Improving Reference Genome Resources Using Long-Read Sequencing Technology

Zea mays is an important crop species and genetic model for elucidating transcriptional networks in plants. Uncertainties about the complete structure of mRNA transcripts, particularly with respect to alternatively spliced isoforms, limit the progress of research in this system. In this study, we used single-molecule sequencing technology to investigate the maize transcriptome. Intact full-length cDNAs from six tissues of the maize inbred line B73 were barcoded, pooled, size-fractionated (<1 kb, 1–2 kb, 2–3 kb, 3–5 kb, 4–6 kb, and 5–10 kb), and sequenced on the PacBio RS II platform with P6-C4 chemistry. The resultant 111,151 transcripts captured ~70% of the annotated genes of the maize RefGenV3 genome assembly. A large proportion of transcripts (57%) represented novel, sometimes tissue-specific, isoforms of known genes, and 3% corresponded to novel gene loci. In other cases, the identified transcripts have improved existing gene models. To validate transcript structures we checked for occurrence of each splice-junction within high-depth Illumina reads generated from matched tissues. Averaging across all six tissues, 90% of splice-junctions were well supported by short-reads in matched tissues. In addition, we identified a large number of novel long non-coding RNAs (lncRNAs) and fusion transcripts, and found that DNA methylation plays important roles in generation of various isoforms. Our results show that the characterization of the maize B73 transcriptome is far from complete, and that maize gene expression is more complex than previously thought.