Expression Profiles Analysis Identification and Interaction of Key Genes and Micrornas in Hepatocellular Carcinoma

Background: Hepatocellular carcinoma is one of the most common malignant tumors, with rapid development and high malignancy. MicroRNAs have been reported to play important roles in hepatocellular carcinoma progression. Aim: To identify the key genes and miRNAs in HCC, and to explore their potential molecular mechanisms. Methods: Gene expression profiles of GSE15471 (mRNA profile) and GSE57555 (miRNA profile) were downloaded from gene expression omnibus, which were analysed using R software and bioconductor packages. The gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichments of DEGs were performed using the DAVID database, and the protein–protein interaction networks of the DEGs were constructed from the STRING database. In addition, targets of differentially expressed miRNAs were predicted by the online resource miRDB. Result: In total, 191 differentially expressed genes were identified, including 142 upregulated and 49 downregulated genes. Functional analysis revealed that these DEGs were associates with wound healing, endodermal cell-cell adhesion, activation of MAPK activity and negative regulation of cell proliferation. In addition, we identified five DEMs, which were upregulated and downregulated. hsa-miR-122-5p may target the PDK4, and hsamiR-21-5p probably targets SPOCK1 and PAIP2B. Conclusions: We applied integrated bioinformatics to identify key pathogenic genes involved in hepatocellular carcinoma and provide new clues for further studies of hepatocellular carcinoma.


Introduction
Hepatocellular carcinoma (HCC) is one of the most common malignant tumors worldwide [1] with more than 0.55 million liver cancer cases annually [2,3].The pathogenesis of HCC is a multistep process involving the multi-Step molecular pathogenesis [4] and cellular events of cancer metabolism and tumor microenvironment.[5,6].Although the great advances of therapeutic approaches targeted at liver cancer, the recurrence and metastasis remain the major factors affecting the prognosis of patients with HCC [7].It is urgent to investigate the molecular mechanisms involved in HCC initiation and progression to develop new treatment approaches.Therefore, bioinformatics analysis of microarray data has been widely used to developing biomarkers in malignancy such as HCC.
In recent years, a large number of studies have been conducted on differentially expressed genes (DEGs) in HCC, and their role in development and progression of HCC have been reported.[8,9].And previous studies revealed that miRNAs Trends Journal of Sciences Research played a significant role in biological processes, molecular functions and different pathways of HCC [10,11], However, it remains unclear about the potential molecular pathway interactions between DEGs and miRNA in the pathogenesis of HCC.We analysed data generated by microarray technology to provides reliable molecular markers between DEGs and DEMs, as well as effective drug targets for treating HCC.

Data source
The gene expression profiles between HCC and normal controls downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/),GSE15471 and GSE57555.The GSE15471, which contains an mRNA expression profile of 39 HCC samples and 39 normal samples.The GSE57555 consist of an miRNA expression profile of 32 tumor samples and 32 normal samples.

Screening for DEGs and DEMs
R software was used to convert and discard the meaningless data.GEO2R was used to screen for miRNAs (differentially expressed miRNAs; DEMs) that were differentially expressed between tumor samples and normal samples.The adjusted P < 0.05 and logFC ≥ 2 for DEGs and P < 0.05 and logFC ≥ 1 for DEMs as a screening criteria.In this process, we needed the R package of limma in the Bioconductor package for identifying differentially expressed genes.

GO and KEGG analyses of DEGs
DAVID Database (https://david.ncifcrf.gov) is a bioinformatics database containing functional interpretation of lists of genes, where the analyses of GO and KEGG can be found.GO is a frequently-used bioinformatics tool because of providing comprehensive information about gene function of individual genomic products through ontology.Molecular functions (MF), biological processes (BP) and cellular components (CC) are three items of GO.KEGG is a set of high-throughput genes and protein pathways.The screening criteria of GO and KEGG analyses of DEGs identified in GSE15471 was P < 0.05, and the analyses were performed through DAVID online tools.

PPI network integration
The STRING database (http://string-db.org/)usually used to confirm the proteins and proteins interaction (PPI) of DEGs.Moreover, the Cytoscape software (http://www.cytoscape.org/) is a network for visualizing PPI network of common DEGs in HCC.Cytoscape Molecular Complex Detection (MCODE) plug-in was used to screen clustered sub-networks.

Prediction of microRNA targets
MiRDB (http://mirdb.org/miRDB/), a database of microRNA targets, was used to predicted the target genes for DEMs in GSE57555.The common genes were further screen by comparing the target genes with DEGs.

Microarray data information and identification of DEGs in hepatocellular carcinoma
The results of the hepatocellular cancer expression microarray datasets GSE15471 was standardized as is shown in Figure 1A and Figure 1B.When the GSE15471 dataset was screened and obtained 191 DEGs by the limma package (corrected P-value<0.05,logFC >2), where 142 upregulated genes and 49 downregulated genes were identified (Table1).The heatmap of the top 100 DEGs are shown in Figure 2. Trends Journal of Sciences Research Table 1.Screening DEGs in HCC by integrated microarray

Gene ontology and pathway enrichment analyses
GO and KEGG analyses to explore the function and pathways of the 191 identified DEGs were performed by using the DAVID.Figure 3A is the analysis of GO term enrichment, where we could see the biological processes category, the molecular function and the cell composition of upregulated genes and downregulated.As is shown in Figure 3B, in the biological process group, the upregulated genes were mainly enriched in extracellular matrix organization, cell adhesion, and proteolysis.The downregulated genes were mainly concentrated in heparan sulfate proteoglycan metabolic process.In the cell composition group, the upregulated genes were mainly enriched in extracellular space, extracellular exosomes, and extracellular matrix.The downregulated genes were mainly enriched in transport vesicle membrane, apical part of cell, and the collagen type V trimer.In the molecular function group, the upregulated genes were mainly enriched in calcium ion binding, serine-type endopeptidase activity, and extracellular matrix structural constituent.The downregulated genes were mainly enriched in extracellular matrix constituent conferring elasticity, oxidoreductase activity, acting on the CH-NH2 group of donors, oxygen as acceptor, N-acetylglucosamine-6-sulfatase activity, and cysteine-type endopeptidase inhibitor activity.These results indicate that most DEGs are significantly enriched in extracellular substance, cell cycle regulation, and regulating endopeptidase activity.
The pathway analysis of KEGG showed that the pathway of extracellular matrix receptor interactions and focal adhesion had the most upregulated genes in seven pathways, meanwhile downregulated genes mainly took part in pancreatic secretion (Table 2).

MicroRNA-differentially expressed gene pairs
After analysing the GSE57555 dataset, Five DEMs that were differentially regulated between the HCC and the normal tissue samples were identified.Of these, three miRNAs were downregulated and two miRNAs were upregulated.The miRDB database was used to predict target genes of the identified DEMs (Table 3).We screened genes with an opposing expression trend depend on the target genes with DEGs from GSE15471, this was all for further analysis.In the downregulated genes, hsa-miR-23a-3p probably targets VCAN, TNFAIP6, SPOCK1, SLC6A14, PLAU, MFAP5, FBN1, CRISPLD1 and COLHA1, while hsa-miR-200b-3p potentially targets ZFPM2, SULF1, GREM1, FN1 and CDH11, and hsa-miR-21-5p may target SPOCK1, CCL20, ASPN.At the same time, in the upregulated genes, hsa-miR-122-5p may target PDK4, but meanwhile has-miR-4286 may target ITGBL1, CYP1B1 and Col5A2.Table 3. Differentially expressed microRNAs in HCC and their potential target genes.

Discussion
The occurrence and development of HCC is a complex and multistep process that relates to a variety of cumulative genetic and epigenetic changes (12,13).In this work, 142 up-regulated genes and 49 down-regulated genes were screened out.The top 10 most significantly upregulated genes were COL10A1 INHBA CTHRC1 SULF1 COL1A2 VCAN THBS2 COL11A1 GJB2 S100P.The top 10 most significantly downregulated genes were CTRL SERPINI2 AQP8 ALB SYCN TMED6 GNMT CELA2B PNLIPRP2 ERP27.In addition, the upregulated genes were mainly involved in the cell adhesion, inflammatory response, extracellular exosome, heparin binding, wound healing and apoptotic process that the downregulated genes were mainly involved in the activation of MAPK activity.MAPK, a family of serine/threonine kinases, play a significant role in the tumor proliferation, differentiation, migration, apoptosis and inflammation (14,15).
There is now well-established in the literature that miRNAs play a vital role in regulating the translation and degradation of mRNAs (16,17).More and more research showed that miRNAs played an important role in cell death, cell proliferation, stress and resistance.In the present work, we identified 3 upregulated DEM and 2 downregulated DEM.We identified potential targets of these DEMs by comparing target genes with DEGs.
The upregulated DEM, hsa-miR-122-5p, potentially targets PDK4.PDK4 is a key mitochondrial enzyme of PDK family that is involved in regulating the switch to aerobic glycolysis by shunting pyruvate metabolism from the mitochondria to the cytoplasm for glycolysis (18,19), and it may play an important role between cellular respiration and cell cycle progression as a prime molecular suspect in crosstalk (20).It was reported that PDK4 in hepatocellular carcinoma cells could be suppressed by the miR-129-5p ( 21), but the role of hsa-miR-122-5p in the regulation of CCR7 in hepatocellular carcinoma pathogenesis still needs further study.
The downregulated DEM, hsa-miR-21-5p, potentially targets the SPOCK1 and SPAI2B.SPOCK1-mediated cell proliferation and migration influence the survival of several tumors, such as non-small cell lung cancer, gastric cancer and pancreatic cancer (22)(23)(24) PAIP2 is a translational repressor capable of inhibiting cap-dependent translation that has been implicated in multiple biological processes, including cell proliferation and differentiation (25,26).However, whether SPOCK1 and PAIP2 are involved in the HCC proliferation and migration is unclear.
In conclusion, our study has profiled DEGs using bioinformatics analyses and to find potential miRNAs to improve our understanding of the pathogenesis of HCC.In this work, 191 DEGs and 5 DEMs were screened.Given their interactions with PDK4 SPOCK1 and PAIP2, respectively, the role of hsa-miR-122-5p and hsa-miR-21-5p in HCC may provide effective targets for the treatment of HCC.However, further molecular biological experiments studies are required to confirm the function of the identified genes obtained from bioinformatics analysis.