New insights of Arabidopsis transcriptome complexity revealed by direct sequencing of native RNAs

The transcriptome profiles of the model plant Arabidopsis thaliana have been extensively studied and charcaterised under different developmental and physiological conditions. However, most of these “RNA-sequencing” datasets have been generated using the sequencing of reverse-transcribed cDNAs from mRNAs that have a relatively short read length. Here, we performed direct RNA sequencing using the latest Oxford Nanopore Technology (ONT) with unusual read length. We demonstrate that the complexity of the A. thaliana transcriptomes has been under-estimated. The ONT direct RNA sequencing technology identified transcript isoforms at a vegetative (14 day old seedlings, stage 1.04) and a reproductive stage (stage 6.00-6) when 10% of the flowers had opened. In-house software called TrackCluster was used to determine alternative transcription initiation (ATI), possible alternative polyadenylation (APA), poly(A) length, alternative splicing (AS), and fusion transcripts. Tombo software was used to detect RNA base modifications. More than 38,500 novel transcript isoforms were identified, including six categories of fusion-transcripts which may result from differential RNA processing mechanisms. Fusion-transcripts are prone to mis-assembly by sequencing with short reads using next-generation-sequencing (NGS). These new transcript isoforms provide important additions to the annotated Arabidopsis genome. The power of ONT in detecting RNA modifications was demonstrated by characterisation of the modifications between mobile mRNAs and total mRNAs. The mobile mRNAs were enriched in m5C modifications, which is consistent with a recent finding that m5C modification in mRNAs is crucial for their long-distance movement. In summary, ONT direct RNA sequencing greatly enhances the identification of novel RNA transcript isoforms and RNA base modifications. Overall design: Arabidopsis thaliana col-0 seeds were surface sterilized with 50% commercial bleach and 0.01% Triton X-100 for 10 minutes, then washed 5 times with sterilized H2O. The sterilized seeds were sown on ½ MS plate with 1% sucrose and 0.6% Agar (sigma, cat#: A1296). Stratification at 4ºC for 48 hours, then placed in a growth chamber with 16h light/8h dark for 14 days. The 14-day seedlings were harvested (two biological replicates, each with one gram) for RNA extraction. Some seedlings were transplanted into soil for floral buds, the soil-based Arabidopsis seedlings were also treated with photoperiod of 16h light/8h dark. After another three-week growing, floral buds from plants equals to principal growth stage 6.0 to 6.1 were harvested, and they were immediately immersed in a foil boat full of liquid nitrogen in a relative larger liquid nitrogen container. Two Harvested biological replicates were frozen at -80degree freezer till RNA isolation. Total RNAs from Arabidopsis seedlings and floral buds were isolated with Trizol according to manufacturer's instruction. The extracted total RNAs were then treated with Dnase I (NEB CAT#:M0303). The treated total RNAs were extracted with acidic phenol (125:24:1, pH4.5, ThermoFisher cat#: AM9720) and precipitated with LiCl. The purified RNAs were subjected to mRNA isolation with Dynabeads mRNA purification kit (ThermoFisher cat#: 61006), around 0.9ug mRNAs for each library were used with Nanopore direct RNA sequencing kit (SQK-RNA002). The prepared libraries were loaded onto R9.4.1 flowcells, and sequencing was performed in MinION sequencing machine. Each library was run for 48 hours.