Groundnuts are predominantly grown in semiarid, tropic and subtropic ecoregions. In sub-Saharan Africa (SSA) groundnuts are grown for their nutritional content, providing proteins and vitamins to humans (Naidu et al. 1999) as well as a raw material in animal feeds (Kana et al. 2013). The cultivation of groundnuts is mainly for subsistence purposes within smallholder farms with no irrigation practiced. One of the limitations of production are viral diseases, in particular groundnut rosette. Groundnut rosette virus (GRV) is a positive, single stranded RNA virus classified within the genus Umbravirus (King et al. 2012). GRV is closely associated with a satellite RNA (sat-RNA) also referred to as helper RNA. The sat-RNA assists in the encapsulation of GRV into a coat protein that is lacking from GRV. This allows for the persistent transmission of GRV by the aphid Aphis craccivora (Naidu et al. 1999). Interestingly, GRV is confined to the SSA region (Waliyar et al. 2007; Okello et al. 2017). The estimated losses attributed to groundnut rosette disease caused by the combination of GRV, sat-RNA and Groundnut rosette assistor virus (GRAV, family Luteoviridae) range from $5 to $250 million USD per year in Nigeria and Zambia respectively (Waliyar et al. 2007). However, to date there have been limited efforts or opportunities to obtain the genomes of GRV. An increase in the availability of the GRV genomes in public databases would facilitate the development of molecular diagnostic tools for use by diagnosticians and plant breeders. In this study, we carried out two field surveys in the western highlands of Kenya across two seasons (2015/2016).

Groundnut leaf samples with chlorotic rosette-like symptoms were collected and stored in silica gel, and subsequently transported to the laboratory for analysis. Total RNA extraction was carried out using the Zymo RNA Miniprep kit (Zymo Research) according to the manufacturer’s instructions. Individual samples were used for cDNA library preparation using the Illumina Truseq Stranded Total RNA kit (Illumina) with plant ribozero depletion. Each library was dual indexed with unique adaptors and subsequently enriched with a PCR stage. Libraries were subsequently quantified and the correct insert size determined using the Tape Station Agilent 2200 (Agilent) using a D1000 screen tape and ladder (Agilent). Libraries with correct insert size and concentration were normalised, pooled and sequenced on rapid mode run using a single flow cell on the Illumina Hiseq 2500 at Macrogen Korea. We obtained a final genome of 4298 nt in length with an average coverage of 4487 times. The complete genome was deposited in GenBank (accession number MG646922). Assembly of this genome was from a total of 13,439,450 reads obtained after trimming to remove low quality bases. De novo assembly of the trimmed reads was carried out in CLC genomic workbench 8.5.1 (CLCGW) using set parameters as previously described (Kehoe et al. 2014). All contigs were subjected to NCBI BLASTn searches. GRV contigs were subsequently mapped to a reference genome (Z69910) using Geneious v. 8.1.8 (Biomatters). In addition, trimmed reads were imported into Geneious and mapped against reference GenBank sequences. Mapping parameters were set as follows: minimum overlap 10%, minimum overlap identity 80%, allow gaps 10% with fine tuning, iteration of up to 25 times. To confirm that the genome sequence was complete we compared the genomes obtained against those in GenBank. A partial genome was also generated in this study (SRF 54; MG646923) with 10,525,186 reads after trimming and only the ORF 3 and ORF 4 were used for phylogenetic analysis.

Sequence similarity search using BLASTn showed nucleotide sequence similarity of 84% to GRV isolates from Malawi (Z69910). Species demarcation within the genus Umbravirus is based on natural host range and a percent nucleotide identity of less than 70% (King et al. 2012). Based on these criteria, the GRV isolate from Kenya was considered to be a member of the Groundnut rosette virus species in genus Umbravirus. The complete genome length of 4298 nucleotides was larger than current reference genomes of 4019 nucleotides from GenBank. The extra 279 bases were found on the 3′ end outside the main ORF. Evolutionary relationships of GRV isolates from this study and reference sequences from GenBank were based on nucleotide sequences of ORF3 and ORF4. These two gene regions are the main markers used in the molecular diagnosis of GRV. Bayesian phylogenetic relationship analysis was carried out using the optimal evolutionary model GTR + I + G on MrBayes 3.2.2 (Ronquist et al. 2012) for 50 million generations, with sampling every 1000 generations with 25% of the sub optimal trees discarded. The consensus tree was viewed in FigTree v1.4.2. The sequences formed distinct clades based on geographical sampling locations (Fig. 1). These genome resources will have a direct impact in the development of molecular diagnostic tools in particular within SSA, where the virus remains pandemic.

Fig. 1
figure 1

Bayesian phylogenetic relationships of groundnut rosette virus (GRV) using nucleotide sequences of the open reading frames 3 (a) and 4 (b). Clade I, groundut rosette virus (GRV) Malawian isolates; clade II, GRV Kenyan isolates; clade III, GRV Nigerian isolates. Accession numbers for GRV Kenyan isolates: SFR 54, MG646923; SRF 57, MG646922

Full size image