This work reports the continued search for new antibiotics in the relatively under investigated region of KwaZulu-Natal, South Africa. A soil bacterium designated strain N8 with antibacterial activity against both Gram-positive and Gram-negative bacteria was isolated from a poultry farm in Pietermaritzburg, South Africa. The organism was one of approximately 2600 strains isolated from various habitats in the KwaZulu-Natal midlands, South Africa during an actinomycete screening programme. The highest number of antimicrobially-active isolates came from a forest soil site whereas the lowest number was present in a riparian soil. Morphological, physiological and cultural characteristics indicated that strain N8 belonged in the genus Intrasporangium. In the literature, members of this actinomycete genus have not been associated previously with antibiotic production. Studies on the influence of different nutritional compounds on antibiotic production showed that the highest antibacterial activities were obtained when glycerol at 1% (w/v) was used as sole carbon source in the presence of mineral trace elements. Using solvent extraction and various chromatographic techniques, the antibiotic produced by strain N8 was recovered from the fermentation broth. The use of a three-solvent system, petroleum ether: acetone: ethyl acetate enhanced the separation of the antibiotic complex in broth. Bioassay results established that the antibacterial agent was in the ethyl acetate fraction (EAF) and chromatographic methods were used in its purification. The chromatographic methods used were: flash column chromatography (FCC), thin-layer chromatography (TLC), and Harrison research chromatotron (HRC). Further purification was carried out by reverse phase high performance liquid chromatography (HPLC). Most of the inactive, coloured material was removed from the antibiotic extract by FCC, while TLC chromatograms run using a range of the most polar to the least polar solvent systems [SS1 (most polar) – SS5 (least polar)] showed best separation of EAF with SS2. TLC chromatograms using SS2 usually showed 3 bands. Bioautograms of SS2-separated EAF revealed that the antibiotic activity was located in the region with an Rf value of 0.56 – 0.64. The Harrison research chromatotron technique also gave good separation of the EAF sample. Preparative HPLC was used as the final purification step for most of the EAF samples. Although, a number of peaks were observed during isocratic-HPLC (IHPLC) runs, they were not as clearly separated as those obtained with gradient-HPLC (GHPLC). Three major peaks PI, PII and PIII with elution times of 3.56 min, 4.53 min and 23.06 min respectively were revealed under GHPLC runs with decreasing concentrations (100% – 50%) of methanol in water. Methanol concentrations between 50% and 70% in water were considered the optimum GHPLC mobile phases. Since these chromatographic methods were all time consuming, required large volumes of solvents, and resulted in low yields of the antibiotic, an alternative procedure producing better results was sought. This led to the development of a procedure combining a three-solvent extraction system with a pH precipitation process which efficiently recovered the antibiotic in solid/crystal form. Using this procedure, sufficient quantities of the antibiotic were recovered from the fermentation broth to permit a degree of structural elucidation. Two types of crystals (brown and pink-yellow in colour) were obtained and their chemical natures established by means of 1H- and GCOSY- nuclear magnetic resonance (NMR) and liquid chromatography-mass spectrometry (LC-MS). On further LC-MS analysis, the brown crystals appeared to be a protein and since it did not show inhibitory activity against any of the test organisms, no further studies were carried out on it. The pink-yellow crystals when suspended in a minimal volume of methanol showed inhibitory activity against S. marcescens confirming that the antibiotic activity resided therein. The LC-MS spectrum of these crystals showed a prominent/base peak at 304.2724 [mass to charge ratio (m/z) in positive mode]. The elemental composition of this compound suggests a molecular formula close to C16H36N2O3 with a molar mass of 304.4686 g/mol. No existing name could be assigned to it from the database of known natural compounds. Hence, the possibility that it is a novel antimicrobial compound cannot be excluded. Characterisation of the antimicrobial substance using GC-MS revealed that it contained at least seven components (A – G). These components were then subjected to mass spectrum analysis and their retention indices compared to computer database listings of known compounds. Components A and B were regarded as representing one compound (possibly isomers) since they have the same molecular weight and formula. Their different retention indices strongly suggest they are indeed isomers. Thus a total of six different compounds were detected in the extract by GC-MS and the molecular formulae assigned to them include: C6H10O (A and B); C6H12O2 (C); C9H14O (D); C8H7N (E); C21H44 (F); and C12H14N2O (G). Since only low probability matches were obtained for A – F and as the sample could not be recovered from the analyser, they were not studied further. The closest match (71% probability) with substances listed in the computer database of natural compounds was for compound G (C12H14N2O) which was thus provisionally identified as N-acetyltryptamine. A structurally related compound known as melatonin is attributed with the ability to inhibit tumour growth in vivo and in vitro. Attempts were made to assign a chemical structure to the antibiotic produced by strain N8 using all the data available. The indications are that it is a tryptamine, the chemical structure of which is postulated to be: In order to monitor the antimicrobial activity of the antibiotic produced by strain N8, bioassays were conducted after all major steps during the isolation and characterization processes. The antimicrobial activity of the pink-yellow crystals was confirmed on the test organisms used during the primary screening phase, namely, Escherichia coli, Pseudomonas fluorescens, Serratia marcescens, Staphylococcus aureus, Enterococcus faecalis and Xanthomonas campestris pv. campestris, and the yeast Candida utilis, indicating that the crude substance had maintained its inhibitory activity against Gram-positive and Gram-negative bacteria, and the yeast tested. The study was extended to include investigations into the use of combinations of the GHPLC separated peaks of the antibiotic (PI, PII and PIII) to improve the efficacy of growth inhibition of the test pathogens for possible use in chemotherapy. Data from these studies showed that PI inhibited the growth of E. coli and X. campestris pv. campestris while PII and PIII inhibited the growth of the latter organism and also that of S. marcescens. Individually, the peaks showed no growth inhibition on Pseudomonas fluorescens but the combination PI+PII+PIII was antimicrobially effective. In all cases, the use of combinations was significantly more effective than the use of any single component alone. For example, the combination of GHPLC PI and PII had a greater growth inhibitory effect (synergic action) against Serratia marcescens than did either alone; the inhibition-zone diameter being double (30mm) that caused by the single peaks (15mm) against S. marcescens. Likewise mixing PI and PIII resulted in a much improved action against X. campestris pv. campestris. These findings may meet the current call by many scientists that all infectious diseases should be treated with a combination of two antibiotics with different mechanisms of action in order to counter the serious problem of emerging bacterial resistance. Since the antibiotic isolated during this study showed activity against both mammalian and plant pathogenic bacteria it is hoped that this work will encourage further investigation in this field in South Africa. The results obtained should impact on the pharmaceutical industry as well as agriculture and will, hopefully, help curb both plant and human infectious diseases in our African communities. This study also confirmed that KwaZulu-Natal soils do harbour rare actinomycetes that produce novel antimicrobial compounds. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2010.

The search for new antibiotics continues in a rather overlooked hunting ground. In the course of screening for new antibiotic-producing microorganisms, seventy-nine isolates showing antimicrobial activity were isolated from soil samples from various habitats in the KwaZulu-Natal midlands, South Africa. Existing methods of screening for antibiotic producers together with some novel procedures were reviewed. Both modified agar-streak and agar-plug methods were used in the primary screens. The use of selective isolation media, with or without antibiotic incorporation and/or heat pretreatment, enhanced the development of certain actinomycete colonies on the isolation plates. Winogradsky's nitrite medium (Winogradsky, 1949), M3 agar (Rowbotham and Cross, 1977), and Kosmachev's medium (Kosmachev, 1960), were found to be selective for actinomycetes. Statistical analysis showed highly significant interactions between isolates, assay media and the test organisms. The diameters of inhibition zones were found to be larger on Iso-sensitest agar (ISTA)[Oxoid, England] than in nutrient agar plates. Of the 79 isolates that showed antimicrobial activity, 44 isolates were selected for confirmatory screening. Of these, 13 were selected for secondary screening. Criteria for selection were based on significant inhibition of at least two test organisms and/or the inhibition of the specifically targeted organisms, Pseudomonas and Xanthomonas species. Following secondary screening eight isolates were considered for further investigation. The isolates were tentatively identified . on the basis of morphological features, using both light microscopy and scanning electron microscopy(SEM); their ability to utilize various carbon sources; and selected physiological and staining tests. Suspected actinomycetes were further characterized on the basis of selected chemical properties using thin layer chromatography (TLC) and high pressure liquid chromatography (HPLC) techniques. High pressure liquid chromatography analysis (Beckman 6300 analyzer) detected the presence of diaminopimelic acid (DAP) in whole-cell hydrolysates of six of the isolates while TLC analysis confirmed the type ofDAP present. The isolates N2, N12, N16, N19 and N35 were tentatively identified as Thermomonospora, Saccharopolyspora, Nocardiodes, Corynebacterium and Promicromonospora, respectively. Isolate N30 was identified as belonging to the coryneform group ofbacteria, possibly an Arthrobacter species. Isolate, N8, tentatively identified as Actinosynnema, was unique among the isolates tested as it showed good antimicrobial activity against all the Gram- positive and Gram-negative bacteria, and yeasts used as test organisms in the present investigation. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 2001.