Most RNA sequencing experiments target RNA transcripts that are either nuclear or mitochondrial protein-coding polyadenylated transcripts, or other non-coding transcripts, such as long non-coding RNA (lncRNA). In eukaryotes, transcripts that are polyadenylated at the 3’ end (mRNA, some lncRNAs, etc) comprise 1-5% of a total RNA extraction, and other non-ribosomal RNA/non-polyadenylated transcripts can make up roughly 1-10% of a total RNA sample. Ribosomal RNA (rRNA) transcripts (5S, 5.8S, 18S, and 28S) make up roughly 80-90% of total RNA samples, each in extremely highly abundant copy numbers (Dual RNA-seq of pathogen and host; AJ Westermann, SA Gorski, J Vogel - Nature Reviews Microbiology, 2012).

Our Direct RNA (SQK-RNA002) and cDNA (SQK-PCS109 and SQK-DCS109) kits are designed to target 3’-polyadenylated transcripts for sequencing. Samples can be enriched for 3’-polyadenylated (poly(A)) transcripts by: removal of rRNA (“ribodepletion”), or selectively isolating poly(A) molecules (poly(A)+). Alternatively, total RNA can be used as an input sample wherein the 3’-polyadenylated transcripts are “self-selected” from the total RNA via the adapters and/or primers that are supplied in the sequencing kits.

RNA was extracted from Saccharomyces cerevisiae cells (S288c) and GM12878 cells using TRIzol-based methods: sequencing libraries were prepared from 1000 ng total RNA or 500 ng poly(A)-enriched RNA using the Direct RNA Sequencing Kit (SQK-RNA002) and run on MinION Mk 1B. Three different enrichment procedures were attempted; one method of ribodepletion and two methods of poly(A)+ isolation. Ribodepletion was performed with 10 µg of extracted total RNA using the RiboMinus™ Eukaryote Kit v2 (Invitrogen™), with ~5-10% of input RNA recovered. Poly(A)+ isolation was performed with 75 µg of extracted total RNA using either the DYNAL™ Dynabeads™ mRNA Purification Kit (Invitrogen™), or the NucleoTrap® mRNA Kit (Macherey Nagel): about 2% of input RNA was recovered.

Alignment was performed by minimap2-x map-ont to the Ensembl transcriptomes keeping primary alignments with minimum mapping quality of five. We observed a higher yield of aligned reads for the enriched yeast samples compared to the yeast total RNA, although this was not the case for the human samples (Figure 1). The length of the poly(A) tail for each read was also measured and variation in the median poly(A) tail length was detected after poly(A)+ isolation (compared with the total RNA control), particularly when using the NucleoTrap® mRNA Kit (Figure 2). This suggests that some bias may be involved during poly(A)+ selection, wherein molecules with longer tails are more readily targeted than molecules with shorter tails.

We recommend that customers are aware that some enrichment methods might influence the results that are obtained upon downstream sequencing.

Figure 1. Aligned read count for total RNA and poly(A)+ enriched RNA from yeast and human samples. We observed a higher yield of aligned reads for the enriched yeast samples compared to the yeast total RNA. Conversely, the human sample preps show comparatively similar aligned read yields among the three different enriched samples with the highest number of aligned reads for the total RNA input.

Figure 2. Measured median poly(A) tail length for total RNA and poly(A)+ enriched RNA from yeast and human samples. We observed an increase in the measured median poly(A) tail length after enrichment by selective isolation of polyadenylated molecules, particularly when using the NucleoTrap® RNA Kit.