Endopolygalacturonases (PGs) are the first cell wall degrading enzymes that are produced when pathogenic fungi encounter the host cell wall (Albersheim and Anderson, 1971). The role that these enzymes play in pathogenicity has been investigated for numerous pathogenic fungi. Although the results are not conclusive, there is evidence for some fungi that these enzymes are significant for their pathogenecity. Furthermore, plants contain polygalacturonase inhibiting proteins (PGIPs) in their cell walls, which are able to inhibit PGs (De Lorenzo et al, 2001; 2002). Colletotrichum SHK2148 is a pathogenic fungus causing anthracnose of lupin plants in South Africa. The identity of the fungus has been described as Colletotrichum tortuosum (Koch, 1996). However, this was based on morphological evidence only. Thus, the classification of the South African lupin- associated Colletotrichum isolates was re-assessed by comparing Colletotrichum SHK2148 on a morphological and molecular level to the recently described Colletotrichum lupinispecies (Nirenberg et al, 2002) as well as previously described Colletotrichum acutatum lupin anthracnose isolates (Talhinas et al, 2002). Based on the culture morphology, ITS and <font face=”symbol”>b</font>â-tubulin sequence data, it was concluded that Colletotrichum SHK2148 groups with C. lupini, more specifically, C. lupini var. setosum. The fungus, renamed Colletotrichum lupini SHK2148, was evaluated for its PG activity in pectin media (pH 5) over a 12 day growth period by using an agarose diffusion assay. The specific PG activity reached its highest level after three days, whereafter it decreased. Previous studies performed at the ARC, revealed that the fungus produced PG activity and this crude activity was inhibited by a PGIP produced in apple. A study was launched to isolate and characterise the gene(s) responsible for PG production. PG gene sequences from Colletotrichum gloeosporioides f.sp. malvae and Colletotrichum lindemuthianum were compared and conserved regions were identified from which primers were designed to amplify a fragment of a PG gene from C. lupini SHK2148. Inverse PCR was used to resolve the 5’ and 3’ sequences of the PG gene whereafter a complete copy of the gene was isolated from the genome of the fungus and characterised. The isolated gene was approximately 1Kb, contained a single intron of 59 bp and was very similar to the PG gene from C. gloeosporioides f.sp. malvae (cmpgII) as well as one of the PG genes (clpg2) from C. lindemuthianum. Southern blot analyses revealed that the gene was present as a single copy in the genome of the fungus. The in vitro expression of the PG gene from C. lupini SHK2148, grown in pectin media (pH 5), was investigated via northern blot analyses as well as RT-PCR, which revealed that the gene was expressed in the same time period that the highest PG activity was observed. A full cDNA copy of the PG gene was isolated using mRNA harvested from mycelia that was grown for 4 days on pectin. The cDNA copy confirmed the predicted intron position of the previously isolated genomic PG gene. Due to the unavailability of a full cDNA copy of the C. lupini SHK2148 PG gene at the time when expression studies were initiated, a complete cDNA copy was constructed by swapping an internal cDNA PG fragment with its counterpart in the complete genomic PG gene copy. The resulting cDNA PG copy was used as a template from which PG constructs were prepared for expression in Pichia pastoris. Constructs containing the PG gene with its native signal peptide, the PG gene with the β-MF signal peptide factor as well as hybrid constructs where the N terminal part of the mature PG proteins of Fusarium moniliforme and C. lupini SHK 2148 were exchanged, were transformed into P. pastoris . No PG activity was observed with an agarose diffusion assay for any of the Pichia clones. SDS-PAGE analyses were used to evaluate total protein isolations from the P. pastoris clones. The supernatant and cells of the clones were subjected to western blot analyses using antibodies directed againstAspergillus niger PG as well as F. moniliforme PG. The only positive hybridisation signal was observed between the A. niger antibody and a protein in supernatant extracts of the P. pastoris clones. However, the size of the hybridising band was very large. This could be due to glycosylation of the C. lupini SHK 2148 PG in P. pastoris, although the size increase is unusually large. The results indicated that it is unlikely that the C. lupini SHK 2148 PG was expressed in P. pastoris transformed with any of these constructs. Copyright / Dissertation (MSc)--University of Pretoria, 2010. / Plant Science / unrestricted