Genome-Wide Analysis of MicroRNA Responses to the Phytohormone Abscisic Acid in Populus euphratica

Abstract

MicroRNA (miRNA) is a type of non-coding small RNA with a regulatory function at the posttranscriptional level in plant growth development and in response to abiotic stress. Previous studies have not reported on miRNAs responses to the phytohormone abscisic acid (ABA) at a genome-wide level in Populus euphratica, a model tree for studying abiotic stress responses in woody plants. Here we analyzed the miRNA response to ABA at a genome-wide level in P. euphratica utilizing high-throughput sequencing. To systematically perform a genome-wide analysis of ABA-responsive miRNAs in P. euphratica, nine sRNA libraries derived from three groups (control, treated with ABA for 1 day and treated with ABA for 4 days) were constructed. Each group included three libraries from three individual plantlets as biological replicate. In total, 151 unique mature sequences belonging to 75 conserved miRNA families were identified, and 94 unique sequences were determined to be novel miRNAs, including 56 miRNAs with miRNA(*) sequences. In all, 31 conserved miRNAs and 31 novel miRNAs response to ABA significantly differed among the groups. In addition, 4132 target genes were predicted for the conserved and novel miRNAs. Confirmed by real-time qPCR, expression changes of miRNAs were inversely correlated with the expression profiles of their putative targets. The Populus special or novel miRNA-target interactions were predicted might be involved in some biological process related stress tolerance. Our analysis provides a comprehensive view of how P. euphratica miRNA respond to ABA, and moreover, different temporal dynamics were observed in different ABA-treated libraries.

Keywords: ABA; Populus euphratica; high-throughput sequencing; microRNA; target.

Figure 1

The Response of Populus euphratica to ABA. (A) Photosynthetic rate, (B) stomatal conductance, (C) intercellular CO₂ concentration, and (D) transpiration rate of P. euphratica in response to ABA treatments for different amounts of time. Data are mean ± SE (n = 6). The values with different lowercase letters are significantly different at the P < 0.05 level.

Figure 2

Length distributions of small RNAs in nine samples. Distribution of the sequence lengths of the sRNA from the nine libraries. Counts are based on unique sequences rather than the number of reads per unique sequence.

Figure 3

Conserved miRNA families in Populus euphratica and across species. Twenty-two representative conserved miRNA families in 23 angiosperms. All miRNAs of P. euphratica were identified based on sRNA sequencing data, and those of other plants were taken from miRBase (Release 21).

Figure 4

Predicted miRNA precursor stem–loop structures of novel miRNA precursors. Precursor structures of four novel P. euphratica miRNAs (miR-n26, miR-n59, miR-n65, and miR-n88) were predicted by the online software MFOLD. The MEFs are shown after the miRNA names. The mature miRNA and miRNA star sequences are marked in red and blue, respectively.

Figure 5

Verification of selected miRNAs by real-time quantitative PCR. Differentially expressed miRNAs identified by sequencing were confirmed by real-time qPCR, and their expression levels were compared among the three groups. The expression level of miRNA in deep sequences was performed with R statistical software package, which was“DESeq2” library used with raw date. ^*P < 0.05, ^**P < 0.01. (A) The miRNA expression for the comparison SL/CL; (B) The miRNA expression for the comparison LL/CL; (C) The miRNA expression for the comparison LL/SL.

Figure 6

GO analysis of miRNA putative target genes. GO annotation categorized all of the predicted miRNA target genes and differentially expressed miRNA target genes into biological processes, molecular functions, and cellular components.

Figure 7

The expression profiles of predicted target genes and their corresponding miRNAs by real-time quantitative PCR. Complementary miRNAs and predicted target gene were confirmed by real-time qPCR. The level of every gene in the control was set at 1.0. Error bars the standard deviation of three replicates. The values with different lowercase letters are significantly different at the P < 0.05 level. (A) The relative expression of miR395a/b with the predicted targets CCG007855.1 and CCG033595.2; (B) The relative expression of miR475a-3p with the predicted targets CCG020392.1; (C) The relative expression of miR6421 with the predicted targets CCG011696.1; (D) The relative expression of miR-n87 with the predicted targets CCG033430.1 and CCG0327391.1; (E) The relative expression of miR408-5p with the predicted targets CCG004054.1, CCG009428.1 and CCG032972.1; (F) The relative expression of miR408-3p with the predicted targets CCG027148.1 and CCG033426.1; (G) The relative expression of miR482a.1 with the predicted targets CCG010532.1; (H) The relative expression of miR-n10 with the predicted targets CCG020696.1; (I) The relative expression of miR-n49 with the predicted targets CCG008392.2; (J) The relative expression of miR6462a-3p with the predicted targets CCG012018.1 and CCG014959.1.