Thesis (MSc)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: Streptomyces spp. are responsible for a large proportion of the world-wide quality deterioration of potatoes causing a potato tuber disease called cornmon scab. Determining the genetic diversity of the Streptomyces spp., especially the main pathogen, S. scabies, has been a prerequisite for the ultimate control of common scab. Techniques responsible for the classification and determination of genetic diversity have improved with advances in DNA technology. Analysis of South African (S.A.) S. scabies isolates has been focusing on the organisms' morphology, physiology, pathogenicity and melanin production, but the classification of S. scabies using DNA techniques has not yet been explored. In this study various DNA techniques were screened for optimal use in determining the genetic diversity within and among isolates of S. scabies. Bacteria had been sampled from the main potato producing regions in S.A. and a few other regions. The techniques explored included RAPDs, AFLPs, RAMS, Rep-PCR, 16S rDNA sequencing and ITS analysis. The first three techniques had to be abandoned due to non-reproducibility between the same isolate extracted on separate occasions and ITS analysis was abandoned due to sequencing difficulties. Of the three Rep-PCR techniques tested (BOX, ERIC and REP), BOX was selected because it produced the clearest and most reproducible results. BOX-PCR and 16S rDNA sequencing were therefore ultimately selected as the methods to analyse the genetic diversity of the S. scabies isolates. Information concerning the pathogenicity of the isolates was supplied by the Vegetable and Ornamental Plant Research Institute of the Agricultural Research Council (VOPI, ARC, Roodeplaat). A brief analysis of the pathogenicity prediction of the isolates in this study was explored with the PCR technique. Presence of the necJ gene was previously shown to be an indication of the pathogenicity within the Streptomyces spp. group. PCR analysis is based on the amplification of a O.72kb fragment (necl) in pathogenic isolates which was absent in non-pathogenic isolates. However, in this study the test for pathogenicity lacked specificity and sensitivity and some of the problems experienced included non-reproducibility between PCR reactions and the presence of the pathogenic fragment in the nonpathogenic isolates (as designated by VOPI, ARC). These observations led to the conclusion that this technique is not an ultimate test for pathogenicity of S. scabies isolates in a South African context. The genetic distances and similarity matrices of the Rep-PCR results were calculated using Nei's genetic distance calculation (Nei M, 1975). Clusters from these matrices were constructed using the unweighted pair group average (UPGMA) with the PAUP4 package. The clusters for the 16S rDNA sequences were formed with the Neighbor Joining (NJ) method and the PAUP4 package. The NJ trees do not take small sequencing differences into account, therefore a Parsimony Network had to be constructed. The trees obtained with the 16S rDNA sequencing techniques grouped most S. scabies isolates into one major group with a 100% bootstrap robustness of this group. More genetic diversity was illustrated by the BOX-PCR technique and the isolates were generally grouped according to their different regions of origin. However, the bootstrap values were low, indicating a lack of robustness regarding the BOXPCR clustering. This was not unexpected as the number of data points employed in the BOX technique is very limited. Both techniques revealed unexpected grouping of a few isolates. Their isolated positions could be attributed to possible misclassification or to the fact that they could be genetically different S. scabies isolates. Streptomyces spp. (other than S. scabies) displayed enough differences to place them in their own distinct groups using both techniques. Comparison of the cluster results obtained in this study did not correlate to the data supplied by the VOPI, ARC (morphology, physiology, pathogenicity and melanin production) which revealed differences between the S. scabies isolates within their respective regions. The lack of diversity displayed by the 16S rDNA technique can be attributed to the fact that only a limited section of the genome is involved making it inappropriate for intra-species genetic diversity analysis. The BOX technique takes various loci within the genome but is still not ideal for a thorough genetic diversity analysis. This study represents the first attempt to determine the genetic diversity of S. scabies in S.A. on DNA level. / AFRIKAANSE OPSOMMING: Streptomyces spp. is verantwoordelik vir 'n _groot deel van die wereld afname in aartappel kwaliteit as gevolg van die aartappelknol siekte bruinskurf. Die bepaling van die genetiese diversiteit tussen die Streptomyces spp., varal die hoof patogeen in die groep, S. scabies, is 'n vooreiste vir die uiteindelike beheer van bruinskurf. Tegnieke verantwoordelik vir die klassifikasie en bepaling van genetiese diversiteit het verbeter met vooruitgang in DNA tegnologie. Analise van Suid Afrika (S.A.) se S. scabies isolate konsentreer op die organisme se morfologie, fisiologie, patogenisiteit en malanien produksie, maar die klassifikasie van S. scabies met die behulp van DNA tegnieke is nog nie uitgevoer me. In hierdie studie is verskeie DNA tegnieke ondersoek vir optimale bepaling van genetiese diversiteit binne en tussen S. scabies isolate van S.A Bakteriee is verkry van die hoof aartappel-produserende areas in S.A. en ook van 'n paar ander areas. Die tegnieke wat in die studie gebruik is, het RAPDs, AFLPs, Rep-PKR, 16S rDNA volgordebepaling en ITS analise ingesluit. Die eersgenoemde drie tegnieke is uitgesluit as gevolg van nie-herhalende resultate tussen dieselfde isolaat geisoleer op verskillende geleenthede. ITS analise is uitgesluit as gevolg van probleme met volgordebepaling. Rep- PKR en 16S rDNA volgordebepaling is uiteindelik gekies as die mees geskikte metodes vir die analise van genetiese diversiteit tussen S. scabies isolate in hierdie studie omdat albei skynbare herhaalbare resultate gelewer het. Inligting met betrekking tot die patogenisiteit van die isolate is voorsien deur die Groente en Sierplant Instituut van die Landbou Navorsinsraad (YOPI, LNR). 'n Vinnige analise van die patogenisiteits voorspelling van die isolate is uitgevoer met die PKR tegniek. Dit is voorheen aangetoon dat die teenwoordiheid van die necJ geen dui op die patogenisiteit in die Streptomyces sp _groep. PKR analise het 'n O.72kb fragment (necl) in patogeniese isolate geamplifiseer wat nie teenwoordig was in niepatogeniese isolate nie. Hierdie toets vir patogenisiteit soos gebruik in hierdie studie was egter onspesifiek en onsensitief en sommige van die probleme wat ondervind is sluit in nie-herhaalbaarheid tussen PKR reaksies en die teenwoordigheid van die patogeniese fragment in die nie-patogeniese isolaat (soos beskryf deur YOPI, LNR). Uit hierdie waarnemings word afgelei dat die tegniek nie 'n geskikte toets vir patogenisiteit van S. scabies isolate in 'n S.A konteks is nie. Die genetiese afstande en ooreenkomstige matrikse van die Rep-PCR resultate is bereken met die genetiese afstand bepaling van Nei (Nei M, 1975). Groepe is gevorm met die "unweighted pair group average" (UPGMA) en PAUP4 pakket. Die "Neighbor Joining" (NJ) groepe is gevorm met die 16S rDNA volgordebepaling data mbv die PAUP4 pakket. Die NJ groepering neem nie klein volgorde verskille in ag nie en gevolglik moes'n "Parsimony Network" opgestel word. Die groepering met die 16S rDNA volgordebepaling het meeste van die isolate in een hoof groep geplaas met 'n 100% "bootstrap" waarde. Meer genetiese diversiteit is met die BOX-PCR tegniek gevind en isolate was oor die algemeen gegroepeer vol gens van hul oorsprong. Die "bootstrap" waardes vir die BOX tegniek was baie laag. Dit was nie onverwags nie, want die hoeveelheid data punte was beperk met die BOX tegniek. Albei tegnieke het 'n aantal afwykende isolate vertoon. Hul ge-isoleerde posisies kan toegeskryf word aan moontlike misklassifikasies van die isolaat. Die moontlikheid dat daar wel genetiese verkille tussen die isolate is, kan egter nie uitgesluit word nie. Streptomyces spp. (uitgesluit S. scabies) het genoeg variasie vertoon om hulle in hul eie groepe met die gebruik van beide tegnieke te plaas. Vergelyking tussen die groepe in die studie stem nie ooreen met die data verkry vanaf YOPl, LNR (morfologie, fisiologie, patogenesiteit en melanien produksie) nie wat verskille tussen S. scabies isolate binne 'n sekere gebied vertoon. Die gebrek aan diversiteit soos vertoon deur die 16S rDNA tegniek kan toegeskryf word aan die feit dat slegs 'n beperkte gedeelte van die genoom ondersoek word, wat dit ongeskik vir intra-species genetiese diversiteit analise maak. Die BOX tegniek neem verskeie loci in die genoom in ag, maar is steeds nie ideaal vir deeglike genetiese diversiteit analise nie. Hierdie studie verteenwoordig die eerste poging om die genetiese diversiteit van S. scabies in S.A. op DNA vlak te bepaal.