Biological and molecular characterization of potyviruses infecting sweet potato
A survey was conducted to identify the viruses that infect sweet potato in the major growing areas in Kenya. Only Sweet potato feathery mottle virus (SPFMV), Sweet potato chlorotic stunt virus (SPCSV), Sweet potato mild mottle virus (SPMMV) and Sweet potato chlorotic fleck virus (SPCFV) were detected. Ninety-two percent and 25% of the symptomatic and asymptomatic plants, respectively, tested positive for at least one of these viruses. SPFMV was the most common and the most widespread, being detected in 74% of the symptomatic plants and in 86% of fields surveyed. Virus incidence was highest (18%) in Kisii district of Nyanza province and lowest (1%) in Kilifi district of Coast province. A hitherto undescribed isolate from sweet potato in Taiwan, referred to as 'sweet potato virus 2' (SPV2), was characterized. Its filamentous particles were 850 nm in length and induced cytoplasmic cylindrical inclusions consisting of pinwheels and scrolls. SPV2 attained strikingly higher titres and was more uniformly distributed in Ipomoea setosa plants dually infected with SPV2 and the crinivirus SPCSV than in plants infected with SPV2 alone. Comparison of the CP and 3' -UTR sequences of SPV2 with those of other potyviruses demonstrated that it is a distinct member of the genus Potyvirus (family Potyviridae). The name Sweet potato virus Y (SPVY) was proposed for this virus. To determine the variability within SPVY, the CP-coding region and 3' -UTR sequences of 13 isolates were sequenced. Sequence comparisons of the CP-coding region of these isolates revealed identities ranging from 82 to 99% at the nucleotide level and 86 to 99% at the amino acid level. Bootstrap analysis of sequences distinguished phylogenetic groups, which partially correlated with the geographical origin of the isolates and reflected differences in host range and symptoms. These results strongly suggest the occurrence of biologically and genetically diverse strains of SPVY. The CP-coding region and the 3' UTR of four Sweet potato virus G (SPV G) isolates from South Africa, China and Portugal was determined. Analysis of CP sequences revealed identities ranging from 81% to 98% and from 90 to 99% at the nucleic and amino acid levels, respectively. Phylogenetic analysis of the CP revealed a cluster comprising closely related isolates from USA, Europe, China and Egypt, as well as two clearly distinct isolates, one from South Africa and the other from China. These data are the first evidence for molecular variability among SPVG isolates and the first report of the occurrence of SPVG in Portugal and South Africa. The variability of geographically diverse SPFMV isolates in the CP gene and the 3' UTR was determined. The CP genes of the isolates sequenced had amino acid sequence similarities ranging from 82 to 98% and nucleotide sequence identities ranging from 80 to 99%. Phylogenetic analysis of the SPFMV sequences revealed two main distinct clusters, strain groups C and RC (sensu lata). The CP sequences of the C group were two amino acids shorter (due to a deletion of6 nucleotides in the N-terminus) than those of the strain group RC SPFMV. Isolates of strain group 0 were also shown to occur in East Africa. A high-affinity monoclonal antibody (MAb 1C4) reacting with all SPFMV isolates tested was obtained following immunisation with purified SPFMV virions of a Kenyan isolate (KY 115/1 S). The determination of the 5' -proxirnal part of the SPFMV genome indicated that the P3 gene could be the most variable part. Alignment of the PI sequences of the five African isolates ofSPFMV, with that of the previously only known SPFMV PI sequence from Japan revealed that the latter has a deletion of 75 nucleotides (25 amino acids) and is thus considerably shorter than the P I of the African isolates. This could indicate a critical source of variation in SPFMV.