A genome-wide identification, characterization and functional analysis of salt-related long non-coding RNAs in non-model plant Pistacia vera L. using transcriptome high throughput sequencing

Abstract

Long non-coding RNAs (lncRNAs) play crucial roles in regulating gene expression in response to plant stresses. Given the importance regulatory roles of lncRNAs, providing methods for predicting the function of these molecules, especially in non-model plants, is strongly demanded by researchers. Here, we constructed a reference sequence for lncRNAs in P. vera (Pistacia vera L.) with 53220 transcripts. In total, we identified 1909 and 2802 salt responsive lncRNAs in Ghazvini, a salt tolerant cultivar, after 6 and 24 h salt treatment, respectively and 1820 lncRNAs in Sarakhs, a salt sensitive cultivar, after 6 h salt treatment. Functional analysis of these lncRNAs by several hybrid methods, revealed that salt responsive NAT-related lncRNAs associated with transcription factors, CERK1, LEA, Laccase genes and several genes involved in the hormone signaling pathways. Moreover, gene ontology (GO) enrichment analysis of salt responsive target genes related to top five selected lncRNAs showed their involvement in the regulation of ATPase, cation transporter, kinase and UDP-glycosyltransferases genes. Quantitative real-time PCR (qRT-PCR) experiment results of lncRNAs, pre-miRNAs and mature miRNAs were in accordance with our RNA-seq analysis. In the present study, a comparative analysis of differentially expressed lncRNAs and microRNA precursors between salt tolerant and sensitive pistachio cultivars provides valuable knowledge on gene expression regulation under salt stress condition.

Figure 1

General flowchart of the pipeline used to identify of lncRNAs from reference transcriptome of P. vera.

Figure 2

The distribution of sequence length in lncRNAs compared with protein coding genes.

Figure 3

Common and specific salt responsive lncRNAs from two P. vera cultivars. (A) Venn diagrams show the number of common and specific differentially expressed lncRNAs between St6-Sc, Gt6-Gc and Gt24-Gc. (B) Venn diagrams show the number of common and specific differentially expressed lncRNAs between St6-Gt6 (differential expressed lncRNAs between Ghazvini and Sarakhs after 6 h salt treatment), Gt6-Gc and Gt24-Gc.

Figure 4

MD-plot showing the log-fold change of differential expressed lncRNAs compare to control and average abundance of each lncRNAs in three studied samples. Significantly up and down differential expressed lncRNAs are highlighted in red and blue, respectively. MD-plot was generated using EdgeR package.

Figure 5

Expression pattern of lncRNAs and coding-RNAs in two P. vera cultivars in different conditions. The size of boxes represents the expression levels variability of the contigs.

Figure 6

The log fold change value of 76 lncRNA-target pairs that were differentially expressed under salt stress predicted by Blast based transcript filtering and LncTar software.

Figure 7

Potential functions of 76 lncRNA-target genes that differential expressed under salt stress predicted by Blast based transcript filtering method and LncTar software. (A) Graphical results of molecular function. (B) Graphical results of biological process. Graphical results were prepared through AgriGO web-based tool.

Figure 8

Differentially expressed lncRNAs in pistachio. (A) Heatmap showing the expression pattern of 17 selected salt responsive lncRNAs compare to control under salt treatment. Heatmap was generated using the pheatmap package (https://cloud.r-project.org/web/packages/pheatmap/). (B) Regulation networks of five selected lncRNAs with the most interaction with their coding target genes. lncRNAs are represented by red nodes, and target genes are represented by blue nodes. The edges represent connections. Interaction networks was generated using the Cytoscape.

Figure 9

Graphical results of molecular function of potential 446 target genes of five selected lncRNAs that were differentially expressed under salt stress predicted by expression pattern based transcript filtering method. Graphical results were prepared through AgriGO web-based tool.

Figure 10

Validation of RNA sequencing results by Real-time PCR. Comparisons between Real-time validation and gene expression profiling data of top five salt responsive selected lncRNAs. The upper and lower graphs represent the Real-time PCR and the RNA-seq results, respectively.

Figure 11

Venn diagrams show the number of common and specific lncRNAs between pre-miRNA and C-lncRNAs.

Figure 12

The log fold change value of 14 miRNA-lncRNAs that differential expressed under salt stress in St6-Sc, Gt6-Gc and Gt24-Gc.

Figure 13

Venn diagram showing common and specific lncRNAs between different samples. (A) Venn diagrams show the common and specific differentially expressed miRNA-lncRNAs between St6-Sc and Gt6-Gc and also (B) between Gt6-Gc and Gt24-Gc.

Figure 14

The pipeline for the identification and target prediction of differential expressed miRNA-lncRNAs. (A) Identification and salt responsive target prediction of specifically differential expressed miRNA family between St6-Sc and Gt6-Gc. (B) Identification and salt responsive target prediction of specifically differential expressed miRNA family between Gt24-Gc and Gt6-Gc.

Figure 15

Gene ontology distribution results of coding target genes of selected miRNA-lncRNA family. (A) Targets of mir164 and 167 (B) Targets of mir393 and 477 (C) Targets of mir160 and 397. GO terms histogram was prepared through WEGO online tool.

Figure 16

Validation of RNA sequencing results of miRNA-lncRNAs by Real-time PCR. Comparisons between Real-time validation and gene expression profiling data of four salt responsive selected miRNA-lncRNAs. The two upper and one lower graphs represent the Real-time PCR and the RNA-seq results, respectively.

Figure 17

Common and genotype-specific salt responsive miRNAs between Ghazvini and Sarakhs and their target genes (transcription factor and transporter).