Two main forms of common bean (P. vulgaris) are grown around the world: the snap bean (green pods harvested), and the dry bean (seeds harvested). In sub-Saharan Africa, dry beans (also referred to as grain) account for 95% of beans grown are used for local consumption, while the snap bean variety, a horticultural crop grown mainly for export accounts for 3% [41]. The remaining 2% is Tepary bean (Phaseolus acutifolius) which is localised in different parts of sub-Saharan Africa [41]. Consequently, the region is highly dependent on dry bean cultivation for its protein needs [25]. Most of the common bean varieties used in our study are primarily grown for grain and seed by smallholder farmers with less than 5 acres of land. An earlier metagenomic study had detected the presence of endornaviruses (PvEV1 and PvEV2) in common bean plants sampled in smallholder farms in peri-urban and rural Kenyan fields [31]. However, that study did not identify the specific common bean varieties harbouring these Endornaviruses.
Our study systematically screened common bean varieties from the east African region to determine if endornaviruses were widespread in popular lines. Using RT-PCR, we detected all three endornaviruses. A key finding was the first detection of PvEV3 in African-grown lines. PvEV3 had previously only been identified in a few North American cultivars [12]. The occurrence of endornaviruses was not linked to geographical location, and it is therefore likely that in east Africa, bean endornaviruses are spread by imports of germplasm from elsewhere and not due to local varietal preferences or inadvertent selection by local plant breeders.
We observed single infections of PvEV1 in 6 varieties. However, we did not observe any single infections of PvEV2, which was consistent with results reported earlier by Okada and colleagues when they screened 50 bean varieties, including wild relatives, in their seminal report on the detection of PvEV3 [12]. We did not observe any single infection of PvEV3, which contrasts with the findings of Okada and colleagues, who observed single PvEV3 infection in two domesticated varieties of bean, ‘Clouseau’ and ‘PI 209488’ of Andean and Mesoamerican origins, respectively, and in four wild bean varieties, all of Mesoamerican origin [12].
We observed mixed infection in six bean varieties. KK022, RWR2245 and SER16 had double infection with PvEV1 and PvEV2, and another three varieties, KK072, MCM 2001 and RWR 2073 of Kenya, Uganda, and Rwanda provenance, respectively, had mixed infection of all three endornaviruses. Okada and colleagues have reported mixed infections, either double or triple, for endornaviruses [12]. We did not observe any double infections involving either PvEV1-PvEV3 or PvEV2-PvEV3 combinations. This was different from the findings of Okada and colleagues, who reported detecting PvEV2-PvEV3 infection in variety ‘Red Rover’, an Andean variety and PvEV1-PvEV3 infection in two landraces (W6 12107 and PI 309885) of Mesoamerican origins [12]. Our screening of 26 popular varieties makes for a modest start. Should screening other bean varieties within the east Africa region for endornaviruses be required, we may likely encounter similar results. According to the Pan African Bean Research Alliance a bean research network active in 32 African countries, over 400 improved varieties adapted to different agro-ecological zones have been released for use in sub-Saharan Africa [41].
RNAseq analysis performed using near-isogenic lines of the cultivar Black Turtle Soup found 132 genes differentially expressed between plants doubly infected with PvEV1 and PvEV2 compared with control plants that were not infected with either of the endornaviruses [42]. Plants harbouring PvEV1 and PvEV2 yielded faster germinating seeds that gave rise to seedlings with longer radicles, had a lower chlorophyll content, but higher carotene content, produced longer pods, and yielded seeds with a greater average mass than control plants [20]. It is possible that the effects on gene expression [42] explain many, if not all, of the changes in plant physiology [20]. Arguably some of these virus-associated traits are beneficial to the plants and could be agronomically valuable. However, some traits conferred by other PVs, such as the amalgavirus STV, which appears to increase susceptibility to infection by other viruses [22], are not.
Using high-throughput sequencing and mapping to consensus sequences, we generated near full-length genomes for the three endornaviruses. Okada and colleagues [14] reported the length for PvEV1 to be 13908 nucleotides and that for PvEV 2 as 14820 nucleotides. The sequences we generated were longer for PvEV1 (14071 nucleotides) but similar for PvEV2 (14820 nucleotides). Meanwhile, our PvEV3 genome length at 15204 nucleotides was shorter than that reported by Okada and colleagues [12] by a single nucleotide in length (15205 nucleotides). These differences in length could be attributed to the methods used to generate the genome information. Pairwise comparisons of nucleotide and protein sequences showed that our reported sequences had sufficient inter-species differences for definitive sequence demarcation along species lines and showed apparent intra-species genetic diversity compared to other genomes in GenBank. Phylogenetic analyses of the whole genomes and helicase regions further evidenced this.
Common bean is susceptible to a range of plant pathogens, including viruses [28, 43–45]. Breeding for resistance has offered protection against some viral diseases. For example, the I gene protects against some strains of bean common mosaic virus; although it renders plants carrying this dominant resistance allele vulnerable to systemic necrosis induction by bean common mosaic necrosis virus [28]. However, for some pathogenic viruses detected in bean, such as CMV [31], there is no effective genetic resistance. Therefore, tools that enhance protection must be found. As noted earlier, research into endornaviruses shows a mixed record of what their presence can cause to plants [18]. Therefore, depending on whether or not these vertically inherited viruses are considered beneficial, their presence could be helpful for bean breeders to ensure or avoid the incorporation of endornaviruses or other PVs into new plant lines. Future studies should determine whether bean endornaviruses confer advantages to the host plant.