Systematic Identification and Characterization of Long Non-Coding RNAs in the Silkworm, Bombyx mori

Long noncoding RNAs (lncRNAs) are emerging as important regulators in various biological processes. However, to date, no systematic characterization of lncRNAs has been reported in the silkworm Bombyx mori. In the present study, we generated eighteen RNA-seq data-sets with relatively high depth. Using an in-house designed lncRNA identification pipeline, 11,810 lncRNAs were identified for 5,556 loci. Among these lncRNAs, 474 transcripts were intronic lncRNAs (ilncRNAs), 6,250 transcripts were intergenic lncRNAs (lincRNAs), and 5,086 were natural antisense lncRNAs (lncNATs). Compared with protein-coding mRNAs, silkworm lncRNAs are shorter in terms of full length but longer in terms of exon and intron length. In addition, lncRNAs exhibit a lower level of sequence conservation, more repeat sequences overlapped and higher tissue specificity than protein-coding mRNAs in the silkworm. We found that 69 lncRNA transcripts from 33 gene loci may function as miRNA precursors, and 104 lncRNA transcripts from 72 gene loci may act as competing endogenous RNAs (ceRNAs). In total, 49.47% of all gene loci (2,749/5,556) for which lncRNAs were identified showed sex-biased expression. Co-expression network analysis resulted in 19 modules, 12 of which revealed relatively high tissue specificity. The highlighted darkgoldenrod module was specifically associated with middle and posterior silk glands, and the hub lncRNAs within this module were co-expressed with proteins involved in translation, translocation, and secretory processes, suggesting that these hub lncRNAs may function as regulators of the biosynthesis, translocation, and secretion of silk proteins. This study presents the first comprehensive genome-wide analysis of silkworm lncRNAs and provides an invaluable resource for genetic, evolutionary, and genomic studies of B. mori.