Investigating Variability in Viral Presence and Abundance across Soybean Seed Development Stages Using Transcriptome Analysis

Abstract

Plant transcriptomes offer a valuable resource for studying viral communities (viromes). In this study, we explore how plant transcriptome data can be applied to virome research. We analyzed 40 soybean transcriptomes across different growth stages and identified six viruses: broad bean wilt virus 2 (BBWV2), brassica yellow virus (BrYV), beet western yellow virus (BWYV), cucumber mosaic virus (CMV), milk vetch dwarf virus (MDV), and soybean mosaic virus (SMV). SMV was the predominant virus in both Glycine max (GM) and Glycine soja (GS) cultivars. Our analysis confirmed its abundance in both, while BBWV2 and CMV were more prevalent in GS than GM. The viral proportions varied across developmental stages, peaking in open flowers. Comparing viral abundance measured by viral reads and fragments per kilobase of transcript per million (FPKM) values revealed insights. SMV showed similar FPKM values in GM and GS, but BBWV2 and CMV displayed higher FPKM proportions in GS. Notably, the differences in viral abundance between GM and GS were generally insignificant based on the FPKM values across developmental stages, except for the apical bud stage in four GM cultivars. We also detected MDV, a multi-segmented virus, in two GM samples, with variable proportions of its segments. In conclusion, our study demonstrates the potential of plant transcriptomes for virome research, highlighting their strengths and limitations.

Keywords: soybean seed development; transcriptome analysis; viral load shifts; viromes; virus.

Figure 1

Images of soybean tissues at five developmental stages. All images were captured by Won Kyong Cho in Gadam-ri, Hoengseong-gun, Republic of Korea. (A) Apical bud, (B) flower bud, (C) open flower, (D) developing pod at five days post-fertilization, (E) developing pod at fifteen days post-fertilization. Red-colored arrows are used to highlight each specific tissue.

Figure 2

The identification of viruses from 40 transcriptomes’ data derived from Glycine max (GM) and Glycine soja (GS). (A) The number of viral contigs for the identified viruses from GM and GS are labeled in blue and orange, respectively. (B) The proportion of identified viruses is shown by combining 40 soybean transcriptomes based on the number of contigs assigned to each virus. The number of viral contigs in each sample derived from GM (C) and GS (D). Bean plots display the number of viral contigs in each developmental stage for GM (E) and GS (F). In the bean plot, the black solid lines represent the median values, while the white lines depict individual data points.

Figure 3

The number of viral reads assigned to each virus and the proportion of viral reads in each sample. (A) The number of viral reads for the identified viruses is shown by combining 40 soybean transcriptomes, with blue and orange colors indicating GM and GS, respectively. (B) The proportion of viruses is shown according to the viral reads assigned to each virus. The proportion of viral reads in each sample derived from GM (C) and GS (D) is displayed. The proportion of viral reads in each developmental stage derived from GM (E) and GS (F) is shown. In the bean plot, the black solid lines represent the median values, while the white lines depict individual data points.

Figure 4

The proportion of identified viruses in each sample is based on the FPKM (fragments per kilobase of transcript per million) value. (A) The FPKM value representing viral transcripts for each virus is shown, with blue and orange colors indicating GM and GS, respectively. (B) The proportion of identified viruses is shown by combining 40 soybean transcriptomes based on the FPKM values. The proportion of viral transcripts in each sample based on the FPKM values is displayed for GM (C) and GS (D).

Figure 5

The proportion of viral transcripts in each developmental stage based on the FPKM values. Violin plots display the distribution of the total amount of viral transcripts derived from four samples for each soybean developmental stage for GM (A) and GS (B). The distribution of viral transcripts in each soybean developmental stage for SMV is shown for both GM (C) and GS (D). The distribution of viral transcripts in each soybean developmental stage for CMV is shown for both GM (E) and GS (F). The distribution of viral transcripts in each soybean developmental stage for BBWV2 is shown for both GM (G) and GS (H).

Figure 6

The comparison of viral transcripts for viruses composed of multiple genome segments. (A) A comparison of viral transcripts among DNA segments of MDV in two different samples. (B) A comparison of viral transcripts among RNA segments of CMV in different samples. (C) A comparison of viral transcripts between two RNA segments of BBWV2 in different samples from GM. (D) A comparison of viral transcripts between two RNA segments of BBWV2 in different samples from GS.

Figure 7

Size distribution of assembled viral contigs from GM and GS. (A) Viral contigs assembled using the Trinity assembler. (B) Viral contigs assembled using the Trinity assembler followed by the CAP3 assembler.

Figure 8

Mapping results of SMV-associated reads from the GS3-AB sample to the SMV reference genome. (A) Alignment outcomes of raw data sourced from the GS3-AB sample (SRR7538229) were computed against the SMV reference genome utilizing BWA software and subsequently visualized using the IGV viewer. Within the visualization, inverted triangles colored in red and blue delineate gap regions where the reads failed to align. A focused view on the mapping results in proximity to nucleotide positions 720 (B) and 1300 (C) within the SMV reference genome is presented. (D) Zoomed-in depiction of the mapped reads, represented by bars of distinct colors. Notably, gray signifies nucleotides not found within the reference genome, indicating insertion or deletion events (indels) relative to the reference sequence. Adenine (Green), thymine (Red), cytosine (Blue), and guanine (Orange) are color-coded representations of nucleotide bases.

Figure 9

Visualization of contigs associated with SMV. (A) Contigs assembled from GM samples using the Trinity assembler aligned to the SMV reference genome. (B) Contigs assembled from GM samples using the Trinity assembler, followed by the CAP3 assembler, aligned to the SMV reference genome. (C) Contigs assembled from GS samples using the Trinity assembler aligned to the SMV reference genome. (D) Contigs assembled from GS samples using the Trinity assembler, followed by the CAP3 assembler, aligned to the SMV reference genome. The obtained viral contigs do not map to the nucleotide regions indicated by the red-colored bars on the SMV reference genome.