MicroRNAs of Bombyx mori identified by Solexa sequencing

Abstract

Background: MicroRNA (miRNA) and other small regulatory RNAs contribute to the modulation of a large number of cellular processes. We sequenced three small RNA libraries prepared from the whole body, and the anterior-middle and posterior silk glands of Bombyx mori, with a view to expanding the repertoire of silkworm miRNAs and exploring transcriptional differences in miRNAs between segments of the silk gland.

Results: With the aid of large-scale Solexa sequencing technology, we validated 257 unique miRNA genes, including 202 novel and 55 previously reported genes, corresponding to 324 loci in the silkworm genome. Over 30 known silkworm miRNAs were further corrected in their sequence constitutes and length. A number of reads originated from the loop regions of the precursors of two previously reported miRNAs (bmo-miR-1920 and miR-1921). Interestingly, the majority of the newly identified miRNAs were silkworm-specific, 23 unique miRNAs were widely conserved from invertebrates to vertebrates, 13 unique miRNAs were limited to invertebrates, and 32 were confined to insects. We identified 24 closely positioned clusters and 45 paralogs of miRNAs in the silkworm genome. However, sequence tags showed that paralogs or clusters were not prerequisites for coordinated transcription and accumulation. The majority of silkworm-specific miRNAs were located in transposable elements, and displayed significant differences in abundance between the anterior-middle and posterior silk gland.

Conclusions: Conservative analysis revealed that miRNAs can serve as phylogenetic markers and function in evolutionary signaling. The newly identified miRNAs greatly enrich the repertoire of insect miRNAs, and provide insights into miRNA evolution, biogenesis, and expression in insects. The differential expression of miRNAs in the anterior-middle and posterior silk glands supports their involvement as new levels in the regulation of the silkworm silk gland.

Figure 1

Reads counts and genomic distribution of the conserved and silkworm-specific miRNAs. (A) Detection of conserved and silkworm-specific miRNAs in the three libraries. (B) Genomic distribution of the conserved and silkworm-specific miRNAs. (C) Comparison of total reads count of the conserved and silkworm-specific miRNAs. (D) Averaged reads count per miRNAs. WB, whole silkworm body; AMSG, anterior-middle silk gland; PSG, posterior silk gland; WB-AMSG-PSG, across the three libraries.

Figure 2

Heterogeneity at the 5' and 3' ends of the sequenced tags. Pre-miRNAs, structures, and multiple isoforms of expressed mature bmo-mir-263a, bmo-let-7a and bmo-mir-317 sequences and their read counts are shown. Underlined sequences are annotated in miRBase 14.0. The boxed sequences have the highest read counts. The annotated mature sequence of bmo-mir-317 was not identified, while the maximal sequence in the whole body was strongly accumulated in the two other libraries, and consequently regarded as the most active isoform.

Figure 3

Annotated miRNA and maximal sequences from different arms of the precursors. The sequencing reads of four miRNAs in the whole body are depicted. Underlined sequences are the annotated mature sequences while the boxed sequences represent the maximum.

Figure 4

Sequence alignment of precursors of conserved paralogous miRNAs. The precursors were aligned using the CLUSTAL X program, and then highlighted with GeneDoc.

Figure 5

Sense and antisense transcripts of silkworm miRNAs. (A) Hairpin structures of the sense and antisense sequences of miRNA precursors. Mature miRNAs and miRNA stars are highlighted in red and blue, respectively. (B) Alignment of precursors. Mature miRNAand star sequences are underlined in red. (C) Schematic diagram of sense and antisense miRNAs and their potential target sites. Only one miRNA/target duplex proximal to CDS is presented (labeled with an asterisk). (D) Sense and antisense miRNAs target each other. All target sites were predicted using RNAhybrid2.2 and filtered with MirTif. The SVM score is higher than 1.0.