Comparative study on three viral enrichment approaches based on RNA extraction for plant virus/viroid detection using high-throughput sequencing

Abstract

High-throughput sequencing (HTS) has become increasingly popular as virus diagnostic tool. It has been used to detect and identify plant viruses and viroids in different types of matrices and tissues. A viral sequence enrichment method prior to HTS is required to increase the viral reads in the generated data to ease the bioinformatic analysis of generated sequences. In this study, we compared the sensitivity of three viral enrichment approaches, i.e. double stranded RNA (dsRNA), ribosomal RNA depleted total RNA (ribo-depleted totRNA) and small RNA (sRNA) for plant virus/viroid detection, followed by sequencing on MiSeq and NextSeq Illumina platforms. The three viral enrichment approaches used here enabled the detection of all viruses/viroid used in this study. When the data was normalised, the recovered viral/viroid nucleotides and depths were depending on the viral genome and the enrichment method used. Both dsRNA and ribo-depleted totRNA approaches detected a divergent strain of Wuhan aphid virus 2 that was not expected in this sample. Additionally, Vicia cryptic virus was detected in the data of dsRNA and sRNA approaches only. The results suggest that dsRNA enrichment has the highest potential to detect and identify plant viruses and viroids. The dsRNA approach used here detected all viruses/viroid, consumed less time, was lower in cost, and required less starting material. Therefore, this approach appears to be suitable for diagnostics laboratories.

Fig 1. Graphical representation of the three RNA extractions (or enrichment) approaches used in this research, i.e. dsRNA, ribo-depleted totRNA and sRNA.

The steps are indicated in orange boxes. The sequencing Illumina platforms are in green.

Fig 2. Venn diagram showing the viruses/viroid detected in all samples using different viral enrichment approaches (dsRNA, ribo-depleted totRNA and sRNA).

The overlapping regions correspond to the number of viruses/viroid detected by more than one approach. The detected viruses were PEMV1: pea enation mosaic virus 1, PEMV2: pea enation mosaic virus 2, WHAV2: Wuhan aphid virus 2, PNYDV: pea necrotic yellow dwarf virus, VCV: Vicia cryptic virus and PhCMoV: Physostegia chlorotic mottle virus, PSTVd: potato spindle tuber viroid, PvEV1: Phaseolus vulgaris alphaendornavirus 1, and PvEV2: Phaseolus vulgaris alphaendornavirus 2.

Fig 3. Percentage of reference sequences recovered by the reads of the RNA-based approaches on each of the normalised nucleotide subsamples (sizes: 1M, 10M, 20M, 30M, 40M and 50M nt).

The means of each approach are shown as blue circle: dsRNA, red triangle: ribo-depleted totRNA and yellow square: sRNA. The means are joined by lines with same colours. The vertical lines represent the standard deviation of the ten replicates. The strips over each graph are divided into two parts (upper: sample number, lower: virus/viroid acronym).

Fig 4. Mean depth of the RNA-based approaches for the detection of the viruses/viroid in this study on each of the normalised nucleotide depth (1M, 10M, 20M, 30M, 40M and 50M nt) for each subsample.

The means of each approach are shown as blue circle: dsRNA, red triangle: ribo-depleted totRNA and yellow square: sRNA. The means are joined by lines with same colours. The vertical lines represent the standard deviation of the ten replicates. The strips over each graph are divided into two parts (upper: sample number, lower: virus/viroid acronym).