After 48 h of incubation, the firefly and Renilla luciferase activities were measured using a dual-luciferase reporter assay (Promega, Madison, WI, USA).

As described, the protein was extracted from cells or tissues using RIPA cell lysis with Protease Inhibitor Cocktail. Mitochondrial protein extraction was performed with the use of a mitochondrial isolation kit (Beyotime Biotechnology), in accordance with the manufacturer’s directions. The proteins were quantified by using the BCA kit, subjected to 12% SDS-PAGE for separation, then transferred to 0.45 μM PVDF membranes (Millipore, USA). The membranes were blocked with skimmed milk and incubated with primary antibodies at 4 ℃ overnight, followed by incubation with the corresponding HRP-conjugated secondary antibody (PeproTech). The bands were visualized by enhanced chemiluminescence. The intensity of protein expression was measured using ImageJ software. The antibodies used for western blot were listed as follows: VIPR1, Bioss, # bs-2982R, 1:1000; GAPDH, Cell Signaling Technology, #5174, 1:1000.

The cell proliferation assays were performed using a CCK-8 Kit (Yeasen, Shanghai, China) and colony formation. Three thousand cells were seeded into each well in a 96-well plate. The CCK-8 solution (10 µl) was added to 100 µl of culture media, and the optical density was measured at 450 nm. For colony formation assay, one thousand cells were seeded into each well in a 6-well plate for one week, then washed twice with PBS, fixed with 4% paraformaldehyde, stained with crystal violet, and the numbers of foci were counted for each well. Three independent experiments were performed.

For migration assay, 5 × 10⁴ cells were suspended in 200 µl of DMEM without serum and placed in the cell culture insert (8 µm pore size; BD Falcon, San Jose, CA) of a companion plate (BD Falcon) with a prewarmed culture medium containing 10% fetal bovine serum. For invasion assay, 1 × 10⁵ cells suspended in serum-free medium were seeded into the upper chamber coated with 1 µg/µl Matrigel (BD Biosciences, USA) of a 24-well transwell plate (8-μm pore size, Corning, NY, USA), then 600 μl DMEM with 10% FBS was added into the lower chamber. After incubation for an indicated period of time at 37 °C in 5% CO₂, the migrating and invading cells on the outside of the upper chamber membrane were then fixed with 4% paraformaldehyde, stained with crystal violet, and counted under a light microscope (100× magnification) in eight randomly selected areas. Three independent experiments were performed.

Previous studies have reported that EZH2 plays a crucial role in HCC. We hereby comprehensively discussed the important function of EZH2 in HCC through bioinformatics analysis [19‐21]. The expression of EZH2 is frequently upregulated in HCC tissues compared to normal tissues (Fig. 2A). The related survival analysis of EZH2 showed that patients with high EZH2 expression have a worse overall survival (OS: HR = 1.99; 95% CI: 1.40–2.84; p < 0.001), disease-specific survival (DFS: HR = 2.44; 95% CI: 1.53–3.89; p < 0.001), and progression-free interval (PFI: HR = 1.81; 95% CI: 1.35–2.42; p < 0.001) (Fig. 2B). The association between EZH2 mRNA expression and copy number is shown in Fig. 2C and D), which suggested that the mRNA expression level of EZH2 is positively related to the copy number of EZH2 in HCC (R = 0.15, p = 4.227e−3). The protein expression of EZH2 in HCC and normal tissues were further evaluated through the HPA database. The immunohistochemistry of EZH2 showed that EZH2 is mainly located in the nuclear and highly expressed in tumor tissues compared with normal tissues (Fig. 2E). These results identified EZH2 as a prognostic factor, which promoted the progression of HCC. Future studies based on EZH2 may help explore more novel biomarkers and treatment strategies in HCC.

Based on the expression of EZH2, the whole HCC cohort was divided into EZH2^high and EZH2^Low groups using median partitioning. The difference analysis was performed to obtain the differentially expressed lncRNAs (DElncRNAs). lncRNAs with |logFC|> 2 and p-value < 0.05 was considered as DElncRNAs. Then, 163 upregulated and 31 downregulated lncRNAs were visualized through the volcano map (Fig. 3A). The top 10 significant downregulated lncRNAs in order of p-value were selected for gene co-expression analysis, and the result showed that there is a strong negative correlation between 9 lncRNAs (LINC00570, AC079061.1, HTR2A-AS1, LDLRAD4-AS1, AC022601.1, PGM5-AS1, FAM83A-AS1, AL391261.2, and TMEM252-DT) and EZH2 (Fig. 3B). We assessed the expression level of 10 lncRNAs between HCC tissues and normal tissues, the result suggested that 5 lncRNAs (AC079061.1, HTR2A-AS1, LDLRAD4-AS1, FAM83A-AS1, and AL391261.2) were highly expressed in normal tissues, while 1 lncRNAs (AC005165.1) was highly expressed in tumor tissues (Fig. 3C). Then, ROC analyses were performed to assess the diagnostic value of these 10 lncRNAs. The results showed that only FAM83A-AS1, HTR2A-AS1, AC079061.1, and LDLRAD4-AS1 have a better diagnostic value in HCC, the area under curve (AUC) of these four lncRNAs were 0.709, 0.853, 0.804, and 0.712, respectively (Fig. 3D). Furthermore, survival analyses were used to evaluate the impact of lncRNAs on the outcome of HCC. Finally, the results suggested that AC079061.1 is the only lncRNA that can affect the prognosis of HCC. Patients with high AC079061.1 expression have a better outcome compared with patients with low AC079061.1 expression (HR = 0.63; 95% CI: 0.44–0.89; p = 0.01) (Fig. 3E).

Based on the ceRNA mechanism, potential miRNAs that may bind to DElncRNAs were first predicted using the LncBase v.2 database (http://​carolina.​imis.​athena-innovation.​gr/​diana_​tools/​web/​index.​php?​r=​lncbasev2%2Findex-predicted). Meanwhile, the differentially expressed miRNAs between HCC tissues and normal tissues were obtained by the “limma” R packages. Finally, 8 miRNAs that may bind to DElncRNAs were identified in HCC by intersecting the LncBase predicted miRNAs with the DEmiRNAs in TCGA (Fig. 4A and B). The expression levels of these 8 miRNAs were confirmed in Fig. 4C, which showed that 2 miRNAs (hsa-miR-4725-3p and hsa-miR-6514-3p) were downregulated in tumor tissues, and 6 miRNAs (hsa-miR-939-5p, hsa-miR-3690, hsa-miR-765, hsa-miR-362-3p, hsa-miR-766-5p, and hsa-miR-4742-3p) were elevated in tumor tissues. Next, ROC analyses were performed to evaluate the diagnostic value of DEmiRNAs. The results suggested that hsa-miR-362-3p (AUC = 0.755), hsa-miR-765 (AUC = 0.720), hsa-miR-939-5p (AUC = 0.713), and hsa-miR-4742-3p (AUC = 0.717) have a better diagnostic value than hsa-miR-766-5p (AUC = 0.596), hsa-miR-3690 (AUC = 0.604), hsa-miR-4725-3p (AUC = 0.588), and hsa-miR-6514-3p (AUC = 0.567) (Fig. 4D). Furthermore, survival analyses were utilized to assess the prognostic value of DEmiRNAs. Results showed that only 2 miRNAs (hsa-miR-3690 (HR = 1.47, p = 0.03) and hsa-miR-765 (HR = 1.43, p = 0.046) affected the prognosis of HCC (Fig. 4E). Collectively, AC079061.1 may regulate the progression of HCC by binding to hsa-miR-765.

In recent years, tumor immune microenvironment has been the focus of research and a hot topic. Many studies have suggested an inseparable relationship between tumorigenesis and the level of immune infiltration [22‐25]. Hence, we report 6 immune-related genes that may bind to hsa-miR-765 by intersecting the miRDB-predicted miRNAs and immune-related gene list with DEmRNA (Fig. 5A and B). The expression level of differentially expressed immune-related genes (DEIRGs) was confirmed in Fig. 5C. The ROC and survival analysis results of 6 DEIRGs suggested that only VIPR1 (AUC = 0.975, HR = 0.67, p = 0.023) has both diagnostic and prognostic values in HCC (Fig. 5D and E).

To further explore the association between lncRNA- AC079061.1, hsa-miR-765, and VIPR1, we first analyzed the subcellular localization of the lncRNA- AC079061.1. The sequence of lncRNA-AC079061.1 was obtained from the LNCipedia database (https://​lncipedia.​org/​db/​transcript/​lnc-EBAG9-9:​1), and the sequence was imported into the lncLocator database to predict the subcellular location of lncRNA- AC079061.1. The results suggested that lncRNA- AC079061.1 was mainly located in the ribosome and cytosol (Fig. 6A). The Spearman correlation analysis was used to assess the correlation between lncRNA- AC079061.1, hsa-miR-765, and VIPR1 since the expression of these genes did not conform to normal distribution. Then, results showed a negative correlation between hsa-miR-765 and lncRNA- AC079061.1 or VIPR1. A positive correlation was also noted between lncRNA- AC079061.1 and VIPR1 (Fig. 6B–D).

Previous studies confirmed that immune cells were involved in the progression of HCC [26‐28]. Thus, we explored the correlation between immune infiltration and these key genes. Initial results suggested that lncRNA-AC079061.1 was negatively correlated with NK CD56 bright cells, TFH, Th2 cells, macrophages, B cells, and Th1 cells, while positively correlated with neutrophils, Tcm, CD8 T cells (Fig. 7A). Subseqent results showed that hsa-miR-765 was negatively correlated with DC cells, cytotoxic cells, neutrophils, CD8 T cells, NK cells, NK CD56dim cells, and Treg, while positively correlated with NK CD56bright cells (Fig. 7B). VIPR1, an immune-related gene, was positively correlated with many immune cells such as NK cells, neutrophils, mast cells, CD8 T cells, and eosinophils, while negatively correlated with Th2 cells and pDC cells (Fig. 7C). Furthermore, we also analyzed the correlation between VIP or ADCYAP1 (VIPR1’s ligand) and the infiltration level of immune cells. Results showed that the VIP and ADCYAP1 were similar to VIPR1, which displayed a positive correlation with the infiltration level of NK cells, iDC, and TEM (Additional file 1: Fig. S1A–D), suggesting the possibility that altered expression of ligands and receptors may lead to the change in immune cell infiltration level and has a dramatic impact on HCC progression. Interestingly, we found that lncRNA AC079061.1, hsa-miR-765, and VIPR1 were all related to neutrophils and CD8 T cells. The correlation analysis showed that there is a positive relationship between lncRNA-AC079061.1/VIPR1 and CD8 T cells/Neutrophils (Fig. 7D and E). These results implied that the lncRNA-AC079061.1/hsa-miR-765/VIPR1 axis may influence the progression of HCC by regulating the immune infiltration level of neutrophils and CD8 T cells.

The potential mechanism of VIPR1 remains unclear and was therefore uploaded to the String database. Then, the interaction network between VIPR1 and related genes was obtained (Fig. 8A). The KEGG and GO functional enrichment analyses were performed to explore the potential VIPR1-related mechanism. The results showed that VIPR1 and its related genes were highly enriched in neuroactive ligand-receptor interaction (adjusted p-value = 1.38e−04), cAMP signaling pathway(adjusted p-value = 0.004), and vascular smooth muscle contraction (adjusted p-value = 0.011) (Fig. 8B). Subsequently, the most highly enriched GO BP terms were cyclic-nucleotide-mediated signaling (adjusted p-value = 2.38e−09), G protein-coupled receptor signaling pathway, coupled to cyclic nucleotide second messenger (adjusted p-value = 5.42e−09), and adenylate cyclase-activating G protein-coupled receptor signaling pathway (adjusted p-value = 1.05e−08). The GO CC terms showed that these genes may be located in the perikaryon (adjusted p-value = 0.068), endocytic vesicle membrane (adjusted p-value = 0.068), and Mitochondria-associated ER Membrane (adjusted p-value = 0.068). Furthermore, the highly enriched GO MF terms were peptide hormone binding (adjusted p-value = 4.89e−07), G protein-coupled peptide receptor activity (adjusted p-value = 9.55e−06), and peptide receptor activity (adjusted p-value = 9.55e−06) (Fig. 8C, Table 1).

To further explore the possible mechanisms of VIPR1, the top 100 VIPR1-related genes were downloaded from the GEPIA database (http://​gepia.​cancer-pku.​cn/​detail.​php?​gene=​VIPR1) for another functional enrichment analysis (Additional file 1: Table S1.). More physiological processes such as complement activation, lectin pathway (adjusted p-value = 2.33e−04), positive regulation of vasculature development (adjusted p-value = 2.33e−04), cytokine receptor binding (adjusted p-value = 1.99e−04), receptor-ligand activity (adjusted p-value = 0.027) were obtained, suggesting that VIPR1 and its related genes play a key role in multiple mechanisms including the immune process. In summary, the results of these functional enrichment analyses suggested that VIPR1, as an immune-related gene, is involved in a variety of physiological processes (such as cytokine binding, inflammation-stimulation, and stress response) by regulating relevant genes, indicating the tremendous research value of VIPR1 in HCC.

Methylation is known to be aberrantly expressed in various forms of tumors. Meanwhile, it is capable of affecting not only the genetic phenotypes of cells but also the life span of humankind. There is a great potential for methylation in tumor diagnosis and treatments [29‐31]. Therefore, multiple methods, as well as databases, were used to assess the correlation between VIPR1 and the progression of HCC. We first analyzed the correlation between VIPR1 expression and methyltransferases in HCC. As shown in Additional file 1: Fig. S2A, the results showed that there is a negative correlation between VIPR1 and DNMT1 (R = −0.150, p = 0.003) /DNMT3A (R = −0.250, p < 0.001). But no significant correlation was noted between VIPR1 and DNMT3B (R = −0.099, p = 0.057). Similarly, DiseaseMeth 2.0 database demonstrated that the methylation of VIPR1 was significantly lower in disease tissues compared with HCC normal tissues (p = 0.021, Additional file 1: Fig. S2B). Additionally, the CpG expression level in Additional file 1: Fig. S2C was obtained from the MethSurv platform, in which 8 methylation sites (cg27510539, cg12828331, cg23003500, cg10970409, cg04322483, cg22694480, cg27309542, cg23517013) in the DNA sequences of VIPR1 were found to be negatively associated with VIPR1 expression levels and 11 methylation sites (cg06783423, cg03791150, cg00328740, cg18559858, cg02061253, cg09475741, cg11484444, cg09996971, cg01919863, cg15066503, cg06290884) were found to be positively associated with VIPR1 expression level, analyzed through the MEXPRESS tool (Additional file 1: Fig. S2D). According to these results, altered methylation of VIPR1 may be an inducer of HCC.

To assess the potential prognostic value of lncRNA AC079061.1, hsa-miR-765, and VIPR1 in HCC, univariate and multivariate cox regression analyses were performed. The results showed that lncRNA AC079061.1, hsa-miR-765, and VIPR1 as independent factors can affect the prognosis of HCC (Fig. 9A–C; Table 2, 3, 4). Then, we constructed prediction models to evaluate the relationship between lncRNA AC079061.1/hsa-miR-765/VIPR1 and the overall survival of HCC. The models were visualized by nomogram (Fig. 9D, E and F). All the results implied that the lncRNA AC079061.1/hsa-miR-765/VIPR1 axis acts as a prognostic indicator that may suppress the development of HCC.

To validate the feasibility of our comprehensive analysis, the expressions of these three genes were detected in the HCC cell lines and normal liver cells. We found that lncRNA and mRNA were frequently downregulated in HCC cells (Figs. 10A and B), while hsa-miR-765 was upregulated significantly in HCC cells (Fig. 10C). Then, we transfected the lncRNA- AC079061.1 siRNA into HCC cells, and the qRT-PCR was performed to confirm the stable downregulation of lncRNA- AC079061.1 (Additional file 1: Fig. S3). Moreover, to evaluate the relationship between lncRNA- AC079061.1 and the malignant phenotype of the tumor, a series of cellular function experiments was performed. The CCK-8 and colony formation assays suggested that the knockdown of lncRNA-AC079061.1 promoted the proliferation of HCC cells (Fig. 10D and E). Then, the transwell assays showed that the invasion and migration ability of si-lncRNA-AC079061.1 HCC cells were enhanced (Fig. 10F). These results indicated that lncRNA- AC079061.1 functions as an antioncogenic lncRNA to inhibit the proliferation, migration, and invasion of HCC cells in vitro. Additionally, to further verify the lncRNA-AC079061.1/VIPR1 axis, we first detected the expression level of hsa-miR-765 and VIPR1 in si- lncRNA- AC079061.1 HCC cells. Results showed that hsa-miR-765 is elevated in si- lncRNA- AC079061.1 cells compared with control cells (Fig. 10G), while the mRNA (Fig. 10G) and protein expression (Fig. 10H) of VIPR1 was decreased as the downregulation of lncRNA- AC079061.1 or upregulation of miR-765. Then, we performed dual-luciferase reporter assays in HCC cells and found that the luciferase activity was significantly downregulated after increasing the hsa-miR-765 expression in MHCC97H and HCCLM3 cells transfected with lncRNA-AC079061.1 or VIPR1 wildtype vectors. These results suggested that lncRNA-AC079061.1 may regulate the expression of downstream gene VIPR1 through sponging hsa-miR-765 (Fig. 10I). Furthermore, we investigated the effects of miR-765 on lncRNA-AC079061.1-mediated antioncogenic processes in vitro. Upregulation of miR-765 inhibited si-lncRNA-AC079061.1 induced proliferation (Fig. 10J), migration, and invasion (Fig. 10K) in MHCC97H cells. Collectively, these results indicated that the lncRNA- AC079061.1/hsa-miR-765/VIPR1 axis may regulate the progression of HCC.

Clinically, chemotherapy remains one of the mainstream treatments for HCC. However, in recent years, HCC patients tend to develop tolerance to common chemotherapy drugs. Therefore, the exploration of novel drugs is currently an essential trend in research. By utilizing the GDSC and CTRP databases, we obtained the correlation between VIPR1 expression and drug sensitivity in pan-cancer. These results provided good reference for future research direction. The results showed that the key genes investigated in the present study did not only correlate with common drugs such as Gefitinib and Afatinib but also with some unreported drugs. These correlations may even help identify new therapeutic drugs for HCC (Additional file 1: Fig. S4).