This thesis represents a continuation of the work on the isolation and structure elucidation of potential drug leads from terrestrial fungal sources that the natural products group at the University of Canterbury has been engaged in. Capillary NMR spectroscopy was involved in the research as the main tool for dereplication and elucidating the structures of novel bioactive compounds as well as for biosynthetic studies. Eleven new compounds including five cyclic peptides, four related pyrones and two diketopiperazines were isolated from the extract of Aspergillus sp. of endophyte collected from Malaysia. The five peptides F8268-A-1 to F8268-A-5 showed excellent P388 (HCT116 (ATCC CCL-247) and human breast cancer, MCF7 (ATCC HTB-22)) activities. Two of the peptides F8268-A-3 and F8268-A-5 were 4,000 times more active when compared with commercial drugs (fluorouracil, cisplatin and tamoxifen). The partial stereochemistries of F8268-A-2 and F8268-A-3 were established by Marfey’s method. Four related pyrones isolated from the same extract were also shown to have good P388 activities. They are related to the known compound NF00659A3. The relative stereochemistries were established from NOSEY experiments and the energy-minimised (MM2) model created using CHEM 3D software. Two new diketopiperazines, F7474-D3 and F7474-D11, also isolated from the Aspergillus extract did not show activity in the P388 assay. F7474-D11 contained the amino acid Me-kynurenine which is the first report of this from a natural source. The absolute stereochemistry of F7474-D11 was elucidated by Marfey’s method. The other diketopiperazine F7474-D3 was similar to the known compound lumpidin, and combined use of ROESY NMR and Marfey’s method established that the constituent amino acids had the unusual R configuration. Dereplication has been greatly improved by the application of capillary NMR. For example, the HPLC analysis and UV library searching of compounds from extracts F8095 and F7855 suggested they contained related compounds belonging to the lasiodiplodin family. However, CapNMR spectroscopic analysis and AntiMarin database searching revealed that the compounds from F8095 were all known polyesters while the compounds from F7855 did belong to the lasiodiplodin family. Two new lasiodiplodins were found in the F7855 extract, (3R,4R)-4-hydroxy-de-O-methyl-lasiodiplodin (F7855-4) and (E)-9-etheno-de-O-methyl-lasiodiplodin (F7855-6). The relative stereochemistries were elucidated from NMR coupling constant analyses. Two new dimers (F7090-A and F7090-B) were elucidated from a New Zealand fungal endophyte. The differences between these two dimers was their stereochemistries. F7090-A had the same stereochemistries for the three stereocentres in both parts, while the stereochemistry of F7090-B was different in the two parts of the dimer. Biosynthetic studies were also carried out using CapNMR methodology. A known compound tetrahydrofuran A and a new compound tetrahydrofuran B from an unidentified New Zealand fungus were used for this study. For the first time an INADEQUATE NMR experiment was successfully carried out using CapNMR spectroscopy, thus demonstrating the capability for carrying out biosynthetic studies on a very small scale (<200 μg of ¹³C-labelled compound). The implementation of efficient dereplication procedures with CapNMR methodology is paramount for those working in the field of natural products. The recent advances that have been made in the dereplication process in the natural products group at the University of Canterbury are given using examples from this research and where necessary from other group members.

Actinomycetes and fungi were successfully isolated from both New Zealand and Malaysian marine invertebrates and classified as facultatively marine based on their ability to grow on both sea water and non-sea water media. Most of the extracts obtained from selected isolates were cytotoxic. A clear preference of the actinomycetes for solid-state fermentation was observed, however, for fungi no significant preference was seen. Three isolates of Streptomyces spp., four Penicillium spp. and two Paecilomyces spp. whose extracts showed good cytotoxicity were selected for further investigation. A small-scale extract obtained from a solid culture of Streptomyces sp. (LA3L2) showed good cytotoxicity and a new cytotoxic metabolite was isolated from a large-scale extract of Streptomyces sp. (LA3L2). This metabolite was characterized as S-methyl 2,4-dihydroxy-6-isopropyl-3,5-dimethylbenzothioate (5.15) and is only the third compound reported to contain the S-methyl benzothioate group. Two known compounds, montagnetol (5.16) and erythrin (5.18), were isolated from a further large-scale cultivation of Streptomyces sp. (LA3L2) and is the first reported actinomycete to produce these lichen-related compounds. In addition, two known inactive metabolites (bohemamine (5.1) and bohemamine B (5.2)) were identified from the small-scale extract. Streptomyces sp. (LA3L2) was also investigated for the effect of temperature and salinity on growth and cytotoxicity and shown to produce bohemamine only at 20 - 28℃ and 4% sea salt concentration on solid media. This isolate gave a low yield of active metabolite under all conditions. Small-scale extracts of two other Streptomyces spp. yielded three known cytotoxic metabolites. These were thiazostatin B (7.14) from Streptomyces sp. (LA5L4) and chromomycin A2 (7.1), chromomycin A3 (7.2) and chromomycin 02-3D (7.3) from Streptomyces sp. (LA3L1). All four Penicillium spp. produced known metabolites. Penicillium sp. (LY1L5) yielded two known metabolites, cycloaspeptide A (7.4) and α-cyclopiazonic acid (7.5). α-Cyclopiazonic acid (7.5) and three other known metabolites (roquefortine A (7.6), cyclopeptin (7.7) and viridicatin (7.8)) were isolated from Penicillum sp. (KK3T23). Penicillium sp. (KK3T8) produced brefeldin A (7.10), while mycophenolic acid (7.12) and brevianamide A (7.11) were produced by Penicillium sp. (KK4T14b). The effect of salinity on growth and cytotoxicity was investigated for the two Penicillium isolates producing the cytotoxic metabolite, α-cyclopiazonic acid (7.5). Saline conditions were not required for growth but metabolite production differed between the two isolates with respect to salinity. Isolate LY1L5 required saline conditions for α-cyclopiazonic production whereas isolate KK3T23 produced the metabolite under non-saline conditions and in concentrations of sea salt up to 6%. Three known compounds, indole-3-carboxylic acid (7.15), indole-3-carboxylate (7.17) and 5-carboxymellein (7.16) were identified from Paecilomyces sp. (PR5L9). Investigation of a small-scale extract obtained from a solid culture of another Paecilomyces sp. (PR10T2) resulted in the isolation and characterization of a unique structure of a symmetrical cyclic depsipeptide, epi-angolide (NAM 6-1). NAM 6-1 was considered as a new compound based on four homoisomeric configurations (A1, A2, A3 and A4). The value of dereplication procedures with respect to the rapid identification of metabolites and enhancement of in-house metabolite libraries is discussed. Structural elucidation of nine known metabolites (7.1, 7.2, 7.3, 7.5, 7.6, 7.7, 7.8, 7.10 and 7.11) was greatly aided by the in-house dereplication techniques using LC-MS-UV and AntiMarin database. A significant advantage was gained by the use of the CapNMR which enabled NMR characterization of very small quantities of metabolites (<20 µg). Approximately <5 µg of materials were required to perform 1D proton NMR experiments for the dereplication of seven known compounds; bohemamine (5.1), bohemamine B (5.2), thiazostatin B (7.14), indole-3-carboxylate (7.17) and 5-carboxymellein (7.16). Approximately 20 µg of materials were needed to acquire 1D and 2D (HSQC, HMBC and NOE) NMR spectra for structural elucidation of the new metabolite, S-methyl 2,4-dihydroxy-6-isopropyl-3,5-dimethylbenzothioate (5.15). Some 8 µg of materials were sufficient to perform 1D and 2D (COSY, HSQC and HMBC) NMR experiments for complete structural characterization of two known metabolites, montagnetol (5.16) and erythrin (5.18). Approximately 10 µg of materials were needed to acquire 1D and 2D NMR (COSY, HSQC and HMBC) experiments for structural elucidation of the new compound, epi-angolide NAM 6-1 (A1, A2, A3 and A4). Rapid identification of known fungal metabolites enabled the in-house HPLC-UV/Rt library to be enhanced by eight metabolites (7.5, 7.6, 7.7, 7.8, 7.10, 7.11, 7.17 and 7.16). An HPLC-UV/Rt library for actinomycete metabolites was successfully established with the insertion of eight known metabolites (5.1, 5.2, 5.16, 5.18, 7.1, 7.2, 7.3 and 7.14).

Actinomycetes and fungi were successfully isolated from both New Zealand and Malaysian marine invertebrates and classified as facultatively marine based on their ability to grow on both sea water and non-sea water media. Most of the extracts obtained from selected isolates were cytotoxic. A clear preference of the actinomycetes for solid-state fermentation was observed, however, for fungi no significant preference was seen. Three isolates of Streptomyces spp., four Penicillium spp. and two Paecilomyces spp. whose extracts showed good cytotoxicity were selected for further investigation. A small-scale extract obtained from a solid culture of Streptomyces sp. (LA3L2) showed good cytotoxicity and a new cytotoxic metabolite was isolated from a large-scale extract of Streptomyces sp. (LA3L2). This metabolite was characterized as S-methyl 2,4-dihydroxy-6-isopropyl-3,5-dimethylbenzothioate (5.15) and is only the third compound reported to contain the S-methyl benzothioate group. Two known compounds, montagnetol (5.16) and erythrin (5.18), were isolated from a further large-scale cultivation of Streptomyces sp. (LA3L2) and is the first reported actinomycete to produce these lichen-related compounds. In addition, two known inactive metabolites (bohemamine (5.1) and bohemamine B (5.2)) were identified from the small-scale extract. Streptomyces sp. (LA3L2) was also investigated for the effect of temperature and salinity on growth and cytotoxicity and shown to produce bohemamine only at 20 - 28℃ and 4% sea salt concentration on solid media. This isolate gave a low yield of active metabolite under all conditions. Small-scale extracts of two other Streptomyces spp. yielded three known cytotoxic metabolites. These were thiazostatin B (7.14) from Streptomyces sp. (LA5L4) and chromomycin A2 (7.1), chromomycin A3 (7.2) and chromomycin 02-3D (7.3) from Streptomyces sp. (LA3L1). All four Penicillium spp. produced known metabolites. Penicillium sp. (LY1L5) yielded two known metabolites, cycloaspeptide A (7.4) and α-cyclopiazonic acid (7.5). α-Cyclopiazonic acid (7.5) and three other known metabolites (roquefortine A (7.6), cyclopeptin (7.7) and viridicatin (7.8)) were isolated from Penicillum sp. (KK3T23). Penicillium sp. (KK3T8) produced brefeldin A (7.10), while mycophenolic acid (7.12) and brevianamide A (7.11) were produced by Penicillium sp. (KK4T14b). The effect of salinity on growth and cytotoxicity was investigated for the two Penicillium isolates producing the cytotoxic metabolite, α-cyclopiazonic acid (7.5). Saline conditions were not required for growth but metabolite production differed between the two isolates with respect to salinity. Isolate LY1L5 required saline conditions for α-cyclopiazonic production whereas isolate KK3T23 produced the metabolite under non-saline conditions and in concentrations of sea salt up to 6%. Three known compounds, indole-3-carboxylic acid (7.15), indole-3-carboxylate (7.17) and 5-carboxymellein (7.16) were identified from Paecilomyces sp. (PR5L9). Investigation of a small-scale extract obtained from a solid culture of another Paecilomyces sp. (PR10T2) resulted in the isolation and characterization of a unique structure of a symmetrical cyclic depsipeptide, epi-angolide (NAM 6-1). NAM 6-1 was considered as a new compound based on four homoisomeric configurations (A1, A2, A3 and A4). The value of dereplication procedures with respect to the rapid identification of metabolites and enhancement of in-house metabolite libraries is discussed. Structural elucidation of nine known metabolites (7.1, 7.2, 7.3, 7.5, 7.6, 7.7, 7.8, 7.10 and 7.11) was greatly aided by the in-house dereplication techniques using LC-MS-UV and AntiMarin database. A significant advantage was gained by the use of the CapNMR which enabled NMR characterization of very small quantities of metabolites (<20 µg). Approximately <5 µg of materials were required to perform 1D proton NMR experiments for the dereplication of seven known compounds; bohemamine (5.1), bohemamine B (5.2), thiazostatin B (7.14), indole-3-carboxylate (7.17) and 5-carboxymellein (7.16). Approximately 20 µg of materials were needed to acquire 1D and 2D (HSQC, HMBC and NOE) NMR spectra for structural elucidation of the new metabolite, S-methyl 2,4-dihydroxy-6-isopropyl-3,5-dimethylbenzothioate (5.15). Some 8 µg of materials were sufficient to perform 1D and 2D (COSY, HSQC and HMBC) NMR experiments for complete structural characterization of two known metabolites, montagnetol (5.16) and erythrin (5.18). Approximately 10 µg of materials were needed to acquire 1D and 2D NMR (COSY, HSQC and HMBC) experiments for structural elucidation of the new compound, epi-angolide NAM 6-1 (A1, A2, A3 and A4). Rapid identification of known fungal metabolites enabled the in-house HPLC-UV/Rt library to be enhanced by eight metabolites (7.5, 7.6, 7.7, 7.8, 7.10, 7.11, 7.17 and 7.16). An HPLC-UV/Rt library for actinomycete metabolites was successfully established with the insertion of eight known metabolites (5.1, 5.2, 5.16, 5.18, 7.1, 7.2, 7.3 and 7.14).