Abstract:Long non-coding RNAs (lncRNAs), whose sequences are approximately 200 bp or longer and unlikely to encode proteins, may play an important role in eukaryotic gene regulation. Although the latest maize (Zea mays L.) reference genome provides an essential genomic resource, genome-wide annotations of maize lncRNAs have not been updated. Here, we report on a large transcriptomic dataset collected from 749 RNA sequencing experiments across different tissues and stages of the maize reference inbred B73 line and 60 from its wild relative teosinte. We identified 18,165 high-confidence lncRNAs in maize, of which 6,873 are conserved between maize and teosinte. We uncovered distinct genomic characteristics of conserved lncRNAs, non-conserved lncRNAs, and protein-coding transcripts. Intriguingly, Shannon entropy analysis showed that conserved lncRNAs are likely to be expressed similarly to protein-coding transcripts. Co-expression network analysis revealed significant variation in the degree of co-expression. Furthermore, selection analysis indicated that conserved lncRNAs are more likely than non-conserved lncRNAs to be located in regions subject to recent selection, indicating evolutionary differentiation. Our results provide the latest genome-wide annotation and analysis of maize lncRNAs and uncover potential functional divergence between protein-coding, conserved lncRNA, and non-conserved lncRNA genes, demonstrating the high complexity of the maize transcriptome.

New lncRNA annotation reveals extensive functional divergence of the transcriptome in maize.