Abstract:
The taxonomy of Fusarium has long been a contentious field. Recent syntheses have brought a measure of order to classification at the species level. At the sub-specific level identification is difficult, requiring biological indexing, and knowledge of genetically isolated groups and their interrelationships is scant. This is particularly true for F. oxysporum, a widespread plant pathogen. Molecular methods, ranging from protein electrophoresis to monoclonal antibodies, have been used for some decades in attempts to find identifying markers for sub-specific groups with limited success. The exponential increase in DNA technology has opened several new avenues for the investigation of these interrelationships and the utility of some was investigated using Fusarium oxysporum f. sp. dianthi as a test system. A range of Fusarium species was screened for DNA restriction fragment length polymorphisms (RFLPs) by hybridisation to Southern blots of random probes cloned from an isolate of F. oxysporum f. sp. dianthi. The RFLPs derived from combinations of probes and restriction enzymes enabled differentiation at species, forma species and isolate levels, and patterns were conserved over time despite phenotypic variation. One probe hybridised only to a subset of the species F. oxysporum, giving a multi-band pattern, and not to isolates of the seven other Fusarium spp. This probe was further used on forty-six isolates of Fusarium from dis- eased carnations which were examined for RFLPs and vegetative compatibility groups (VCGs). Both RFLPs and VCGs divided the isolates into two major and three minor corresponding groups. Testing of subsets of the groups showed the major groups to be pathogenic. The two methodologies gave equivalent results for the genetic separation of populations, and it is suggested that in many cases these methods will be superior to differential hosts for the establishment of taxonomic units. Investigation of this highly specific probe found it to be derived from nuclear DNA. Subclones were explored, and the portion responsible for the multiband pattern was sequenced. Computer searches failed to find homology with any known sequence, nor were there any features of particular note in the sequence. It was noted that two sub-clones could be used to provide a very specific dot blot test which would determine pathogenicity of Fusarium isolates from diseased carnations. This technique was developed into a rapid, non-radioactive dot blot test for general use.