Effect of emodin on long non-coding RNA-mRNA networks in rats with severe acute pancreatitis-induced acute lung injury

Abstract

Long non-coding RNAs (lncRNAs) contribute to disease pathogenesis and drug treatment effects. Both emodin and dexamethasone (DEX) have been used for treating severe acute pancreatitis-associated acute lung injury (SAP-ALI). However, lncRNA regulation networks related to SAP-ALI pathogenesis and drug treatment are unreported. In this study, lncRNAs and mRNAs in the lung tissue of SAP-ALI and control rats, with or without drug treatment (emodin or DEX), were assessed by RNA sequencing. Results showed both emodin and DEX were therapeutic for SAP-ALI and that mRNA and lncRNA levels differed between untreated and treated SAP-ALI rats. Gene expression profile relationships for emodin-treated and control rats were higher than DEX-treated and -untreated animals. By comparison of control and SAP-ALI animals, more up-regulated than down-regulated mRNAs and lncRNAs were observed with emodin treatment. For DEX treatment, more down-regulated than up-regulated mRNAs and lncRNAs were observed. Functional analysis demonstrated both up-regulated mRNA and co-expressed genes with up-regulated lncRNAs were enriched in inflammatory and immune response pathways. Further, emodin-associated lncRNAs and mRNAs co-expressed modules were different from those associated with DEX. Quantitative polymerase chain reaction demonstrates selected lncRNA and mRNA co-expressed modules were different in the lung tissue of emodin- and DEX-treated rats. Also, emodin had different effects compared with DEX on co-expression network of lncRNAs Rn60_7_1164.1 and AABR07062477.2 for the blue lncRNA module and Nrp1 for the green mRNA module. In conclusion, this study provides evidence that emodin may be a suitable alternative or complementary medicine for treating SAP-ALI.

Keywords: acute lung injury; acute pancreatitis; emodin; lncRNA; mRNA.

Figure 1

Pathologic characteristics of lung tissue from SAP‐ALI rats with drug treatment. A, The experimental design of this study. B, Haematoxylin‐eosin staining of lung tissue. A representative image of haematoxylin‐eosin staining of lung tissue was selected for each group. The scale bar is 100 μm. C, The analysis of pathological scores. D, Quantitative analysis of amylase (AMY). E, Quantitative analysis of TNF‐α. F, Quantitative analysis of IL‐6. G, Analysis of blood gas. PaO2 (left) and PaCO2 (right) in blood was measured. Data are presented as the means ± standard deviation (SD). Student's t test was performed to compare 6 and 24 hours for each group with significance set at a P‐value of <.05. *P < .05, **P < .01. Different letters on each bar (lowercase for 6 h and uppercase for 24 h, respectively) indicate significant difference between two groups (Tukey HSD, P < .05)

Figure 2

Effect of drug treatment on the gene expression profile of lung tissue from SAP‐ALI rats. A, The pipeline for identification and expression analysis of mRNA and LncRNA genes expressed in rat lung tissue. B, Numbers of expressed mRNA and lncRNA genes expressed in at least one sample (with FPKM > 0). C, Principal component analysis (PCA) of 30 distinct samples based on the expression level of mRNA genes. The samples were clustered by treatment time (6 and 24 h). D, Heat map of correlation coefficients of 30 samples according to expression level of mRNA genes. E, Distribution of the number of differentially expressed mRNAs between two groups. F, Distribution of the number of differentially expressed lncRNAs between two groups

Figure 3

RT‐qPCR validation of DEGs and DELncRNAs. Relative expression levels of DEG and DELncRNAs by RT‐qPCR (up) and RNA‐seq (FPKM) (down). Data are presented as the mean ± standard deviation (SD). Student's t test was performed to compare 6 and 24 h with significance set at a P‐value of <.05. *P < .05, **P < .01

Figure 4

The co‐expression pattern of differentially expressed genes by weighted gene co‐expression correlation network analysis (WGCNA). A, Hierarchical clustering heat map of all samples based on all differentially expressed mRNA genes. B, Expression modules of differentially expressed mRNAs by WGCNA. C, Dendrogram of all differentially expressed mRNAs by hierarchical cluster analysis. Each co‐expressed gene was assigned to a module colour. D, Cluster analysis and heat map of each gene co‐expression module based on their correlations. E, Heat map of the correlations between each gene co‐expressions module (colour names) with traits (time, control, SAP, DEX and EMO). Pearson correlation coefficients with P‐values < .05 (in brackets) are presented. F, Eigengene pattern of each gene co‐expression module is presented by heat map

Figure 5

Functional analysis of mRNA genes from selected co‐expressed modules. A, The top10 GO analysis terms for mRNA genes from six co‐expressed modules. B, Functional analysis of mRNAs in green module. Genes enriched for two types of terms including response to chemicals and regulation of cell proliferation and apoptotic processes are presented. Green rhombuses are functional terms with size representing statistical significance. The red circles are the mRNAs. The line represents mRNAs involved terms with the degree of shade, the mRNA involvement and the number of terms involved. C, RT‐qPCR validation of the mRNA expression levels of Cdkn1a. RT‐qPCR (up) and RNA‐seq (FPKM) (down) are presented. Data are presented as means ± standard deviation (SD). Student's t test was performed to compare 6 and 24 h with significance set at a P‐value of <.05. *P < .05, **P < .01

Figure 6

Analysis of lncRNAs‐mRNAs co‐expression networks. A, Interactions between each lncRNA module and each mRNA module are shown. B, GO analysis terms of mRNAs co‐expressed with each lncRNA module are presented by Heat map. The top GO analysis term for mRNAs of each lncRNA module is shown. Colour degree represents the normalization of statistical significance of each terms (−log 10 [Q value]). C, Heat map showing expression patterns of lncRNAs from the blue module. D, The top10 Go analysis terms for co‐expressed mRNAs with LncRNAs from the blue module. E, The co‐expression network of lncRNAs from the blue module and co‐expressed mRNAs from black and green modules. LncRNAs are on the right and co‐expressed mRNAs are on the left. F, RT‐qPCR validation of the mRNA expression levels of Tbx2, Nrp1 and LncRNAs AABR07062477.2 and Rn60_7_1164.1. RT‐qPCR (up) and RNA‐seq (FPKM) (down) are presented. Data are presented as means ± standard deviation (SD). Student's t test was performed to compare 6 and 24 h with significance set at a P‐value of <.05. *P < .05, **P < .01