Variability in P1 gene redefines phylogenetic relationships among cassava brown streak viruses

Cassava cuttings collected from CBSD-symptomatic plants in Malawi and Tanzania (Table 1) during national surveys in 2013 (under the auspices of each country’s agricultural research institutes). The plants were classified by having symptoms that were consistent with CBSD (feathery chlorosis along veins in leaves and brown streaks/lesions along the plant stem), or potentially were coinfected with agents causing both CBSD and CMD (mosaic, mottling, misshapen and twisted leaflets) and were taken to The Leibniz Institute – Deutsche Sammlung von Mikroorganismen and Zellkulturen GmbH (DSMZ) Plant Virus Department, where they were maintained under greenhouse conditions. Total RNA was extracted from the virus-infected leaves of the cassava plants using the cetyl trimethyl ammonium bromide method [21] with modifications described previously [22] or using an RNeasy Mini kit (Qiagen). Nucleic acids were quantified using a Nanodrop spectrophotometer, and about 2.0 × 10⁻⁵ μg/mL nucleic acid was used for virus detection by RT-PCR as detailed in Winter et al. [5]. A cDNA fragment, the partial sequence of the P1 gene, was amplified using virus specific primer sets designed by Winter et al. [5]. The reactions were performed in a GeneAmp 9700 PCR thermal cycler (Applied Biosystems, Foster City, CA, USA) set with the following conditions: 42 °C for 30 min for reverse transcription, followed by heat denaturation at 94 °C for 5 min; and then 35 cycles of amplification comprising the following: denaturation at 94 °C for 1 min, annealing at 52 °C for 1 min, extension at 72 °C for 1 min, followed by a single cycle of final extension at 72 °C for 10 min. All RT-PCR products were purified using a Qiagen gel extraction kit, ligated into the pDrive U/A cloning vector (Qiagen) and subsequently electroporated into Escherichia coli DH5α cells. The clones were Sanger sequenced in both orientations. A single consensus sequence for each isolate was verified to be CBSV by blastn searches of GenBank (https://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). The resulting nucleotide sequences were submitted to GenBank (pending accession numbers, Table 1).

Percentage nucleotide identities were computed in Geneious Software v10.0.5 [23]. A matrix of nucleotide identities was produced using the Sequence Demarcation Tool v1 [24]. Putative recombination events were detected using nine recombination detection programs within the RDP4 package (http://​darwin.​uvigo.​es/​rdp/​rdp.​html): RDP, GENECONV, MaxChi, Chimaera, Bootscan, Siscan, PhylPor, LARD, and 3Seq [25]. Analyses were carried out using default settings (except sequences were set to linear) and the Bonferroni correction P-value cut-off of 0.05. Only breakpoints supported by at least three methods were considered further [26].

Phylogenetic relationship among P1 regions of CBSV isolates (Table 1) was determined. The sequences were aligned using ClustalW [27] in MEGA 7 [28] and edited manually. The alignment was trimmed to give all sequences uniform length. MEGA 7 was used to construct maximum likelihood (ML) phylogenetic trees, and editing was done in FigTree v1.4.2 (http://​tree.​bio.​ed.​ac.​uk/​software/​figtree/​). The trees were created using a GTR nucleotide substitution model, and the best tree was bootstrapped with 1000 replicates [29].