Analysis of Differentially Expressed Genes in Endothelial Cells Following Tumor Cell Adhesion, and the Role of PRKAA2 and miR-124-3p

Abstract

Tumor cell adhesion to the endothelium is one pattern of tumor-endothelium interaction and a key step during tumor metastasis. Endothelium integrity is an important barrier to prevent tumor invasion and metastasis. Changes in endothelial cells (ECs) due to tumor cell adhesion provide important signaling mechanisms for the angiogenesis and metastasis of tumor cells. However, the changes happened in endothelial cells when tumor-endothelium interactions are still unclear. In this study, we used Affymetrix Gene Chip Human Transcriptome Array 2.0. and quantitative real-time PCR (qPCR) to clarify the detailed gene alteration in endothelial cells adhered by prostate tumor cells PC-3M. A total of 504 differentially expressed mRNAs and 444 lncRNAs were obtained through chip data analysis. Gene Ontology (GO) function analysis showed that differentially expressed genes (DEGs) mainly mediated gland development and DNA replication at the biological level; at the cell component level, they were mainly involved in the mitochondrial inner membrane; and at the molecular function level, DEGs were mainly enriched in ATPase activity and catalytic activity. Kyoto Encyclopedia of Genes and Genomes (KEGG) signal pathway analysis showed that the DEGs mainly regulated pathways in cancer, cell cycle, pyrimidine metabolism, and the mTOR signaling pathway. Then, we constructed a protein-protein interaction functional network and mRNA-lncRNA interaction network using Cytoscape v3.7.2. to identify core genes, mRNAs, and lncRNAs. The miRNAs targeted by the core mRNA PRKAA2 were predicted using databases (miRDB, RNA22, and Targetscan). The qPCR results showed that miR-124-3p, the predicted target miRNA of PRKAA2, was significantly downregulated in endothelial cells adhered by PC-3M. With a dual luciferase reporter assay, the binding of miR-124-3p with PRKAA2 3'UTR was confirmed. Additionally, by using the knockdown lentiviral vectors of miR-124-3p to downregulate the miR-124-3p expression level in endothelial cells, we found that the expression level of PRKAA2 increased accordingly. Taken together, the adhesion of tumor cells had a significant effect on mRNAs and lncRNAs in the endothelial cells, in which PRKAA2 is a notable changed molecule and miR-124-3p could regulate its expression and function in endothelial cells.

Keywords: PRKAA2; bioinformatic analysis; long non-coding RNA; miR-124-3p; tumor-endothelial adhesion.

FIGURE 1

Hierarchical clustering of aberrant expressed mRNAs (A) and lncRNAs (B) detected in the TC–EC model and EC alone. Every column represents a cell sample, and every row represents an mRNA or lncRNA probe. Red color indicates overexpression genes and green color indicates low-expression genes.

FIGURE 2

GO analysis of 504 differentially expressed mRNAs in TC–EC interaction, in which gland development, DNA replication, urogenital system development, morphogenesis of a branching epithelium, morphogenesis of a branching structure, branching morphogenesis of an epithelial tube, mesenchymal cell development, nuclear DNA replication, telomere maintenance via semi-conservative replication, and negative regulation of inclusion body assemble are the top 10 biological process (A); mitochondrial inner membrane, transferase complex (transferring phosphorus-containing groups), replication fork, and desmosome are main cell components (B); ATPase activity, catalytic activity acting on DNA, DNA-dependent ATPase activity, and helicase activity are the enriched molecular functions (C). P < 0.05 was considered as statistically significant.

FIGURE 3

Pathway analysis of 504 differentially expressed mRNAs in TC–EC interaction. (A) The most enriched 11 (-LgP > 3) of 40 pathways among up-regulated mRNAs (-LgP > 1). (B) The most enriched pathways among down-regulated mRNAs (-LgP > 1). The bar graphs represented the enrichment of these mRNAs. The value of (-LgP) was P-value taking the negative logarithm with base 10. (C) The interaction between the KEGG pathways of 504 differentially expressed mRNAs. The circle represents the pathway, and the size of the circle represents the degree of the pathway in the network. The larger the degree, the greater the circle, representing more pathway interaction. Lines between pathways represent their interrelationships. Red pathways represent the relative mRNAs that were upregulated, blue pathways represent the relative mRNAs that were downregulated, yellow represents the pathways with both up- and downregulated mRNAs.

FIGURE 4

Differential mRNA–mRNA interaction network. (A) The network graph of mRNA–mRNA interaction. The circle represents mRNA; the size of the circle represents the betweenness centrality value. The bigger the value, the more signal transmission is involved, indicating that the mRNA has greater ability to regulate mRNA interactions. (B) Expression level of the top 10 differential mRNAs between the TC–EC model and EC according to the differential mRNA–mRNA interaction network analysis.

FIGURE 5

mRNA–lncRNA co-expression network analysis of key mRNAs or lncRNAs in EC alone (A) and TC–EC model (B). In the network, the circle represents mRNA, the rhombic represents lncRNA. Red represents the upregulated mRNAs/lncRNAs, blue represents the downregulated mRNAs/lncRNAs. Those mRNA/lncRNAs without interaction relationships were not displayed in the network. (C) Expression level of the top 10 differential mRNAs or lncRNAs between the TC–EC model and EC according to the mRNA–lncRNA co-expression network analysis.

FIGURE 6

(A) The relative expression of PRKAA2 was significantly upregulated in the TC-EC model, compared with EC alone. (B) The prediction of miRNAs targeted at PRKAA2, using miRDB, RNA22, and Targetscan database. (C) qPCR analysis of relative expression levels of 4 predicted miRNAs targeted at PRKAA2, in which the expression of miR-124-3p was significantly decreased in endothelial cells adhered by PC-3M. (D) Predicted binding sites within the 3’UTR region of PRKAA2 with miR-124-3p using the Targetscan database. The data are presented as the means ± SD. (n = 3, **P < 0.01, ****P < 0.0001, TC-EC vs. EC).

FIGURE 7

PRKAA2 expression increased with miR-124-3p knockdown in endothelial cells adhered by PC-3M. (A) The expression level of miR-124-3p in endothelial cells under fluorescence microscope by green fluorescence protein. BF, bright field; GFP, green-fluorescent protein; NC-miR-124-3p: negative control of lentiviral vector; miR-124-3p-down: miR-124-3p knockdown with lentiviral vector. (B) qPCR analysis of relative expression levels of PRKAA2 in HUVECs transfected with LV-miR-124-3p-down. The results show that the expression level of PRKAA2 mRNA in endothelial cells adhered by PC-3M increased greatly with the knockdown of miR-124-3p.

FIGURE 8

MiR-124-3p targeted PRKAA2 3’UTR in HEK 293T cells. pcDNA3.1(-)-miR124-3p was co-transfected with luciferase reporter constructs containing pGL3-PRKAA2-3’UTR-WT or pGL3-PRKAA2-3’UTR-MUT in 293T cells; Firefly luciferase activity for each construct was normalized to the co-transfected Renilla luciferase construct presented miRNA-124-3p-NC transfected group. The data are presented as the means ± SD (n = 3, ***P < 0.001, PRKAA2-3’UTR-WT + miR-124-3p vs. PRKAA2-3’UTR-WT + miR-124-3p-NC).