Abstract
Abstrract : With an ever increasing world population, there is growing pressure on agricultural production to keep up with the demand for food, fuel and fiber products. Agricultural intensification has resulted in the conversion of approximately one-half of the Earth’s terrestrial surface to agricultural crops and range-land. This has significantly changed the natural landscape around agricultural borders, with the development of agro-ecological interface’ between wild ecosystems and crop or pasture lands. These areas are populated with crop plants, weeds of crop systems and non-crop plants that consist of both native and introduced plants. One very important aspect of this change in landscape, is the influence of microroganisms, particularly plant viruses, that move across the interface between crop and wild plants (virus reservoirs) and significantly impact agricultural production. Due to the potential role of these essentially unknown or poorly studied plant viruses in crop diseases and yield loss, they should be a focus of study. The aim of this project, was to characterize the circular viral DNA viral diversity in infecting indigenous or non-cultivated plant species in various cropping systems in South Africa, using a metagenomics approach. A survey was carried out in 2014 and 2015 in three Provinces (Limpopo, Mphumalanga and KwaZulu-Natal) of South Africa. Samples were collected from a total of 230 plants growing alongside cultivated crops and screened for viral infection. The viral diversity in these samples were explored using viral amplification by polymerase chain reaction (PCR), rolling circle amplification (RCA), RCA/restriction fragment length polymorphisms) (RCA/RFLP) and the results extended by next generation sequencing (NGS). Initial screening of the plants by PCR, using universal/degenerate begomovirus primers, found that 24% of the non-cultivated plants tested positive for possible begomovirus infection. To further reveal the full circular, viral diversity of the collected samples, they were screened by RCA/RFLP, followed by NGS for virus identification. The use of the RCA/RFLP was found to be a rapid, reproducible method to screen a large number of plant samples for viral biodiversity. Because of the high cost of NGS, a combined strategy was used, where initially RCA positive samples were pooled, and sequenced in one NGS run (NGS 2014, Pool-1 and 2), and after further screening by PCR, the plant samples suspected to contain circular viral molecules, were sequencing in individual NGS runs (NGS 2015). The pooled NGS, strategy was found to be an effective method for viral discovery, as it did allow the detection and characterization of both previously known viruses and previously unknown, novel virus genomes (Table 4.2). The drawback of the pooling-strategy was that for most viral contigs, the full viral genome could not be assembled and, post-NGS, it required a lengthy follow up procedure to trace the original plant samples, where the viral molecules originated from. 4 The data provided by the individual NGS runs (NGS 2015) allowed for the identification of both previously known viruses (four strains or variants of Tomato Curly Stunt Virus), and previously unknown, novel virus species (one monopartite and two bipartite begomoviruses, two betasatellite molecules and one genomovirus) in four different, uncultivated host species (Table 4.3). Previous studies have shown the widespread presence of ToCSV throughout South Africa and indeed, four ToCSV genomes were recovered in this study, both from tomato plants and weed species. The two ToCSV genomes recovered from the weeds species, Malvastrum coromandelianum and Acalypha indica, represent new natural host report for this virus and should be targeted for removal as they are acting as begomovirus reservoir in these cropping areas. The novel begomovirus species identified in this study, include one monopartite begomovirus, named Malvastrum curly stunt virus (MalCSV), that was identified in M. coromandelianum, along with an isolate of Tomato leaf curl Togo betasatellite (ToLCTGB) and one bipartite begomovirus, named Sida corlifolia golden mosaic virus (SiCGMV) that was identified in S. corlifolia, along with a novel betasatellite that was termed Sida corlifolia yellow mosaic betasatellite (SiCYMB).The MalCSV was identified from a symptomless host, but the SiCGMV and SiCYMB viral molecules were detected in a weed plant displaying symptoms of viral infection, i.e. bright yellow and green mosaic coloration of the apical leaves. The SiCGMV isolate displayed unique ‘NW-like’ genetic features and phylogenetic analysis grouped SiCGMV with other Corchoroviruses, a subgroup of whitefly transmitted viruses, genetically distinct, and basal to all other begomoviruses. These Old World (OW) viruses, with unique New Word (NW) genetic features, are likely due to distinct evolutionary histories or genetic isolation in their host species. It also lends support to the idea that NW begomoviruses may have originated in the OW and were subsequently disseminated to the NW, and that Corchoviruses may be a remnant of such NW begomoviruses that once populated the OW. Furthermore, this study provided the first reports of beta-satellite molecules in South Africa. A new bipartite begomovirus was identified in Phaseolus vulgaris (bean), for which the name Phaseolus vulgaris begomovirus was proposed. Phylogenetic analysis grouped this viral isolate with ScCBV, a member of the‘legumoviruses’ group. Begomoviruses infecting legumes often cluster phylogenetically as a group between the OW and NW begomoviruses and have collectively been referred to as ‘legumoviruses’. No symptoms were observed on the bean plant at the time of collection, thus further investigation into the biological characteristics and possible economic impact of this newly identified virus is required...
M.Sc. (Biochemistry)