Genome-wide analysis of microRNAs in rubber tree (Hevea brasiliensis L.) using high-throughput sequencing

Abstract

MicroRNAs (miRNAs) are short RNAs with essential roles in gene regulation in various organisms including higher plants. In contrast to the vast information on miRNAs from many economically important plants, almost nothing has been reported on the identification or analysis of miRNAs from rubber tree (Hevea brasiliensis L.), the most important natural rubber-producing crop. To identify miRNAs and their target genes in rubber tree, high-throughput sequencing combined with a computational approach was performed. Four small RNA libraries were constructed for deep sequencing from mature and young leaves of two rubber tree clones, PB 260 and PB 217, which provide high and low latex yield, respectively. 115 miRNAs belonging to 56 known miRNA families were identified, and northern hybridization validated miRNA expression and revealed developmental stage-dependent and clone-specific expression for some miRNAs. We took advantage of the newly released rubber tree genome assembly and predicted 20 novel miRNAs. Further, computational analysis uncovered potential targets of the known and novel miRNAs. Predicted target genes included not only transcription factors but also genes involved in various biological processes including stress responses, primary and secondary metabolism, and signal transduction. In particular, genes with roles in rubber biosynthesis are predicted targets of miRNAs. This study provides a basic catalog of miRNAs and their targets in rubber tree to facilitate future improvement and exploitation of rubber tree.

Fig. 1

Size distribution of small RNAs in rubber tree leaves. The percentage of small RNAs of a particular length in the total small RNA population is shown

Fig. 2

The read frequencies of conserved miRNAs in rubber tree leaves. The total number of reads corresponding to a particular miRNA family was normalized against the number of total small RNA reads and expressed in transcripts per million (TPM)

Fig. 3

Diagrams of miRNA-target pairing between two conserved miRNAs (miR160 and miR172) and 20 novel miRNAs and their selected predicted targets. A–U and G–C hydrogen bonds are represented by two dots and the G–U hydrogen bonds are represented by a single dot. The numbers in the target genes represent the positions of the target sites within the indicated EST sequences. EC606382 is a GenBank accession number for a rubber tree EST. EC606382 encodes a protein identical to rubber elongation factor (GenBank accession P15252 from Hevea brasiliensis). The accession numbers for the other targets are from either the Euphorbia esula (leafy spurge) DFCI gene index release 1 or the Manihot esculenta (cassava) DFCI gene index release 1. TC12051 is a cassava EST that encodes a protein homologous to Hevamine A from rubber tree (GenBank accession P23472 from Hevea brasiliensis)

Fig. 4

Northern blot analysis of rubber tree miRNA accumulation in leaves. Total RNA (10 μg) from mature (m) and young (y) leaves from two rubber tree clones PB260 (260) and PB217 (217) was resolved in 15 % polyacrylamide/8 M urea/0.5x TBE gels and subsequently transferred to membranes for northern blot analysis. Membranes were hybridized with an oligonucleotide antisense to each miRNA. 5S rRNA served as a loading control. In some cases, membranes were stripped and used for several hybridizations such that one 5S blot was the loading control for all the miRNAs shown above

Fig. 5

Stem–loop RT-PCR analysis of rubber tree miRNA accumulation in leaves. The expected size of the RT-PCR products is 60 bp. For hbr-cand02 and hbr-cand13, PCR were conducted for 35 cycles with annealing at 64 °C. For hbr-cand05 and hbr-cand12, PCR were conducted for 28 and 35 cycles, respectively, with annealing at 60 °C