PacBio Full-length Transcriptome Analysis Provides New Insights into Transcription of Chloroplast Genomes

Abstract

Background: Chloroplast and mitochondrial DNA (cpDNA and mtDNA) are apart from nuclear DNA (nuDNA) in a eukaryotic cell. The transcription system of chloroplasts is different from those of mitochondria and eukaryotes. In contrast to mtDNA and nuDNA, the transcription of cpDNA remains unclear, mainly because the identification of the cp transcription initiation sites (TISs) and transcription termination sites (TTSs)remains unresolved. Results: In the present study, we characterized the transcription of cp genes with more comprehensive and accurate information using PacBio full-length transcriptome data of Arabidopsis thaliana. The main results included the discovery of four types of artifacts, the validation and correction of cp gene annotations, the discovery of polyA-like sites as TTSs and the exact identification of “GA”- and “GT”-type TISs on the genome scale. Notably, we proposed a new model to explain how the cptranscription initiates and terminates at the genome level. Conclusions: Four types of artifacts, degraded RNAs and splicing intermediates deserve the attention of every scientist who is working with PacBio full-length transcriptome data, as these contamination sequences could mislead the downstream analysis. The cp transcription initiates at multiple promoters and terminates at polyA-like sites. Our study provides new insights into the cp transcription and new clues to reveal the origins of promoters, TISs, TTSs and polyA tails of eukaryotic genes.