WO2008001306A2 - Streptomycete and bioactive compound produced thereby - Google Patents

Abstract

The present invention relates to an isolated microorganism of the strain type SPRT (=DSM 44925T = NRRL B-24448T). The microorganism preferably includes genetic material of GenBank accession number DQ141528 (16S-rDNA-sequence). The invention extends to a method of producing 2,5-diphenyloxazole or variants or derivatives thereof, the method comprising recovering the 2,5-diphenyloxazole or variants or derivatives thereof from a microorganism of the strain typeSPRT (=DSM44925T = NRRL B-24448T).

Description

STREPTOMYCETE AND BIOACTIVE COMPOUND PRODUCED THEREBY

BACKGROUND OF THE INVENTION

This invention relates to an isolated organism and to compounds recovered therefrom. In particular this invention relates to 'Streptomyces polyantibioticus' strain SPR and 21-57 being an antibacterial compound produced by the organism. The invention further relates to the use of 21-57 as an antibacterial agent.

Tuberculosis (TB) has become the leading cause of natural death in South Africa, killing 1000 people/month, most of whom are in their most economically productive years. South Africa is one of the 22 most burdened countries in the world, with an average incidence rate of 536 TB cases per 100 000 people in the population - the third highest incidence in the world (Global Tuberculosis Control, WHO Report, 2005). To further compound the problem, it is estimated that 61% of all TB-positive cases are attributable to patients that are also HIV-positive (50% in 2003; Global Tuberculosis Control, WHO Report, 2005).

According to Kim et a/. (2003), the family Streptomycetaceae contains the genera Streptomyces, Kitasatospora and Streptacidiphilus. All of these genera are phylogenetically related, but differ in chemical composition and in the formation of distinct phylogenetic clades in phylogenetic trees. The genus Kitasatospora (initially Kitasatosporia) was first proposed in 1982 by Omura et al. and was subsequently included in the genus Streptomyces (Groth et al., 2003). It was re-established in 1997 (Zhang et al.) and has subsequently been recognised as a genus separate from the genus Streptomyces.

The genus Streptomyces was first described in 1943 by Henrici & Waksman. After a period of classification and re-classification of genera such as Actinopycnidium, Actinosporangium, Chainia, Elytrosporangium, Microellobosporia, Kitasatoa and Streptoverticillium (Anderson & Wellington, 2001), the genus Streptomyces today consists of 519 validly published species (Euzeby, 2005). The systematics of the genus Streptomyces at species level is, however, still in a state of confusion and the genus is believed to be overclassified - by far having the highest amount of validly described species (Groth et al., 1999).

The genus Streptomyces has long been known as a rich source of antibiotics. Members of the genus Streptomyces are aerobic, Gram positive and have a DNA G+C mol% of 69 - 78%, the highest reported for actinomycetes (Anderson & Wellington, 2001). They produce extensively branching substrate and aerial mycelia; sporangia are not produced. With age, the aerial mycelium differentiates into chains of spores. Spore chain morphology and spore surface ornamentation are often used as distinguishing features amongst members of this genus. The genus is characterized as having a cell wall chemotype I (Lechevalier & Lechevalier, 1970) - LL-DAP and glycine are present in their peptidoglycan, and no characteristic sugars are present in whole-cell sugar hydrolysates (Anderson & Wellington, 2001; Lechevalier & Lechevalier, 1970). Saturated iso- and ante/so-fatty acids, hexa- or octahydrogenated menaquinones with nine isoprene units, and a phospholipid pattern of diphosphatidylglycerol, phosphatidylethanolamine, phosphatidylinositol and phosphatidylinositol mannosides, are characteristic of this genus (Locci, 1989). Furthermore, streptomycetes have an A3γ peptidoglycan type and lack mycolic acids (Locci, 1989).

Most antibiotics are secondary metabolites with a typically low molecular weight (ranging from 150 - 5000 Da) and are chemically diverse. They could comprise only carbon and hydrogen atoms, but mostly they contain carbon, hydrogen, oxygen and nitrogen atoms, while some also contain sulphur, phosphorus or halogens. In addition, almost all organic chemical functional groups are represented (Lancini et al., 1995). The role of secondary metabolites in bacteria has been disputed for a long time (Berdy, 2005). In actinomycetes, they are typically produced after vegetative growth and when the producer is entering a phase of dormancy or spore formation (Challis & Hopwood, 2003). Various hypotheses exist to explain the role of antibiotics in the antibiotic producers. Some researchers believe that these compounds play no role at all and are simply by-products of bacterial metabolism (Lancini et al., 1995). Lenski & Riley (2002) proposed that the ability of actinomycetes to produce antibiotics can explain the competitive effects in soil resulting in the increase of some bacteria and the decrease of others, effectively driving biodiversity of soil bacteria. Another potential role is that antibiotics could assist in the regulation of certain metabolic processes or even play a role in communication between antibiotic-producing organisms (Lancini et al., 1995).

The search for novel antibiotics has increased over the past few years after a period when pharmaceutical companies focussed more on anticancer agents or compounds used in the treatment of cardio-vascular diseases. The need for novel antibiotics with novel mechanisms of action has become urgent with the increased occurrence of bacterial resistance to known antibiotics and the emergence of new and old diseases (especially in the case of tuberculosis) (Berdy, 2005; Sandvoss et al., 2001). Berdy (2005) believes that microorganisms (especially actinomycetes) have a great ability to produce a vast array of structurally new molecules with new mechanisms of action which can be used either directly or as templates for drug design. The basis for his belief is the fact that natural compounds have already been pre-screened by nature and through evolution have been selected for their usefulness.

According to Lancini et al. (1995), the process of isolating, purifying and identifying novel compounds involves a range of professionals: microbiologists, fermentation technologists, bioengineers and chemists, and encompasses all the processes involved from the isolation of the bacterium from a soil sample to a vial containing the purified compound. Due to the high re-isolation incidence of known antibiotics, it is essential to determine as soon as possible whether or not a compound is novel (Lancini et al., 1995). The identification of antibiotics from fermentation media is a labour-intensive and time-consuming process (Berdy, 2005; Sandvoss et al., 2001; Wilson, 2000). The extraction of crude samples with organic solvents results in the extraction of a range of natural products and therefore requires further purification steps to obtain the pure compound of interest. The pure compound is then submitted for structural elucidation to determine whether the compound has been isolated before (Sandvoss et al., 2001). Most companies involved in high-throughput screening programs therefore make use of multiple-hyphenated systems, e.g. high performance liquid chromatography coupled to nuclear magnetic resonance spectroscopy coupled to mass spectrometry (HPLC-NMR-MS). Data obtained from this rapid screening technique are compared to databases with information on known compounds, resulting in a quick determination of whether or not a compound is novel (Sandvoss et al., 2001 ; Wilson, 2000). This type of specialised equipment is, however, expensive and limited to large research facilities or pharmaceutical companies.

A need exists for antibacterial agents useful in the treatment of bacterial infections and microorganisms useful in the production of such agents.

SUMMARY OF THE INVENTION

According to a first aspect to the present invention there is provided an isolated microorganism of the strain type SPR^T(= DSM 44925^T = NRRL B-24448^1").

Preferably the microorganism is an actinomycete strain. More preferably the microorganism is from the genus Streptomyces, The microorganism may be from the family Streptomycetaceae and is most preferably Streptomyces polyantibioticus.

The microorganism preferably includes genetic material of GenBank accession number DQ141528 (16S-rDNA-sequence) (http://www.ncbi.nlm.nih.gov).

According to a second aspect to the present invention there is provided a biologically pure culture having identifying characteristics of GenBank accession number DQ141528 (16S-rDNA-sequence).

According to a third aspect to the present invention there is provided the actinomycete strain SPR^T(= DSM 44925^T = NRRL B-24448^T).

According to a fourth aspect to the present invention there is provided a method of producing 2,5-diphenyloxazole or variants or derivatives thereof, the method comprising recovering the 2,5-diphenyloxazole or variants or derivatives thereof from a microorganism of the strain type SPR^T(= DSM 44925^T = NRRL B-24448⁷). According to a fifth aspect to the present invention there is provided 2,5- diphenyloxazole or variants or derivatives thereof recovered from a microorganism having identifying characteristics of GenBank accession number DQ141528 (16S- rDNA-sequence).

According to a sixth aspect to the present invention there is provided the use of 2,5- diphenyloxazole or variants or derivatives thereof as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

According to a seventh aspect to the present invention there is provided a composition for use as an antimicrobial agent, the composition comprising an effective amount of 2,5-diphenyloxazole or variants or derivatives thereof.

According to a eighth aspect to the present invention there is provided 2,5- diphenyloxazole or variants or derivatives thereof for use as an antimicrobial agent.

According to a ninth aspect to the present invention there is provided 2,5- diphenyloxazole or variants or derivatives thereof for use in the manufacture of a medicament for use as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

According to a tenth aspect to the present invention there is provided the use of 2,5- diphenyloxazole or variants or derivatives thereof in the manufacture of a medicament for use as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

According to an eleventh aspect to the present invention there is provided an isolated polynucleotide of GenBank accession number DQ141528 (16S-rDNA-sequence).

According to a twelfth aspect to the present invention there is provided an isolated polynucleotide sequence comprising a polynecleotide of GenBank accession number DQ141528 (16S-rDNA-sequence).

The microorganism has been deposited in two international culture collections, namely, the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ, the German Culture Collection, http://www.dsmz.de) and the Actinobacterial Culture Collection of the Agricultural Research Service of the US Department of Agriculture (http://nrrl.ncaur.usda.gov).

DESCRIPTION OF PREFERRED EMBODIMENTS

In the description, reference will be made to the following figures in which:

Figure 1 is a Cryo-scanning electron micrograph of strain SPR^T grown on Middlebrook 7H9-glucose agar at 30⁰C for 90 days. Clustered sporangiophores bearing mature sporangia are clearly visible, with smooth, rod-shaped spores contained within the sporangia. The bar represents 1 μm.

Figure 2 is an unrooted 16S-rDNA phylogenetic tree obtained by the neighbour- joining method, showing the position of strain SPR^T among its phylogenetic neighbours and representatives of sporangium-producing genera. The 16S-rDNA sequence of Arthrobacter globiformis JCM 1332^T was used as an out-group. All sequences were edited to the longest common region (1409 bp). GenBank sequence accession numbers are given in parentheses. Numbers at the nodes show the % bootstrap values. Asterisks indicate the clades that were conserved when the neighbour-joining, minimum evolution and maximum parsimony methods were used in constructing the phylogenetic tree.

Figure 3 is an rRNA operon copy number of strain SPR^T. Lanes: 1 & 3 - genomic DNA digested with Sail; 2 - genomic DNA digested with Seal; 4 - genomic DNA digested with EcoRV; 5 - genomic DNA digested with SspBI; 6 - genomic DNA digested with SπaBI. Genomic DNA fragments were fractionated on a 0.7% agarose gel for 24 h, transferred to a nylon membrane and hybridized with the rDNA probes. Lanes 1 and 2 represent digested genomic DNA hybridized with the 16S-rDNA probe, and lanes 3 to 6 represent digested genomic DNA hybridized with the 23S- rDNA probe.

Figure 4 is schematic diagram of the isolation and purification of the bioactive compound, 21-57.

Figure 5 is bioautography against M. tuberculosis H37Rv. Kanamycin was used as a positive control. Amounts spotted are indicated. Areas of white indicate areas of inhibition of growth; darker areas represent areas where bacteria are alive. Figure 6 is a) the molecular structure of 21-57 as determined by X-ray crystallography (structure provided by Associate Professor Susan Bourne, X-ray Crystallography Unit, Department of Chemistry, University of Cape Town) and b) the molecular structure of 21-57 with IUPAC numbering which was followed for the assignment of NMR spectra data (Table 3).

Figure 7 is a schematic representation of the biosynthetic process involved in oxazole and thiazole production (Roy et al., 1999).

Figure 8 illustrates chemical structures of natural products which contain one or more oxazole moieties.

Figure 9 is a) the biosynthetic gene cluster and structure of yersiniabactin produced by Yersinia pestis; and b) the function of a truncated HMWP2 with the final products (hydroxyphenyl) thiazoline-cysteine adduct and (hydroxyphenyl) thiazoline carboxylate (Roy et al. 1999).

Figure 10 shows the gene sequence of "streptomyces polyantibioticus" strain SPR^T (SEQ ID NO:1).

As part of a screening program for antibiotic-producing actinomycetes, an actinomycete with unusual morphology was isolated from soil collected from the banks of the Umgeni River, Kwa-Zulu Natal, South Africa.

Strain SPR^T was isolated on Middlebrook 7H9 agar [Difco Laboratories; supplemented with 10mM glucose; albumin-dextrose (AD) supplement omitted] after incubation at 28°C for 7 days. Following isolation, strain SPR^T was maintained on Middlebrook 7H9-glucose agar.

Antimicrobial activity was determined by the sloppy-agar overlay technique. For this technique, the isolate was stab-inoculated into Middlebrook 7H9-glucose agar, Czapek Solution agar (CZ) (Atlas, 1993) and yeast extract-malt extract (ISP 2 - Shirling & Gottlieb, 1966) agar plates, in duplicate. Plates were incubated for 10 days at 30⁰C (the duplicate set at 37°C), and overlaid with 6 ml Luria sloppy agar (Sambrook et al., 1989) containing the test bacterium. Activity was tested against various Gram positive and Gram negative bacteria: Bacillus coagulans ATCC 7050^T, Enterococcus faecium vanA (vancomycin resistant), Enterococcus phoeniculicola JLB-1^T, a Micrococcus sp. (clinical isolate), Mycobacterium aurum A+ and a Streptococcus sp. (clinical isolate); weak activity was exhibited against Citrobacter braaki strain 90 (clinical isolate), Enterobacter cloacae strain 67 (clinical isolate), Escherichia coli ATCC 25922 and Klebsiella oxytoca strain K52 (clinical isolate; resistant to augmentin and cefuroxime).

Strain SPR^T was cultured in 500 ml Hacene's Medium (HM; Hacene & Lefebvre, 1995) for 10 days at 30⁰C on a rocking shaker. The culture was filtered through a coffee filter (House of Coffees, 1X4 sized filters). The mycelial mass of the isolate was extracted with methanol and activity determined by bioautography.

The morphological characteristics of strain SPR^T were determined using standard methods (Locci, 1989). The isolate was grown on Middlebrook 7H9-glucose agar for 90 days at 30⁰C and the morphology observed under a light microscope and by cryo- scanning electron microscopy.

Standard physiological tests were performed as described by Locci (1989). ISP media were prepared as described by Shirling & Gottlieb (1966). Antibiotic resistance was determined by incorporation of the antibiotics into Bennett's medium agar plates (Atlas, 1993) at the recommended concentrations (Locci, 1989). The concentrations of the non-standard test antibiotics were as follows: capreomycin (20 μg/ml), cefotaxime (100 μg/ml), D-cycloserine (50 μg/ml), kanamycin (10 μg/ml) and viomycin (8 μg/ml). Physiological characteristics were determined after growth at 3O⁰C (unless otherwise stated) for the recommended incubation periods. All carbon^' sources for carbon utilization tests were filter-sterilised and tested at the concentrations recommended by Locci (1989) and Shirling & Gottlieb (1966). Utilization of organic acids, acid production from carbohydrates and resistance to lysozyme were determined as described by Gordon et al. (1974).

Standard techniques were used for the determination of catalase and oxidase activity. Staining was performed using the standard Gram staining technique. Acid- fast staining (1% sulphuric acid used in decolourisation step) and acid-alcohol-fast staining (Ziehl-Neelsen) were performed using standard methods.

Freeze-dried cells used in chemotaxonomic tests were obtained from a 500 ml ISP 2 culture of strain SPR^T, which was cultivated on a rocking shaker at 30⁰C for 5 days. The DAP isomer and whole-cell sugar pattern were determined by the method of Hasegawa et al. (1983), with the exception that freeze-dried cells were used instead of colonies from agar plates.

The fatty acid methyl esters were prepared according to Chou et al. (1998) and were analysed by gas chromatography on a Zebron GC column, ZB-1 (0.25 mm by 30 m). The temperature of the column was fixed at 210⁰C and the temperature of the injector was programmed to increase from 150°C to 22O⁰C at a rate of 4°C per minute. The peaks were identified by comparison with fatty acid methyl ester standards (FAM Standard: MIX GLC-80; SEPULCO). Mycolic acids were isolated and analysed as described by Minnikin et al. (1975). Phospholipids were isolated as described by Minnikin et al. (1984). Menaquinones were analysed by the DSMZ Identification Service (Braunschweig, Germany) using an HPLC separation method. Phospholipids were analysed by two-dimensional thin-layer chromatography (TLC) on silica gel 60 F₂₅₄ (Merck) plates as described by Komagata & Suzuki (1987). The peptidoglycan type was determined as described by Komagata & Suzuki (1987) - analysis was performed by TLC. The base composition (G+C mol%) was determined in 1.0 X standard saline citrate (SSC) using the thermal denaturation method described by Mandel & Marmur (1968).

The 16S-rDNA was amplified by the polymerase chain reaction using the universal bacterial 16S-rDNA primer F1 (Cook & Meyers, 2003) and primer R6 5'- AAGGAGGTGITCCAICC-3' [modified from primer p1525r of Chun & Goodfellow (1995); I = inosine]. PCR conditions were as described by Cook & Meyers (2003). The amplified DNA was purified for sequencing using a QIAquick PCR purification kit (Qiagen). For phylogenetic analysis, reference strains were chosen from the BLAST (Altschul et al., 1997) results as well as from sporangiate genera. For the construction of phylogenetic trees, the software packages MEGA (Molecular Evolutionary Genetics Analysis) version 2.1 (Kumar et al., 2001) and CLUSTALX version 1.81 (Thompson et al., 1997) were used. Unrooted phylogenetic trees were constructed using the neighbour-joining (Saitou & Nei, 1987), minimum evolution and maximum parsimony methods (Takahashi & Nei, 2000), and evaluated by bootstrap resampling (1000 replications).

Genomic DNA was digested singly with restriction endonucleases: Sa/I and Seal for the determination of 16S-rRNA operon multiplicity; and Sa/I, EcoRV, SspBI and SnaBI for the determination of 23S-rRNA operon multiplicity. The digested genomic DNA was electrophoresed on a 0.7% agarose gel for 24 h. Copy numbers for both the 16S-rRNA gene and 23S-rRNA gene were determined by Southern hybridization analysis (Sambrook et a/., 1989). Digoxigenin-dUTP (Roche) labelled DNA probes were used. The probe for the determination of 16S-rRNA copy number was amplified from strain SPR^T using the universal bacterial primers F1 and R5 (Cook & Meyers, 2003), and the probe for the determination of the 23S-rRNA copy number was amplified using the primers described by Wang & Zhang (2000): forward primer 5'- CCGATGAAGGACGTGGGA-3' (positions 46 to 63*) and reverse primer 5'- ACCAGTGAGCTATTACGC-3' (positions 1212 to 1195^*). Amplified, unlabeled 16S- rDNA and 23S-rDNA were used as positive controls. Hybridization and visualization were performed as recommended in the Roche DIG manual (http://www.roche.com). * Streptomyces ambofaciens ATCC 23877^T 23S-rRNA gene numbering

From cryo-scanning electron microscopy (Figure 1), the presence of sporangia borne on clustered sporangiophores was clearly visible. Smooth, almond-shaped spores were visible within the sporangia. The sporangiophores were typically 11 - 18 μm in length and the sporangia had a diameter of 6 - 10 μm, with the spores being approximately 2 μm in length. The sporangiophores and sporangia had "lollipop" appearances and clustered in groups of four to six (Figure 1). When plates containing mature sporangia were flooded with water, no spore motility was observed under a light microscope.

A comparison of the chemotaxonomic characteristics of strain SPR^T with those of sporangiate genera and the phylogenetically related non-sporangiate genera, Streptomyces and Kitasatospora, are summarized in Table 1. LL-DAP and glycine were detected in amino acid analysis of strain SPR^T and the whole-cell sugar hydrolysate yielded galactose, glucose and traces of ribose and mannose, a similar whole-cell sugar pattern as observed for members of the genus Actinoalloteichus and the genus Actinomadura. Mycolic acids were not detected and strain SPR^T had an A3γ peptidoglycan type [presence of LL-DAP and glycine as determined by TLC - Schleifer & Kandler (1972)]. The predominant fatty acids were determined to be of the /so and ante/so-branched type (28% /-C16:0; 18% a/-C15:0; 15% C16:0; 14% ai- C17:0; 10% /-C15:0; 8% /-C17:0), similar to the pattern observed in the genera Oerskovia, Micropolyspora, Saccharopolyspora, Streptomyces, Glycomyces and Kitasatospora. Phosphatidylglycerol, phosphatidylethanolamine, diphosphatidylglycerol, phosphatidylinositol, phosphatidylinositol mannoside and two unknown phospholipids were detected by 2D TLC analysis. The predominant menaquinones were determined to be MK-9 (H₄) (60%) and MK-9 (H₆) (40%), with traces of MK-9 (H₂) (<3%), the same pattern observed for the genera Actinomadura and Spiήllospora. The G+C mol% was 74.4% (± 0.2%) when determined in 1.0 X SSC (in duplicate).

Table 1 : A comparison of the chemotaxonomic characteristics of strain SPR , sporangiate genera and the genus Streptomyces

Data taken from Goodfellow (1989) and Locci (1989); data on the genus Kitasatospora was taken from Chung et al. (1999) and Groth et al. (2003). a: DAP = 2,6-diaminopimelic acid b: Phospholipid pattern: PI = phosphatidylglycerol; PII = phosphatidylethanolamine; PIII = phosphatidylcholine (with phosphatidylethanolamine, phosphatidylmethylethanolamine and phosphatidylglycerol variable); PIV = phospholipids containing glucosamine (with phosphatidylethanolamine and phosphatidylmethylethanolamine variable) c: Fatty acid pattern: 2c = iso- and anfe/so-branched and saturated fatty acids; 3a = saturated, unsaturated, /so- (variable) and methyl-branched fatty acids; 3c = saturated, unsaturated, iso-, anteiso- (variable) and methyl-branched fatty acids

* PG = phosphatidylglycerol; PE = phosphatidylethanolamine; Pl = phosphatidylinositol; PIM = phosphatidylinositol mannosides; DPG = diphosphatidylglycerol

A 1459 bp 16S-rDNA sequence was obtained for strain SPR^T. A BLAST (Basic Local Alignment Search Tool) search showed 97% homology to members of the genus Streptomyces. A phylogenetic tree of streptomycete strains, representatives of the sporangiate genera, the type species of the genera Kitasatospora and Streptacidiphilus, and strain SPR^T (Figure 2) showed that strain SPR^T clustered with members of the genus Streptomyces.

From Figure 3, it is clear that strain SPR^T contains seven copies of the 16S-rRNA and 23S-rRNA genes. The only enzyme which did not give a clear result was EcoRV (lane 4). Most members of the genus Streptomyces have 6 copies of the rRNA operon {rrndb; Klappenbach et al., 2001), but Streptomyces venezuelae has seven copies (Klappenbach et al., 2000; La Farina et al., 1996). Some members of the Gamma proteobacteria, such as Escherichia coli, Shigella flexneri, Pseudomonas putida, Samonella spp. and Yersinia pestis are known to have seven copies of the rRNA operon (Acinas et al., 2004), with some strains of E. coli having up to 36 copies (Klappenbach et al., 2001). Other bacterial strains with seven copies of the rRNA operon, include Oceanobacillus iheyensis HTE831^T, Streptococcus agalactiae NEM316 and Streptococcus agalactiae 2603 V/R (Acinas et al., 2004). Klappenbach et al. (2001) showed that there is no phylogenetic relatedness in species with the same number of rRNA operons and that the copy number is dependent on the natural environment of the bacteria.

Copy numbers of four and more have been detected in soil actinomycetes, which is not surprising, since Klappenbach et al. (2000) showed that a high copy number infers an advantage in fluctuating environments, such as soil, where a high copy number would ensure the ability to quickly utilize newly introduced resources. The fact that strain SPR^T has seven copies of the rRNA operon only reflects the potential ability of the strain to outcompete other strains with a lower copy number in its natural environment, and in no way reflects its phylogenetic position within the class Actinobacteria.

Phylogenetically, strain SPR^T is clearly related to the genus Streptomyces, with homology values of 97% to validly described members of the genus. Members of the genus Kitasatospora used to be classified as streptomycetes due to their high degree of 16S-rRNA sequence similarity, even though they clearly have a different cell wall chemotype (Anderson & Wellington, 2001). With the introduction of new strains with similar chemotaxonomic characteristics, it soon became clear that the group formed a distinct phylogenetic clade and the genus Kitasatospora was resurrected (Zhang ef a/., 1997). Furthermore, the phylogenetic clustering of members of a sporangiate genus among members of a non-sporangiate genus has been observed before: members of the sporangiate genus Spirillospora cluster among members of the non- sporangiate genus, Actinomadura (Zhang ef a/., 2001).

Strain SPR^T clearly differs from members of the genus Streptomyces: morphologically in the production of sporangia, which resemble those produced by members of the genus Streptosporangium, and chemically in containing characteristic sugars in its whole-cell hydrolysate that are similar to the sugars observed for members of the genera Actinoalloteichus and Actinomadura. The rest of the chemical characteristics of strain SPR^T are similar to those of the genus Streptomyces, with the exception of the menaquinone pattern. The predominant menaquinones observed for strain SPR^T, were MK-9 (H₄₁H₆), which is distinct from most streptomycetes [MK-9 (H₆₁H₈)] and the same as the genus Actinomadura. The tetra-hydrogenated MK-9 menaquinone has been detected in some streptomycetes, including S. hebeiensis DSM 41837^T (Xu ef a/., 2004), S. scabrisporus NRRL B- 24202^T (Ping et a/., 2004), S. scopiformis A25^T (Li et a/., 2002), S. sodiiphilus CIP 107975^T (Li ef a/., 2005), S. thermocoprophilus B19^T (Kim ef a/., 2000) and S. yeochonensis NRRL B-24245^1* (Kim ef a/., 2004). S. scabrisporus NRRL B-24202^T, however, is the only streptomycete that has MK-9 (H₄) as one of its predominant menaquinones.

Based on the polyphasic taxonomic analysis described, strain SPR^T is proposed to be a member of the genus Streptomyces. Streptomyces polyantibioticus strain SPR^T (= DSM 44925^T = NRRL B-24448^1") is proposed as the type strain.

Description of Streptomyces polyantibioticus sp. nov.

Streptomyces polyantibioticus (po.ly.an.ti.bi.o'ti.cus. N. L. fern. adj. polyantibioticus - referring to the ability to produce numerous antibiotics).

Strain SPR^T forms brown substrate mycelium and fluffy, white aerial mycelium on inorganic salts-starch agar (ISP 4). A leathery, dark brown substrate mycelium forms on ISP 2, but no sporulation occurs. Brown substrate mycelium with white aerial mycelium forms on oatmeal agar (ISP 3). The colour of the substrate mycelium is not pH sensitive. Good growth is observed on Middlebrook 7H9-glucose agar. A dark brown diffusible pigment is produced on glycerol-asparagine agar (ISP 5) and melanin is produced on peptone-yeast extract-iron agar (ISP 6) and tyrosine agar (ISP 7).

Very strong antibiosis is exhibited against Bacillus coagulans ATCC 7050^T, Enterococcus faecium vanA (vancomycin resistant), Enterococcus phoeniculicola JLB-1^T, a Micrococcus sp. (clinical isolate), Mycobacterium aurum A+ and a Streptococcus sp.(clinical isolate); weak activity was exhibited against Citrobacter braaki strain 90 (clinical isolate), Enterobacter cloacae strain 67 (clinical isolate), Escherichia coli ATCC 25922 and Klebsiella oxytoca strain K52 (clinical isolate; resistant to augmentin and cefuroxime). The methanol extract from mycelial mass was analysed by thin layer chromatography on silica gel 60 F₂5₄ (Merck) plates using ethyl acetate:methanol (100:15, v/v) as the mobile phase. Five anti-M. aurum A+ activity spots were detected by bioautography.

Strain SPR^T grows in the presence of 0.1% 2-phenylethanol (but not 0.3%), 0.0001% crystal violet, 7% NaCI (but not 10%) and 0.1% phenol, but not in the presence of sodium azide (0.01%). Growth was also observed in the presence of lysozyme, capreomycin (20 μg/ml), cefotaxime (100 μg/ml), cephaloridine (100 μg/ml), D- cycloserine (50 μg/ml), penicillin G (10 i.u.) and viomycin (8 μg/ml), but not in the presence of gentamicin (100 μg/ml), kanamycin (10 μg/ml), lincomycin (100 μg/ml), neomycin (50 μg/ml), oleandomycin (100 μg/ml), rifampicin (50 μg/ml), streptomycin (100 μg/ml), tobramycin (50 μg/ml) and vancomycin (50 μg/ml). Growth was observed at 4°C, 30°C and pH 4.3, but not at 37°C and 45°C. Strain SPR^T uses DL- α-amino-n-butyric acid, L-arginine, L-cysteine, L-histidine, L-hydroxyproline, L- phenylalanine, L-serine, L-threonine, L-valine and potassium nitrate as sole nitrogen sources; L-methionine is weakly utilized. It uses D(-) fructose, D(-) lactose, D(+) cellobiose, D(+) galactose, D(+) glucose, D(+) mannose, D(+) melibiose, D(+) xylose, meso-inositol, raffinose, salicin, sodium acetate (0.1%), sodium citrate (0.1%) and trehalose as sole carbon sources (ribose is utilized weakly), but does not utilize adonitol, D(-) mannitol, D(+) melezitose, inulin, L(+) arabinose, L(+) rhamnose, sucrose and xylitol.

H₂S production occurs and nitrate is reduced (weakly). Lecithinase and lipase activity are observed on egg-yolk agar, but not protease activity. Pectin is hydrolysed, but hippurate is not. Strain SPR^T degrades adenine, aesculin, arbutin, casein, hypoxanthine, L-tyrosine, Tween 80, xanthine and DNA, but not allantoin, cellulose, gelatin, guanine, starch, urea and xylan. Acid production from adonitol, α-methyl-D- glucoside, L(+) arabinose, D(+) cellobiose, dulcitol, meso-erythritol, D(+) galactose, D(+) glucose, glycerol, meso-inositol, D(-) lactose, D(-) mannitol, D(+) mannose, D(+) melezitose, D(+) melibiose, raffinose, D(+) sorbitol, trehalose and D(+) xylose were not observed. Strain SPR^T is able to utilize the organic acids, sodium acetate, sodium citrate, sodium formate, sodium gluconate, sodium lactate, sodium malate, sodium succinate and sodium tartrate. Sodium butyrate is utilized weakly and sodium benzoate, sodium maleate, sodium mucate, sodium oxalate, sodium salicylate and sodium sorbate are not utilized.

The G+C content of the DNA is 74.4% (± 0.2%) (1.0 X SSC).

The type strain is strain SPR^T (= DSM 44925^T = NRRL B-24448^T).

In this study, an antibacterial agent was isolated, purified and designated 21-57. The activity of 21-57 was determined and its anti-tubercular activity prompted the use of classical structure determination methods, including nuclear magnetic resonance (NMR) spectroscopy, mass spectrometry (MS), infra-red spectroscopy (IR) and X-ray crystallography to elucidate the structure of the compound. In addition, other physico- chemical properties of the compound were determined which contributed to the description of the compound. The screening of chemical databases led to the identification of the compound.

Fermentation and isolation

¹S. polyantibioticus' strain SPR was maintained on Middlebrook 7H9-glucose agar. After inoculation, the plate was incubated for 10 days at 28°C before use.

Spores and mycelial mass were taken from 10 day old agar plates and heavy suspensions were made in nine Universal containers containing 5 ml sterile distilled water each. The suspensions were used to inoculate nine seed cultures of 100 ml Hacene's medium (HM; 5.0 g glucose, 4.0 g yeast extract powder, 10.0 g malt extract and 1.0 g sodium chloride per litre of distilled water, pH 7.0) (Hacene & Lefebvre, 1995) in 1 litre Erlenmeyer flasks. The flasks were incubated on a rocking shaker at 3O⁰C for 48 hours. The seed cultures were used to inoculate nine 5 litre Erlenmeyer flasks each containing 500 ml of HM. The flasks were incubated on rocking shakers at 30⁰C for a further 192 hours (8 days) for a total fermentation time of 10 days. Isolation and purification of 21-57 is outlined in Figure 4 (method adapted from Oki et al., 1979). After the confirmation of culture purity by standard Gram staining, the cultures were pooled and filtered through coffee filters (House of Coffees, 1x4 sized filters). The antibacterial compounds were extracted from the mycelial mass with methanol and from the culture filtrate with ethyl acetate by stirring overnight at room temperature on magnetic stirrers.

After concentration to 100 ml (by evaporation in a fumehood), the methanol and ethyl acetate extracts were combined, the pH was adjusted to 7.0 (monitored with pH strips, Merck) and the sample was re-extracted with 1 litre of toluene by stirring overnight on a magnetic stirrer at room temperature. The toluene extract was concentrated to 100 ml (by evaporation) and extracted (overnight, stirring at room temperature) with 1 litre of sodium acetate buffer (pH 3.5). The toluene layer was concentrated to 10 ml (by evaporation) and 2 ml was applied to a 40 cm x 4 cm (length by diameter) silica gel G (Merck) column. Toluene was used as the elution solvent and 250 5-ml fractions were collected in glass test tubes.

The activity of the fractions was detected through bioautography with Mycobacterium aurum A+ as the test organism. The active fractions were pooled, dried down and re- dissolved in methanol. White crystals precipitated out, leaving a yellow compound in solution. The methanol was removed and the crystals were allowed to dry. The crystals were re-dissolved in 2 ml toluene and subjected to another 40 cm x 4 cm silica gel G (Merck) column with toluene as the elution solvent for further purification. Once again, 5-ml fractions were collected (150 in total). Bioactivity was determined as before. Fractions 21 to 57 were combined, containing 110 mg of the purified bioactive compound, designated 21-57. The yield of 21-57 was 24.44 mg per litre of culture. Reagent grade organic solvents were used in all extractions.

Biological properties

All test bacteria were grown in LB (Luria-Bertani) broth except for Mycobacterium bovis BCG (Tokyo), Mycobacterium smegmatis LR222 and Mycobacterium tuberculosis H37Rv. The mycobacteria were grown in Middlebrook 7H9 broth supplemented with 0.05% (v/v) Tween 80 and one tenth volume AD supplement (5% BSA, 2% glucose solution, filter sterilized). The test bacteria used in this study were: the Gram positives - Bacillus coagulans ATCC 7050^T, Bacillus megaterium NCIB 2602, Enterococcus faecium VanA (vancomycin resistant), Enterococcus phoeniculicola JLB-1^T, a Micrococcus sp., Mycobacterium aurum A+ (fast-growing, non-pathogenic), Mycobacterium bovis BCG (Tokyo) (fast-growing, non-pathogenic vaccine strain), Mycobacterium smegmatis LR222 (fast-growing, non-pathogenic), Mycobacterium tuberculosis H37Rv (ATCC 27294; slow-growing, pathogenic), Staphylococcus aureus ATCC 25923 (quality control strain for antimicrobial testing), and a Streptococcus sp. (clinical isolate); the Gram negatives - Acinetobacter calcoaceticus C91, Citrobacter braaki 90 (clinical isolate), Enterobacter cloacae 67 (clinical isolate), Escherichia coli ATCC 25922 (strain used for antibiotic susceptibility testing), Escherichia coli ATCC 35218 (strain used for antibiotic susceptibility testing), Escherichia coli CA 84-39 (streptomycin resistant), Escherichia coli E4 (cephalosporin resistant), Klebsiella oxytoca K52 (resistant to augmentin and cefuroxime), Klebsiella pneumoniae K58 (resistant to cephalosporins and augmentin), Proteus mirabilis 87 (clinical isolate) and Pseudomonas aeruginosa ATCC 27853 (strain used for antibiotic susceptibility testing).

All test bacteria were incubated at 37°C, shaking overnight except the enterococci and M. aurum A+ which were incubated for two days. M. bovis BCG (Tokyo) and M. smegmatis LR222 were incubated standing at 37°C for 7 days with occasional manual shaking; M. tuberculosis H37Rv was incubated at 37°C for 9 days, standing with intermittent agitation in an incubator in a Biosafety Level 3 laboratory, Institute of Infectious Diseases and Molecular Medicine, University of Cape Town (all tests involving M. tuberculosis H37Rv were performed in this Biosafety Level 3 laboratory). After confirmation of the purity of the cultures through standard Gram staining (or Ziehl-Neelsen staining for M. tuberculosis H37Rv), the optical density of the test bacterial cultures were determined at 600 nm on a Beckman DU-64 spectrophotometer and adjusted to 0.5 with LB or Middlebrook 7H9 broth depending on the culture (except for M. tuberculosis H37Rv which was not diluted; OD ~ 0.6 by comparison with M. smegmatis LR222 cultures of known OD).

The antimicrobial spectrum of the purified compound was determined through bioautography. 21-57 is highly non-polar which made the determination of minimal inhibitory concentrations in liquid difficult and standard methods, such as the agar disk diffusion method (Kirby-Bauer method), could not be used in the determination of the antibacterial activity of 21-57. A range of microgram amounts (5 - 100μg) of the antibiotic were spotted onto silica gel 60 F₂₅₄ (Merck) plates in 5 μl volumes using Gilson pipettes and allowed to dry for 20 minutes in a fumehood. The same volume (5μl) of organic solvent was spotted as a negative control to ensure that it was not the organic solvent in which 21-57 was dissolved that was killing the bacteria.

Structure and physical properties

The melting point of 21-57 was determined with a Reichart-Jung Thermovar melting point apparatus fitted with a digital thermometer. The infra-red absorption spectrum was recorded as a chloroform solution with a Perkin-Elmer FT-IR spectrometer and the mass spectrum was determined using MALDI-TOF spectrometry (provided as a service by the Department of Molecular and Cell Biology, University of Cape Town) and was confirmed by high resolution mass spectrometry - service provided by the School of Chemistry, University of Witwatersrand (Johannesburg, South Africa). ¹H and ¹³C NMR spectra were recorded on a 600 MHz Bruker NMR spectrometer - service provided by the Department of Chemistry, University of Stellenbosch (South Africa). Chemical shifts are given in S values (ppm) with tetramethylsilane as an internal standard. All spectra were determined in CDCI₃ (deuterated chloroform) and CD₃CN (deuterated acetonitrile). Microanalysis was performed as a service by the Department of Chemistry, University of Cape Town. The UV/VIS absorption spectrum was determined with a Beckman DU-64 spectrophotometer (21-57 dissolved in methanol).

Crystals for X-ray crystallography were prepared by dissolving 10 mg of 21-57 in 400 μl of methanol. The container was covered with Parafilm and the Parafilm was perforated with a sterile needle to allow for the slow evaporation of the solvent. The sample was left at 4°C for 5 days.

Single crystal X-ray diffraction data were measured on a Nonius KappaCCD diffractometer using graphite-monochromated Mo Ka radiation (0.7107 A at 113K). In each case, a series of frames were recorded, each of width 1 ° in Φ or in ω (with K ≠ 0) to ensure completeness of the data collected to θ>27°. The unit cell was indexed from the first ten frames and positional data were refined along with diffractometer constants to give the final cell parameters. Integration and scaling [DENZO, Scalepack (Otwinowski & Minor, 2000)] resulted in a unique data set corrected for Lorentz-polarization effects and for the effects of crystal decay and absorption by a combination of averaging of equivalent reflections and an overall volume and scaling correction. The structure was solved using SHELXS-97 and refined in SHELXL-97 (Sheldrick, 1997) using full-matrix least squares methods with the aid of the program XSEED (Barbour, 2003).

All non-hydrogen atoms were modelled anisotropically, while all hydrogen atoms were assigned an isotropic thermal parameter 1.2 times that of their parent atom and refined using a 'riding' model. Images of crystal and molecular structures were rendered using the program POV-RAY (POV-RAY™ for Windows, 2003). (Details kindly provided by Associate Professor Susan Bourne of the X-ray Crystallography Unit, Department of Chemistry, University of Cape Town).

For thin-layer chromatography (TLC), silica gel 60 F₂₅₄ plates (Merck) were used. To analyse the purity of the compound, 21-57 was chromatographed on TLC plates with 100% toluene as the mobile phase and sprayed with 50% sulphuric acid or cerium (IV) ammonium sulphate [30 ml sulphuric acid added to 500 ml distilled water, 63.0 g of cerium (IV) ammonium sulphate added, and diluted to 1 litre with distilled water] and heated at 100⁰C for colour visualization. The compound was also detected with iodine vapours and UV irradiation at 254 nm. Visualization reagents for the presence of sugar groups [2 ml aniline (Saarchem), 3.3 g phthalic acid (Sigma), dissolved in 100 ml water-saturated n-butanol] and primary amino groups (0.1% ninhydrin in acetone, w/v) were also used.

21-57 (2,5-diphenyl-1,3-oxazole, PPO): White crystalline solid, m.p. 70 - 72°C (similar to DOSE, 2^nd Electronic Version, 2004); R_f = 0.47 (silica gel, 100% CHCI₃); IR(CHCI₃) v_max 3071.03, 3011. 51 , 1590.58, 1483.39, 1448.08, 1133.10, 1026.97cm^"1; ¹H NMR (600MHz, CD₃CN) δ 8.18 (dd, J = 7.7, 1.5 Hz, 2 H), 7.86 (dd, J = 8.2, 1.3 Hz, 2 H), 7.62 - 7.53 (m, 6 H), 7.44 (t, J = 7.4 Hz, 1 H); ¹³C NMR (600 MHz, CD₃CN) δ 161.15, 151.25, 130.31, 128.92, 128.80, 128.42, 128.02, 127.46, 126.27, 124.19, 123.44; HRMS calculated for C₁₅H₁₁NO [M + H⁺] 221.08406, found 221.08178 (similar to Nicolaou et al., 2004).

Antibacterial activity of 21-57

The minimum inhibitory amounts of the bioactive compound, 21-57, against various test bacteria are shown in Table 2. The compound seems to have a broad spectrum of activity, but no specific pattern of activity, as seen in its ability to inhibit the growth of some Gram positives and some Gram negatives. No activity was observed against the following bacteria: the Gram positives - Bacillus coagulans ATCC 7050^T, Enterococcus phoeniculicola JLB-1^T, Staphylococcus aureus ATCC 25923 and a Streptococcus sp. (clinical isolate); and the Gram negatives - Acinetobacter calcoaceticus C91 , Citrobacter braaki 90, Enterobacter cloacae 67, Escherichia coli CA 84-39 (streptomycin resistant), Escherichia coli E4 (resistant to cephalosporins), Klebsiella oxytoca K52 (β-lactamase producer), Klebsiella pneumoniae K58 (resistant to all cephalosporins and augmentin) and Proteus mirabilis 87.

The compound exhibited activity against three different mycobacteria of which M. bovis BCG (Tokyo) was the most sensitive, followed by M. tuberculosis H37Rv (bioautography results shown in Figure 5) and M. aurum A+.

Structure elucidation and physical properties

The physico-chemical properties of 21-57 are listed in Table 3. The compound was purified as a white crystalline solid and was more readily soluble in non-polar organic solvents (e.g. hexane, toluene, ethyl acetate and chloroform) than in polar solvents (e.g. ethanol and methanol) and insoluble in water. In order to characterise 21-57, reactions to various spray reagents on TLC plates were investigated. The compound showed no reaction to ninhydrin (test for the presence of primary amino groups) or to a sugar visualization reagent.

A positive reaction was observed with iodine vapours (presence of double bonds) and 50% sulphuric acid. Most compounds appear black after exposure to acidified cerium (IV) ammonium sulphate and heat. 21-57, however, appeared white, indicating that the compound was possibly already in an oxidized state. The compound also showed absorbance at 254 nm when the TLC plate was irradiated with UV rays at 254 nm.

Table 2: Minimum inhibitory amount (μg) of the bioactive compound, 21-57, as determined against various test bacteria

* This is the highest amount tested against this strain;

• 5μg was the lowest amount tested against all the test bacteria and ≤ 5 indicates that the bacteria were inhibited by this amount, but could be sensitive to lower amounts β >100 shows that the growth of the test bacteria was not inhibited at amounts up to 100μg

Table 3: Physico-chemical properties of 21-57

The melting point of the compound was 70 - 72⁰C and the UV/VIS spectrum of this compound showed strong absorbance at 303 nm and 216 nm. The molar extinction coefficients are indicated in Table 3 as determined in methanol. The IR spectrum indicated the presence of C=C-H (>3000 cm^"1), conjugated cyclic -C=N and C=C bond stretches (1400 - 1600 cm^"1), C-O (1000 - 1200 cm^"1) and aromatic C-H bends (700 - 900 cm^"1). The molecular formula of 21-57, C₁₅H₁₁NO, was determined from the microanalysis and high-resolution MS and MALDI-TOF MS results. The structure of 21-57 was elucidated from X-ray crystallography data and confirmed by ¹H NMR, ¹³C NMR, a DEPT experiment, and 2D NMR spectroscopy (¹H-¹H COSY, HMQC and HSQC).

The crystal structure of 21-57 is shown in Figure 6. Confirmation of the structure by X-ray crystallographic data allowed for a search of the NCBI Pubmed compound database, and 21-57 was identified as the known structure 2,5-diphenyloxazole.

The ¹H and ¹³C spectra and the DEPT experiment confirmed the presence of fifteen carbon atoms and eleven protons in 21-57, of which all the carbons are aromatic and four are quaternary carbons at δ 161.15 (C5), 151.25 (C2), 128.02 (C11) and 127.46 (C21). Assignment of the ¹H NMR and ¹³C NMR spectra is indicated in Table 4. The NMR data corresponded to the data obtained by Nicolaou et al. (2004) for 2,5- diphenyloxazole.

Table 4: NMR spectral data of the antibacterial agent 21-57 and assignments (all data collected at 600MHz in CD₃CN)

Abbreviations for signals: dd = doublet of a doublet; t = triplet; and m = multiplet

Compound 21-57 was isolated from the fermentation broth and mycelial mass of ¹S. polyantibioticus' strain SPR. The structural elucidation of the compound allowed for the identification of 21-57 as 2,5-diphenyloxazole (PPO). PPO is best known for its properties as a major component of scintillation fluid or as a luminophore (lonescu et al., 2005; Semenova et al., 2004; Hariharan et al., 1997; Agnew et al., 1995). It is widely used in laser dyes and as "labels and probes for biomedical assays" (Semenova et al., 2004). The production of this compound from a natural source, namely a bacterium, has never been reported. It is produced by synthesis in industry.

The full assignment of the carbon and proton NMR signals was carried out with reference to the crystal structure of 21-57 as shown in Figure 6. Initially all the NMR spectra were determined in CDCI₃, but some of the ¹H signals overlapped in such a way that made assignment difficult. When the NMR spectra were determined in CD₃CN, certain proton signals shifted in such a way as to allow further analysis. For descriptive purposes, the phenyl ring to the left of the oxazole ring was designated ring B, while the phenyl ring to the right was designated ring A. The NMR data corresponded well with that of Nicolaou et al. (2004), although a detailed assignment of signals has not been reported.

Within the oxazole ring, there are two quaternary carbons. The carbon at position C2 would be affected the most by the deshielding effect of the adjacent oxygen and nitrogen and would therefore have a chemical shift furthermost downfield of all the carbons. The carbon resonating at δ 161.15 was therefore assigned to C-2. The other quaternary carbon is at position C-5. This carbon is expected to have the second highest chemical shift downfield due to the deshielding effect of the oxygen in the oxazole ring. The carbon with chemical shift δ 151.25 was therefore assigned to this position.

From the ¹H-¹H COSY analysis, the proton at δ 7.62 (m) is not coupled to any of the other protons in 21-57. This proton is bonded to the carbon at δ 123.44 as determined from HSQC data. From the HMQC data, it was determined that this carbon has no long-range association with any other protons, but that the proton at δ 7.62 has long-range association with C-2 (δ 161.15) and C-5 (δ 151.25) the two quaternary carbons in the oxazole ring, thereby affirming the assignment of this carbon and proton to position 4. Of the other two quaternary carbons, the carbon at δ 128.02 is further downfield than the carbon at δ 127.46, suggesting that it is also influenced by the deshielding effect of the oxygen and nitrogen on the oxazole ring and can be assigned as the carbon within ring A at position C-1 '. The carbon at δ 127.46 can therefore be assigned as the carbon within ring B that is bonded to the oxazole ring (C-1"). The two protons resonating at δ 8.18 (dd) are shifted the furthest downfield of all the protons, suggesting that their chemical shift is influenced by the electron-withdrawing effect of the oxygen and nitrogen in the oxazole ring. Furthermore, although there is some fine structure in the signal, it is essentially a doublet of doublets with one large and one small coupling, suggesting ortho- and metø-coupling. The signal was therefore assigned to H-2' and H-6', with evidence for small long-range para- couplings. From the HSQC data, the two carbons at δ 126.27 are directly associated with protons H-2' and H-6', and the signal is therefore assigned to C-2' and C-6'. The ¹H-¹H COSY data also indicated that H-2' and H-6' are coupled to a complex signal centred at δ 7.59 and integrating for three protons. This three proton multiplet was therefore assigned to the remaining three protons H-3', H-4' and H-5' of ring A. From the HSQC spectrum, it was observed that this multiplet at δ 7.59 is correlated with two carbons at δ 128.80 and one other carbon at δ 130.31. Since C-3' and C-5' are in identical environments, these are assigned to the former, while C-4' is assigned to the latter.

The carbon at δ 127.46 was assigned to C-1" in ring B as mentioned above. From the HMQC spectrum, this carbon has long-range association with four protons, two with a chemical shift of <5 7.86 (dd) and two others with a chemical shift of δ 7.56 (m). As argued above for the signal at J 8.18, the signal at δ 7.86 can be assigned to H-2" and H-6" in ring B. From the HSQC spectrum, this signal is correlated with a carbon signal at δ 124.19, which is therefore assigned as C-2" and C-6" in ring B. The two protons at δ 7.56 are bonded to two carbons with chemical shift δ 128.92 (as seen from the HSQC spectrum). These carbons were assigned to C-3" and C-5". The remaining carbon at C-4" resonates at δ 128.42, which according to the HSQC spectrum is bonded to the proton resonating as a triplet at δ 7.35. Furthermore, the HMQC spectrum confirmed long-range association of H-4" and C-2", C-3", C-5" and C-6", confirming its presence and assignment in ring B.

Even though the two phenyl rings are structurally identical, they differ significantly in their chemical and magnetic environments. Ring A is influenced by the electron withdrawing effect of the oxygen and nitrogen of the oxazole ring, whereas ring B is relatively electron-rich, resulting in the unique chemical shifts observed. Furthermore, the fact that H-4 seems not to be long-range coupled to any of the protons in ring B suggests that although the molecule seems planar in crystal form, there might be a degree of rotation in solution around the C-5-C-1" bond, presumably to relieve some steric strain caused by the interaction of H-4 and H-2" in the planar conformation. To understand how strain SPR is able to synthesise 21-57, one must consider the components of the compound and how they are synthesised in nature. The core of 21-57 contains an oxazole moiety. Oxazoles are 5-membered aromatic heterocycles which contain both oxygen and nitrogen (Yeh, 2004). Oxazoles are produced by enzymatic post-translational modifications of peptide-based natural products. Non- ribosomal peptide synthetases are often involved (Walsh, 2004; Keating et al., 2000). As can be seen in Figure 7, the cyclodehydration of serine or threonine yields a dihydroheteroaromatic oxazoline. A two-electron oxidation of the oxazoline in turn, yields an oxazole as a core structure. Reduction of the carbon - nitrogen bond in oxazoline creates an oxazolidinone ring (Yeh, 2004; Roy et al., 1999; Milne et al., 1998). All three of these oxidation states can be found in natural products. If the starting block was cysteine, the corresponding process would yield thiazolines, thiazoles and thiazolidinones (Roy et al., 1999; Milne et al., 1998). The process of oxazole/thiazole production requires 5 ± 1 ATP/GTP molecules per oxazole/thiazole produced (Milne et al., 1998).

Naturally occurring oxazoles were initially considered to be rare until increased research in the 1980s proved their ubiquity in nature (Yeh, 2004). Noltemeyer et al. (1982) suggested that the screening for secondary metabolites in streptomycetes is often one-sided and that interesting molecules, such as oxazole derivatives, are often not purified because of their general lack of bioactivity, generating the perception that oxazoles are rare in nature.

In Table 5 and Figure 8, most of the known oxazole-containing natural bioactive compounds are listed and shown. The main reason why so many researchers are interested in oxazoles and thiazoles, is their ability to interact with a wide variety of intracellular bacterial targets, especially proteins, RNA and DNA (Walsh, 2004; Milne et al., 1998). These abilities are clear from the diverse biological properties of the oxazole-containing natural bioactive compounds in Table 5.

Something that is interesting to note, is the fact that many of these compounds have been isolated from marine sponges, sea squirts or sea plumes. Berdy (2005) cautions that many compounds isolated from marine invertebrates are often attributable to their microbial symbionts, especially if one considers that symbionts typically make up 40 to 60% of the biomass of marine organisms. Furthermore, a warning sign to look out for is the relatively low amounts isolated from the actual marine organisms, e.g. diazonamide A is often the focus of chemical synthesis because it is so scarce and not readily isolated from its known source (Berdy, 2005; Yeh, 2004).

Certain synthetic antibiotics also contain oxazole functional groups. Some belong to the well-known antibiotic class, β-lactams, e.g. the isoxazolylpenicillin oxacillin. Its side-chain containing the oxazole functional group provides steric hindrance to the binding of penicillinases. These penicillins have a stronger activity in vitro than that of methicillin, but due to their short half-lives in serum, their activity in vivo is similar to that of methicillin (Lancini et al., 1995).

Two sulphonamide drugs, sulfamethoxazole and sulfisoxazole, are derivatives of sulfanilamide, where one hydrogen of sulfanilamide is replaced by a methyloxazole and a dimethyloxazole, respectively (Prescott et al., 1996).

Oxazolines and oxazolidinones are also represented in antibacterials. Phenyloxazolines have the ability to inhibit the growth of certain Gram negatives by inhibiting lipid A production, a molecule that is found in the outer membrane of most Gram negatives (Jackman et al., 2000). The oxazolidinones were first identified in the 1980s (Tanitame et al., 2004). This new class of antibacterials is synthetic and acts during an early stage in protein synthesis - a novel mechanism of action. They typically target Gram positive bacteria, including sensitive and resistant strains of Mycobacterium tuberculosis, methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus faecium and penicillin-resistant Streptococcus pneumoniae (Sood et al., 2005; Sbardella et al., 2004). The most well-known oxazolidinones are linezolid and eperezolid.

Table 5: Natural compounds with oxazole functional groups, their origin and their biological function

Figure 8 contains some of the chemical structures of the compounds mentioned in Table 5. References for the chemical structures in Figure 8 are the same as the references indicated in Table 5.

According to Lancini et al. (1995) there are numerous steps to follow to determine the biosynthetic pathway of a new antibiotic, namely: 1. Identification of the primary metabolites involved; 2. Isolation and identification of any intermediates produced in the process; 3. Identification of any enzymes which would be capable of catalyzing these reactions; and 4. Identification of the biosynthetic gene cluster, including the sequence and organisation of the genes involved. Ideally these steps are followed as listed, but with genetics and recombinant DNA technology, step 4 is often accomplished first before the acquisition of biochemical evidence (Lancini et al., 1995).

Roy et al. (1999) discussed the biosynthesis of yersiniabactin, a siderophore produced by Yersinia pestis. Synthesis is typically achieved via a non-ribosomal peptide synthetase mechanism. The biosynthetic gene cluster consists of two catalytic subunits, high molecular weight protein 2 and 1 (HMWP2 and HMWP1). HMWP2 comprises the initiation subunit with the aryl carrier protein (ArCP) that via the catalytic action of salicoyl-AMP ligase (YbtE) is acylated with salicylate. The remainder of the subunit consists of the C domain and the A domain which activates cysteine as Cys-AMP. Immediately downstream is the PCP domain as can be seen in Figure 9a. If YbtE, HMWP2, ATP, cysteine and salicylate are incubated together, a time-dependent catalytic process results in the production of a (hydroxyphenyl) thiazoline-cysteine adduct and a (hydroxyphenyl) thiazoline carboxylate (Figure 9b). In theory, if the A domain had the ability to activate L-phenylserine in a similar way as cysteine, one of the final products obtained, would be a (hydroxyphenyl) oxazoline carboxylate with an additional phenyl group attached (from the L-phenylserine). Further decarboxylation, dehydroxylation and a two electron oxidation of the oxazoline would result in the formation of 2,5-diphenyloxazole. For strain SPR to synthesise 2,5-diphenyloxazole, it would have to have a truncated form of HMWP2 or similar to HMWP2, the ability to utilise cysteine would be replaced by L-phenylserine and additional enzymatic processes would have to occur (decarboxylation, dehydroxylation and oxidation reactions).

It is not certain whether the oxazole plays any role in the mode of action of the larger molecules in which it is present, e.g. pristinamycin. According to Lancini et al. (1995), there are three ways to study the mode of action of new antibiotics: 1. In intact cells; 2. In partially purified cell-free systems; and 3. In one or more purified enzyme systems. There are, however, two basic rules which one can follow in predicting the mode of action: 1. If the novel antibiotic has a similar structure to that of a known antibiotic it can be assumed that they will have the same mode of action; and 2. If the novel antibiotic has a similar structure to that of an intermediate in bacterial metabolism, one could predict that the antibiotic could act as an antagonist to that intermediate. It is therefore possible that 21-57 could inhibit bacterial growth via the intercalation of DNA, especially when one considers the fact that the molecule has the ability to be planar, similar to that of acridine orange (three fused heterocyclic rings that allow it to act as a DNA intercalating agent), but this will remain purely speculative until experimentally proven.

21-57 was isolated and purified due to the great capacity of strain SPR to produce antibacterial agents. The antibacterial ability of PPO has not previously been tested, probably because its fluorescence ability was considered to be its most essential quality. According to the Material Safety Data Sheet (MSDS) for PPO, the compound exhibits a low level of toxicity, with an LD₅₀ of 750 mg/kg of mouse, given intraperitoneally. This compares favourably with the LD₅₀ values of known antibiotics: erythromycin, LD₅₀ = 660 mg/kg; tetracycline, LD₅₀ = 200 - 300 mg/kg; and rifampicin, LD₅₀ = 340 mg/kg (all determined for intraperitoneal application in mice) (Lancini et al., 1995). The LD₅₀ values for other medically important antibiotics are: chloramphenicol LD₅₀ = 1320 mg/kg; penicillin G LD₅₀ = 3490 mg/kg; streptomycin LD₅₀ = 1400 mg/kg and lincomycin LD₅₀ = 1000 mg/kg (Lancini et al., 1995).

An aspect of PPO which has to be considered is the fact that it is metabolised by aryl hydrocarbon hydroxylase (AHH) in humans. The enzyme AHH is associated with cancer formation - induction of the enzyme leads to the metabolism of polycyclic aromatic hydrocarbons into carcinogenic reactive metabolites (Ahokas et al., 1987; Mutch et al., 1985). Due to its fluorescence ability, PPO was tested as a substrate for AHH in vivo and its metabolism resulted in the production of another fluorescent compound. It can therefore be used to study the mechanism of action and inducibility of AHH. Mutch et al. (1985) also determined that PPO is oxidatively metabolised by the mono-oxygenase enzyme cytochrome P-450 in mice and several cytochrome P- 450 isozymes in human liver microsomes. PPO is obviously recognised as a potentially toxic agent and is therefore targeted by the mono-oxygenases. This could preclude the use of PPO as an anti-tubercular agent.

In conclusion, 21-57 or 2,5-diphenyloxazole, could potentially be used as an anti- tubercular agent: it has a low cytotoxicity, it is active against M. tuberculosis H37Rv in vitro, and is available in large amounts from a variety of chemical suppliers. The compound can also be used as a starting point for the chemical synthesis of other oxazole-containing compounds which can be developed with the aim of generating other anti-tuberculosis drugs. The in vivo activity of PPO against M. tuberculosis needs to be determined as recommended by Sood et al. (2005) and the possibility of degradation in the human body by AHH needs to be considered/tested before any further development of this compound as a potential anti-tubercular drug is considered.

The following references are included herein by reference.

Agnew, K.A., McCarley, T.D., Agbaria, R.A. & Warner, LM. (1995). Phase transition pattern of 2,5-diphenyloxazole/γ-cyclodextrin (PPO/γ-CD) self-assembly aggregates. Journal of Photochemistry and Photobiology A: Chemistry 91 , 205 - 210.

Ahokas, J.T., Davies, C, Jacobsen, N., Karki, N.T., Philippides, A. & Treston, A.M. (1987). The metabolism of 2,5-diphenyloxazole (PPO) in human lymphocytes and rat liver microsomes. Pharmacology & Toxicology 61 , 184-190. Barbour, LJ. (2003). X-Seed - A software tool for supramolecular crystallography. Journal of Supramolecular Chemistry 1 , 189 - 191.

Berdy, J. (2005). Bioactive microbial metabolites. The Journal of Antibiotics (Tokyo) 58, 1 - 26.

Canu, A. & Leclercq, R. (2001). Overcoming bacterial resistance by dual target inhibition: the case of streptogramins. Current Drug Targets - Infectious Disorders 1 , 215 - 225.

Challis, G.L. & Hopwood, D.A. (2003). Synergy and contingency as driving forces for the evolution of multiple secondary metabolite production by Streptomyces species. Proceedings of the National Academy of Sciences USA 100, 14555-14561.

DOSE (The Dictionary of Substances and Their Effects), 2^nd Electronic Edition (2004). The Royal Society of Chemistry/Knovel Corp. http://www.knovel.com

Egawa, Y., Umino, K., Tamura, Y., Shimizu, M. & Kaneko, K. (1969). Sulfomycins, a series of new sulfur-containing antibiotics. I. Isolation, purification and properties. The Journal of Antibiotics (Tokyo) 22, 12 - 17.

Hacene, H. & Lefebvre, G. (1995). AH17, a new non-polyenic antifungal antibiotic produced by a strain of Spirillospora. Microbios 83, 199 - 205.

Hariharan, C, Vijaysree, V. & Mishra, A.K. (1997). Quenching of 2,5- diphenyloxazole (PPO) fluorescence by metal ions. Journal of Luminescence 75, 205 - 211.

Ichiba, T., Yoshida, W. Y. & Scheuer, PJ. (1991). Hennoxazoles: Bioactive bisoxazoles from a marine sponge. Journal of American Chemical Society 113, 3173- 3174.

lonescu, S., Popovici, D., Balaban, A.T. & Hillebrand, Wl. (2005). Experimental and theoretical study of 2,5-diaryloxazoles whose aryl are para-substituted phenyl groups. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy ARTICLE IN PRESS. Jackman, J. E., Fierke, C.A., Tumey, L.N., Pirrung, Wl., Uchiyama, T., Tahir, S. H., Hindsgaul, O. & Raetz, C.R.H. (2000). Antibacterial agents that target lipid A biosynthesis in Gram-negative bacteria. The Journal of Biological Chemistry 275, 11002-11009.

Kanzaki, H., Wada, K-I., Nitoda, T. & Kawazu, K. (1998). Novel bioactive oxazolomycin isomers produced by Streptomyces albus JA3453. Bioscience, Biotechnology and Biochemistry 62, 438 - 442.

Keating, T.A., Marshall, CG. & Walsh, CT. (2000). Reconstitution and characterization of the Vibrio cholerae vibriobactin synthetase from VibB, VibE, VibF, and VibH. Biochemistry 39, 15522-15530.

Lancini, G., Parenti, F. & GaIIo, G.G. (1995). Antibiotics. A multidisciplinary approach. New York: Plenum Press.

Lenski, R.E. & Riley, M.A. (2002). Chemical warfare from an ecological perspective. Proceedings of the National Academy of Sciences (USA) 99, 556-558.

Lindquist, N. & Fenical, W. (1991). Isolation and structure determination of diazonamides A and B, unusual cytotoxic metabolites from the marine ascidian Diazona chinensis. Journal of American Chemical Society 113, 2303-2304.

Milne, J.C, Eliot, A.C., Kelleher, N.L. & Walsh, CT. (1998). ATP/GTP hydrolysis is required for oxazole and thiazole biosynthesis in the peptide antibiotic microcin B17. Biochemistry 37, 13250 - 13261.

Mutch, E., Woodhouse, K.W., Williams, F.M., Lambert, D., James, O.F.W. & Rawlins, M. D. (1985). 2,5-Diphenyloxazole as a probe for microsomal mono- oxygenation in human and rat liver. Xenobiotica 15, 599-603.

Nicolaou, K.C., Hao, J., Reddy, M.V., Rao, P.B., Rassias, G., Snyder, S.A., Huang, X., Chen, D.Y-K., Brenzovich, W.E., Giuseppone, N., Giannakakou, P. & O'Brate, A. (2004). Chemistry and biology of diazonamide A: second total synthesis and biological investigations. Journal of American Chemical Society 126, 12897 - 12906. Supporting information downloaded from http://pubs.acs.org Noltemeyer, Wl., Sheldrick, G. M., Hoppe, H-U. & Zeeck, A. (1982). 2-ethyl-5-(3- indoly!)oxazole from Streptomyces cinnamomeus discovered by chemical screening. Characterization and structure elucidation by X-ray analysis. The Journal of Antibiotics (Tokyo) 35, 549 - 555.

Oki, T., Kitamura, I., Yoshimoto, A., Matsuzawa, Y., Shibamoto, N., Ogasawara, T., Inui, T., Takamatsu, A., Takeuchi, T., Masuda, T., Hamada, M., Suda, H., Ishizuka, M., Sawa, T. & Umezawa, H. (1979). Antitumor anthracycline antibiotics, aclacinomycin A and analogues. I. Taxonomy, production, isolation and physicochemical properties. The Journal of Antibiotics (Tokyo) 32, 791 - 800.

Otwinowski, Z. & Minor, W. (2000). DENZO and SCALEPACK. In International Tables for Crystallography VoIF, ed. M. G. Rossman & E. Arnold. Dordrecht: Kluwer.

POV-RAY™ for Windows (Version 3.1) (2003). Persistence of Vision Pty. Ltd. http://www.poyray.org/

Prescott, L.M., Harley, J.P. & Klein, D.A. (1996). Chapter 33: Antimicrobial Chemotherapy. In Microbiology 3^rd edition, ed. E.M. Sievers, pp. 656 - 675. Boston: Wm. C. Brown Publishers.

Rodriguez, I.I. & Rodriguez, A.D. (2003). Homopseudopteroxazole, a new antimycobacterial diterpene alkaloid from Pseudopterogorgia elisabethae. Journal of Natural Products 66, 855-857.

Roy, R.S., Gehring, A.M., Milne, J.C., Belshaw, P.J. & Walsh, CT. (1999).

Thiazole and oxazole peptides: biosynthesis and molecular machinery. Natural Product Reports 16, 249 - 263.

Sandvoss, M., Weltring, A., Preiss, A., Levsen, K. & Wuensch, G. (2001).

Combination of matrix solid-phase dispersion extraction and direct on-line liquid chromatography-nuclear magnetic resonance spectroscopy-tandem mass spectrometry as a new efficient approach for the rapid screening of natural products: Application to the total asterosaponin fraction of the starfish Asterias rubens. Journal of Chromatography A 917, 75 - 86. Sbardella, G., Mai, A., Artico, M., Loddo, R., Setzu, M.G. & La CoIIa, P. (2004).

Synthesis and in vitro antimycobacterial activity of novel 3-(1H-pyrrol-1-yl)-2- oxazolidinone analogues of PNU-100480. Bioorganic & Medicinal Chemistry Letters 14, 1537-1541.

Semenova, O.N., Galkina, O.S., Patsenker, L.D., Yermoienko, I.G. & Fedyunyayeva, I. A. (2004). Experimental and theoretical investigation of the reaction of 2,5-diphenyl~1 ,3-oxazole and 2,5-diphenyl-1 ,3,4-oxadiazole dimethylamino derivatives with the Vilsmeier reagent. Functional Materials 11, 67 - 75.

Sheldrick, G.M. (1997). SHELXS97 and SHELXL97, suite of programs for crystal structure determination and refinement, University of Gδttingen.

Shin-ya, K., Wierzba, K., Matsuo, K-I., Ohtani, T., Yamada, Y., Furihata, K., Hayakawa, Y. & Seto, H. (2001). Telomestatin, a novel telomerase inhibitor from Streptomyces anulatus. Journal of American Chemical Society 123, 1262-1263.

Sood, R., Rao, M., Singhal, S. & Rattan, A. (2005). Activity of RBx 7644 and RBx 8700, new investigational oxazolidinones, against Mycobacterium tuberculosis infected murine macrophages. International Journal of Antimicrobial Agents 25, 464 - 468.

Tanitame, A., Oyamada, Y., Ofuji, K., Fujimoto, Wl., Suzuki, K., Ueda, T., Terauchi, H., Kawasaki, M., Nagai, K., Wachi, M. & Yamagishi, J-I. (2004).

Synthesis and antibacterial activity of novel and potent DNA gyrase inhibitors with azole ring. Bioorganic & Medicinal Chemistry 12, 5515-5524.

Walsh, CT. (2004). Polyketide and nonribosomal peptide antibiotics: modularity and versatility. Science 303(19 March), 1805 - 1810.

Wilson, I.D. (2000). Review: Multiple hyphenation of liquid chromatography with nuclear magnetic resonance spectroscopy, mass spectrometry and beyond. Journal of Chromatography A 892, 315 - 327.

Yeh, V.S.C. (2004). Recent advances in the total synthesis of oxazole-containing natural products. Tetrahedron 60, 11995 - 12042. Yun, B.S., Hidaka, T., Furihata, K. & Seto, H. (1994). Promothiocins A and B, novel thiopeptides with a tipA promoter including activity produced by Streptomyces sp. SF2741. The Journal of Antibiotics (Tokyo) 47, 510 - 514.

Claims

1. An isolated microorganism of the strain type SPR^T (= DSM 44925^T = NRRL B-24448^T).

2. A microorganism according to claim 1 which is an actinomycete strain.

3. A microorganism according to either one of claims 1 and 2 from the genus Streptomyces,

4. A microorganism according to claim 3 from the family Streptomycetaceae.

5. A microorganism according to claim 4 wherein the microorganism is Streptomyces polyantibioticus.

6. A microorganism according to claim 5 including genetic material of GenBank accession number DQ141528 (16S-rDNA-sequence) (SEQ ID NO:1).

7. A biologically pure culture having identifying characteristics of GenBank accession number DQ141528 (16S-rDNA-sequence).

8. Actinomycete strain SPR^T (= DSM 44925^T = NRRL B-24448⁷).

9. A method of producing 2,5-diphenyloxazole or variants or derivatives thereof, the method comprising recovering the 2,5-diphenyloxazole or variants or derivatives thereof from a microorganism of the strain type SPR^T(= DSM 44925^T = NRRL B-24448^T).

10. 2,5-diphenyloxazole or variants or derivatives thereof recovered from a microorganism having identifying characteristics of GenBank accession number DQ141528 (16S-rDNA-sequence).

11. Use of 2,5-diphenyloxazole or variants or derivatives thereof as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

12. A composition for use as an antimicrobial agent, the composition comprising an effective amount of 2,5-diphenyloxazoIe or variants or derivatives thereof.

13. 2,5-diphenyloxazole or variants or derivatives thereof for use as an antimicrobial agent.

14. 2,5-diphenyloxazole or variants or derivatives thereof for use in the manufacture of a medicament for use as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

15. Use of 2,5-diphenyloxazole or variants or derivatives thereof in the manufacture of a medicament for use as an antimicrobial agent including use in the treatment of influenza, tuberculosis and/or other bacterial infections.

16. An isolated polynucleotide of GenBank accession number DQ141528 (16S-rDNA-sequence).

17. An isolated polynucleotide sequence comprising a polynecleotide of GenBank accession number DQ141528 (16S-rDNA-sequence).

18. An isolated microorgansim according to the invention, substantially as hereinbefore described or exemplified.

19. An isolated polynucleotide sequence according to the invention, substantially as hereinbefore described or exemplified.