Differentially expressed lncRNAs in liver tissues of TX mice with hepatolenticular degeneration

Abstract

Wilson's Disease (WD), an ATP7B-mutated inherited disease that affects copper transport, is characterised by liver and nervous system manifestations. Long non-coding (ln-c) RNAs are widely involved in almost all physiological and pathological processes in the body, and are associated with numerous diseases. The present study aimed to elucidate the lncRNA-mRNA regulation network in a TX WD mouse model using RNA sequencing (RNA-seq). lncRNA expression profiles were screened using RNA-seq and real-time polymerase chain reaction, and differentially expressed lncRNAs and mRNAs were identified. To analyse the biological functions and pathways for the differentially expressed mRNAs, gene ontology and pathway enrichment analyses were performed. A significantly correlated lncRNA-mRNA relationship pair was calculated by CNC analysis to construct differential lncRNA and mRNA co-expression networks. A total of 2564 significantly up-regulated and 1052 down-regulated lncRNAs, and 1576 up-regulated and 297 down-regulated mRNAs, were identified. These genes were found to be associated with key processes such as apoptosis, and KEGG analysis revealed enrichment in the drug metabolism-cytochrome P450 pathway, PPAR signalling pathway, Notch signalling pathway, and MAPK signalling pathway. The identified differential lncRNAs may be involved in the pathogenesis and development of WD liver injury.

Figure 1

Haematoxylin and eosin staining in the control group showed clearly structured hepatic lobules, and the hepatocytes were stationary in the centre; veins radiated throughout the tissue, and the central veins, the arteriovenous structure, and the bile duct appeared normal. In the model group, there was extensive necrosis of hepatocytes and normal hepatic lobule structure disappeared; a large amount of inflammatory cell infiltration was evident. Masson staining showed that, in the control group, only a few collagen fibres were found in the manifold area and central vein. The liver tissue of the model group had extensive hyperplasia, extending from the manifold area to the surrounding area, forming a complete pseudolobular structure of different sizes.

Figure 2

(a) In total, we identified 62,559 lncRNA transcripts in which fragments per kilobase of exon per million reads mapped (FPKM) were above 0 among any one of 4 CN samples or 4 tx-j samples; further, 50,079 lncRNAs were expressed in both groups in which the FPKM was above 0 among any one of 4 CN samples and 4 tx-j samples (a). (b) The most common type of lncRNAs was intergenic, the next were intronic sense, exonic sense and exonic antisense, and intronic antisense was the least type. In addition, we analysed the distribution of identified lncRNAs on the mouse chromosomes; the 62,559 lncRNA transcripts could be found in all chromosomes, including ChrX and ChrY, and chromosome 2 included the most lncRNAs (b). Almost all chromosomes (excluding Chr19, ChrX, and ChrY) could generate more than 2000 lncRNA transcripts (b).

Figure 3

(a,b) Display the lncRNAs’ expression by use of volcano plots and heatmap. Supervised hierarchical cluster analysis shows that the different mRNA can correctly distinguish the model group from the control group.

Figure 4

(a) After selecting the Pearson’ s correlation coefficients of these lncRNAs and mRNAs equal to or > 0.98 (Table 1), a network is founded in each of the aforementioned groups using the Cytoscape program. The protein-coding genes belonging to the whole lncRNA-mRNA co-expression network were enriched with GO and pathway analysis, and the top 30 terms were presented (b,c).

Figure 5

As shown in (a), ENSMUST00000181536 (Snhg19) correlated with 222 protein-coding genes, and the enrichment analysis showed that Snhg18 may be related to leukocyte transendothelial migration, phagosome, glutathione metabolism, regulation of actin cytoskeleton, natural killer cell mediated cytotoxicity, focal adhesion, Fc epsilon RI signalling pathway, cell adhesion molecules, apoptosis, and Fc gamma R-mediated phagocytosis. As shown in (b), ENSMUST00000150851 (Meg3) correlated with 200 protein-coding genes, and KEGG enrichment analysis pointed out that it may participate in sphingolipid metabolism, regulation of actin cytoskeleton, glycosaminoglycan degradation, complement and coagulation cascades, adherens junction, leukocyte transendothelial migration, galactose metabolism, haematopoietic cell lineage, chemokine signalling pathway, p53 signalling pathway, B cell receptor signalling pathway, Fc epsilon RI signalling pathway, ABC transporters, mitogenactivated protein kinase (MAPK) signalling pathway, and apoptosis. As shown in (c), ENSMUST00000129245 (Meg3) correlated with 169 protein-coding genes, that enriched to dozens of KEGG pathway, such as gap junction, ABC transporters, peroxisome, glutathione metabolism, GnRH signalling pathway, complement and coagulation cascades, peroxisome proliferator-activated receptor (PPAR) signalling pathway, metabolism of xenobiotics by cytochrome P450, phagosome, tryptophan metabolism, fatty acid degradation, carbohydrate digestion and absorption, Notch signalling pathway, ribosome, and toll-like receptor signalling pathway. KEGG analysis revealed that two lncRNA transcripts were both related to ABC transporters and complement and coagulation cascades, although these two transcripts did not share the same gene of interaction. As shown in (d), ENSMUST00000136359 (H19) correlated with 169 protein-coding genes, including vitamin digestion and absorption, galactose metabolism, carbohydrate digestion and absorption, retinol metabolism, phagosome, gap junction, arachidonic acid metabolism, butanoate metabolism, complement and coagulation cascades, antigen processing and presentation, PPAR signalling pathway, ABC transporters, Notch signalling pathway, bile secretion, and MAPK signalling pathway.