Profiling of MicroRNAs and Their Targets in Roots and Shoots Reveals a Potential MiRNA-Mediated Interaction Network in Response to Phosphate Deficiency in the Forestry Tree Betula luminifera

Abstract

Inorganic phosphate (Pi) is often lacking in natural and agro-climatic environments, which impedes the growth of economically important woody species. Plants have developed strategies to cope with low Pi (LP) availability. MicroRNAs (miRNAs) play important roles in responses to abiotic stresses, including nutrition stress, by regulating target gene expression. However, the miRNA-mediated regulation of these adaptive responses and their underlying coordinating signals are still poorly understood in forestry trees such as Betula luminifera. Transcriptomic libraries, small RNA (sRNA) libraries, and a mixed degradome cDNA library of B. luminifera roots and shoots treated under LP and normal conditions (CK) were constructed and sequenced using next-generation deep sequencing. A comprehensive B. luminifera transcriptome derived from its roots and shoots was constructed, and a total of 76,899 unigenes were generated. Analysis of the transcriptome identified 8,095 and 5,584 differentially expressed genes in roots and shoots, respectively, under LP conditions. sRNA sequencing analyses indicated that 66 and 60 miRNAs were differentially expressed in roots and shoots, respectively, under LP conditions. A total of 109 and 112 miRNA-target pairs were further validated in the roots and shoots, respectively, using degradome sequencing. Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of differential miRNA targets indicated that the "ascorbate and aldarate metabolism" pathway responded to LP, suggesting miRNA-target pairs might participating in the removing of reactive oxidative species under LP stress. Moreover, a putative network of miRNA-target interactions involved in responses to LP stress in B. luminifera is proposed. Taken together, these findings provide useful information to decipher miRNA functions and establish a framework for exploring P signaling networks regulated by miRNAs in B. luminifera and other woody plants. It may provide new insights into the genetic engineering of high use efficiency of Pi in forestry trees.

Keywords: Betula luminifera; Pi deficiency; abiotic stress; degradome; miRNA; transcriptome.

FIGURE 1

Changes in phenotype (A), P concentration (B), anthocyanin content (C), and APase activity (D) in B. luminifera after 7 days of Pi starvation. Bars indicate means ± SE (n = 3). P-values were obtained from t-tests between LP and CK conditions. **P < 0.01; *P < 0.05.

FIGURE 2

Expression profiling of –Pi-responsive DE mRNAs in B. luminifera. (A) The numbers of DE mRNAs in roots and shoots after 7 days of Pi starvation compared with control plants were determined by P ≤ 0.05 and log₂FC ≥ 1. Red and blue columns represent upregulated and downregulated DE mRNAs, respectively. (B) Venn diagrams showing the numbers of overlapping and specific DE mRNAs obtained across four comparisons. R, roots; S, shoots.

FIGURE 3

Global profiling of processed small RNAs (sRNAs) in B. luminifera. (A) Frequencies of processed sRNAs are expressed as percentages of the total number of redundant sequences. (B) Frequencies of processed sRNAs are expressed as percentages of the total number of non-redundant sequences. Four samples including CK_Roots (black bars), LP_Roots (white bars), CK_Shoots (striped bars), and LP_Shoots (dotted bars) in roots and shoots under normal and phosphate-deficient (–Pi) conditions.

FIGURE 4

Validation of the expression profiles of miRNAs in B. luminifera shoots and roots identified by sRNA-seq using qRT-PCR. (A) Relative expression of ten LP-responsive miRNA in shoots after 3, 7, and 15 days of LP treatment; (B) Relative expression of ten LP-responsive miRNA in roots after 3, 7, and 15 days of LP treatment. Expression level is represented by the ratio of LP treatment to CK. The value of relative expression level above one indicates that miRNAs are upregulated by Pi starvation, and the value below one shows that miRNAs are downregulated by Pi starvation. The experiments were repeated three times. Error bars indicate standard deviation.

FIGURE 5

Validation of miRNA target genes. (A) Validation of five miRNA targets by 5′ RACE. Matches between miRNAs and their corresponding target genes are indicated by straight lines. G:U wobbles are represented by circles. Arrows indicate cleavage sites, and the clone frequencies are shown. (B) Two NFYA genes were validated by transiently overexpressing BlMIR169a and BlMIR169c in B. luminifera leaves.

FIGURE 6

The putative functional networks for –Pi-responsive miRNAs and their corresponding target genes in B. luminifera. The network depicts the relationship of –Pi-responsive miRNAs and target genes from their expression patterns upon Pi deficiency in shoots (A) and roots (B). Triangle indicates miRNAs and circles are target genes. Red represents upregulated, green indicates downregulated, and gray means unchanged under Pi starvation.

FIGURE 7

qRT-PCR-derived expression analyses of eight miRNAs and their target genes under –Pi stress. To show the expression patterns of miRNAs and their corresponding target genes more directly, the ratio (K value) is used in the figure instead of individual relative expression levels. K is the ratio of the relative expression levels of LP (–Pi treatment) and CK (normal treatment); i.e., K = LP/CK. Each experimental time has a corresponding control group. A K-value > 1 indicates upregulated expression compared to the control, while a K-value < 1 indicates downregulated expression compared to the control. *P < 0.05; **P < 0.01. The expression levels of miR159a-GAMYB (A), miR397a-laccase1 (B), and BlmiR40-5′-MIEL1-like (C) were detected in the shoots, and miR169c-NFYA1 (D), miR171e-SCL27 (E), miR7122b-PPR (F), miR395b-APS1 (G), and miR399c-PLDD (H) in the roots were measured. All expression levels were normalized to TUA and MDH in shoots and roots, respectively. The experiments were repeated three times. Error bars indicate standard deviation.

FIGURE 8

Hypothetical model for the functions of miR169, miR395, miR397, and miR399 in response to Pi starvation. miR169a/b/c, miR395b, miR397a/b/c, and miR399c showed common and unique expression profiles in the shoots and roots when subjected to low Pi availability, then specifically inhibited expression of various target genes, thus affecting downstream biological processes. Arrows and blunted lines indicate positive and negative interactions, respectively. The left and right columns represent the log2(LP/CK) of the transcriptional levels in shoots and roots, respectively.