Each of the kits provided by Oxford Nanopore Technologies for RNA and cDNA library preparation comprise reagents that rely on the presence of a 3’ polyadenylated (poly(A)) tail on template molecules to successfully generate sequencing libraries. However, most prokaryotic RNA transcripts lack a 3’ polyadenylated (poly(A)) tail, including ribosomal (rRNA), transfer (tRNA) and protein-coding (mRNA) transcripts. In eukaryotes, mRNA transcripts are 3’ polyladenylated, however other transcripts of possible interest lack a poly(A) tail, such as some long-non-coding RNAs (lncRNAs), rRNAs and tRNAs.

It is possible to add a 3’ poly(A) tail to non-polyadenylated RNA transcripts enzymatically in vitro using E. coli poly(A) polymerase (NEB Cat No.: M0276L). The reaction incubation time at 37°C determines the length of the resulting 3’ poly(A) tails added to the RNA transcript.

Our recommended sample enhancement protocol starts with extracted, non-polyadenylated RNA to produce a clean, 3’ polyadenylated sample. This is suitable for every downstream RNA sequencing application, such as the Direct RNA Sequencing Kit (SQK-RNA002) and the cDNA Sequencing Kits (SQK-PCS111, SQK-PCB111.24 and SQK-DCS109), with a short preparation time of approximately 20 minutes. Our recommended polyadenylation protocol can be found here.

Tip: E. coli poly(A) polymerase will indiscriminately add an adenosine homopolymer to the 3’ end of any RNA transcript, therefore it is important to consider isolating the transcript(s) of interest before polyadenylating. For example, when targeting prokaryotic protein-coding mRNA transcriptomes for sequencing, we strongly recommend removing all prokaryotic rRNA and tRNA transcripts, which comprise 97-98% of a total RNA extraction before polyadenylating the remaining 2-3% of non-polyadenylated mRNAs. For prokaryote ribodepletion enrichment, we recommend using the ThermoFisher RiboMinus™ Transcriptome Isolation Kit for bacteria (Cat No: K155004).

While this ribodepletion kit is similar in scope and procedure to the 'Enrichment of polyadenylated molecules from a sample of total RNA by depletion of background rRNA' protocol, please follow the specific instructions for the bacterial kit if starting with prokaryotic total RNA.

Experiment I. Polyadenylation of non-polyadenylated control transcript.

A poly(A) tail was added to a 3’ non-polyadenylated RNA transcript, 300 nt in length and synthesized by in vitro transcription (IVT), using E. coli poly(A) polymerase (NEB M276L) and differing incubation times. For each time point (done in triplicate): 1 µg of RNA was polyadenylated (per protocol) and incubated (0, 0.5, 1, 5, 20 and 30 minutes) at 37°C before stopping the reaction with the addition of EDTA (10 mM final concentration). A negative control was incubated for 30 minutes at 37°C with no enzyme added. Samples were purified via a SPRI bead clean-up to remove EDTA and a final recovery yield was determined using an Invitrogen Qubit HS RNA kit and Qubit 4.0 fluorometer. For each sample, 100 ng was visualised using an Agilent Bioanalyzer 6000 Nano RNA kit and analysed using the mRNA Nano v2.5 software (Figure 1).

Figure 1. Polyadenylation with E. coli poly(A) polymerase. Polyadenosine homopolymer tails were added to the 3’ end of a non-polyadenylated IVT RNA transcript, 300 nucleotides long, using E. coli poly(A) polymerase. Incubations were carried out for 0, 0.5, 1, 5, 20 and 30 minutes at 37°C. A negative (-) control with no enzyme added was also incubated at 37°C.

An RNA sample of 300 ng from select time points were sequenced using the SQK-RNA002 kit on a MinION for 30 minutes. Poly(A) tail lengths were measured using Nanopore poly-ng pipeline and nanopolish. Incubation times in the range of 0.5–1.5 minutes resulted in a mean poly(A) tail length of 50–100 nucleotides in length (data not shown) and produced the highest sequencing read counts (Figure 2).

We recommend an optimum incubation time of 0.5–1.5 minutes, with a maximum E. coli poly(A) polymerase reaction incubation time of 5 minutes. Incubation times longer than 5 minutes may result in lower total sequencing yield.

Figure 2. Total read counts. Direct RNA Sequencing Kit (SQK-RNA002) total read counts for samples that were treated with E. coli poly(A) polymerase for the times indicated and run on a MinION flow cell for a total of 30 minutes.

Experiment II. Prokaryotic transcriptome.

As recommended in polyadenylation protocol, 40 µg of Escherichia coli ATCC 8739 total RNA was ribodepleted to enrich for prokaryotic mRNA; 3.75% (1.5 µg) of the total RNA input (now ribodepleted mRNA) was recovered.

Poly(A) tails were added to 500 ng of total and ribodepleted mRNA using E. coli poly(A) polymerase (as per protocol provided). Each reaction was quenched after 1.5 minutes by the addition of EDTA to a final concentration of 10 mM. Negative controls were done for each with no enzyme added. The EDTA was removed via a SPRI bead purification step (per protocol) and the 3’-polyadenylated mRNA recovery yield was measured using an Invitrogen HS RNA Qubit kit and Qubit 4.0 fluorometer. For each of the four sample conditions (total RNA and ribodepleted: +/- polyadenylation), 100 ng was sequenced using the Direct RNA Sequencing Kit (SQK-RNA002) on MinION with a 48 hour run-time. cDNA libraries were also prepped using the cDNA Sequencing Kit (SQK-PCS111), with 5 ng of ribodepleted polyadenylated RNA or 200 ng of total RNA and sequenced for 48 hours.

Reads were mapped to the E. coli (ATCC 8739) genome using minimap2 (Li, Bioinformatics, 2018) and per-gene counts estimated against the E. coli reference annotation. As expected, samples that were untreated with E. coli poly(A) polymerase (PAP) had a lower number of reads mapping to E. coli in both the direct RNA (Figure 3a) and cDNA preps (Figure 3b). Conversely, the reverse is observed when the E. coli transcripts are polyadenylated for both preps, demonstrating the augmentation of the sample by polyadenylation enables successful prepping. The number of E. coli reads observed in the non-polyadenylated libraries are likely to be transcripts that had a high number of naturally coded 3’ adenosines. For both library prepping methods, ribodepletion before polyadenylation produces a higher number of reads that uniquely map to the E. coli genome. The lower number of reads in both polyadenylated total RNA preps (direct RNA and cDNA) is likely due to poor mapping of rRNA transcripts to the multiple rDNA loci encoded in the genome, which are considered to be ambiguous mappers and discarded when Q-score >10 is set.

Figure 3. Number of mapped reads per species. Reads generated using (a.) direct RNA (SQK-RNA002) and (b.) cDNA (SQK-PCS111) sequencing were mapped to the E. coli ATCC 8739 genome.

To assess tail-lengths we used the alignment results and Oxford Nanopore fast5_research tools to sort the raw (fast5) data. Poly(A) tail lengths for the E. coli reads were then estimated using tailfindr (Krause et al., 2019). tailfindr was chosen as it can estimate poly(A) tails for both direct RNA and cDNA data, while nanopolish is restricted to direct RNA data only. The median poly(A) tail-lengths for the two polyadenylated libraries was estimated to be approximately 50–80 nucleotides (Figure 4, a and b). This is concurrent with the expected length range when incubating with E. coli poly(A) polymerase (PAP) for 0.5–1.5 minutes. Samples that were not polyadenylated were estimated to have negligible tail lengths.

Figure 4. Poly(A) tail length estimation. Poly(A) tail lengths of the E. coli transcripts were measured for a.) direct RNA (SQK-RNA002) and b.) cDNA (SQK-PCS111) reads using tailfindr.

References

Heng Li. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018 Sep 15;34(18):3094-3100

Maximilian Krause, Adnan M Niazi, Kornel Labun, Yamila N Torres Cleuren , Florian S Müller, Eivind Valen. tailfindr: alignment-free poly(A) length measurement for Oxford Nanopore RNA and DNA sequencing. RNA 2019 Oct;25(10):1229-1241

Change log

Version	Change
v3, April 2022	Grammatical correction
v2, March 2022	Updated product code error
v1, March 2022	Initial publication