2.2. Detection of PKS and NRPS Genes

Type I and Type II polyketide synthases (PKS-I and PKS-II) and nonribosomal peptide synthetases (NRPS) are biosynthetic systems involved in the synthesis of an important number of microbial natural products classes including important drugs with antibiotic, antifungal or anticancer activity. Whereas the diversity of new chemical structures with biological activities produced by sponges supports their significance as reservoir of new therapeutic agents, today sponge-associated microbial communities are considered to be responsible for the biosynthesis of many of these agents. The presence of PKS and NRPS genes has been previously used to reveal the biosynthetic potential of natural products in bacterial isolates [13]. We evaluated their occurrence in the genomes of the 44 bacteria using degenerate PCR primers targeting conserved motifs in these genes to infer their biosynthetic potential (Table 1).

Gene sequences corresponding to PKS-II were the most frequent and were detected in 29 of the 44 strains (66%), whereas NRPS and PKS-I sequences were only identified in 16 (36.4%) and 9 strains (20.5%) respectively. The strains of Kocuria stand out in their relative richness in these biosynthetic systems compared to those detected in Micrococccus.

The observed frequencies of NRPS and PKS-I gene sequences detected in marine Micrococcaceae (36.4% NRPS and 20.5% PKS-I) are lower than those observed in our previous studies with other terrestrial actinobacteria groups such as lichen Pseudonocardiaceae (68.2% PKS-II; 86.4% NRPS and PKS-I) or tropical soil Micromonosporaceae (42% NRPS, 77% PKS-I). Despite these results, NRPS and PKS sequences have been reported in marine actinobacteria isolated from a wide variety of environments ranging from marine caves, coral reef sediments, the deep sea Mariana Trench to north sea sediments, and with a high occurrence variability reaching up to 94% of the strains [31,32,33]. When comparing our results with other reports including all lineages of sponge-associated actinomycetes, our strains differ in the reduced number of PKS and NRPS gene sequences detected. In fact 70%–87% of the actinomycetes isolated from Norwegian marine sponges, Antarctic deep-sea sponges or the marine sponge Iotrochota sp. collected in the South China Sea contained PKS-I and NRPS genes [33,34,35].

So far, few Micrococcaceae strains have been reported for their capacity to produce natural products with biopharmaceutical potential. Li et al., (2012) reported actimicrobial activities produced by two endophytic actinobacteria from Artemisia annua, a Kocuria strain with type I PKS and NRPS gene sequences and one Micrococcus strain with type II PKS [36]. Similarly Gontang et al. (2010) detected the presence of NRPS genes in 2 of the 5 Kocuria strains isolated from marine sediments [14]. On the contrary, none of the Kocuria strains isolated from the Challenger Deep sediments from the Mariana Trench or the M. luteus isolated from the marine sponge Halichondria panacea gave any positive PKS or NRPS amplification products [12,32]. In our study, it is interesting to show that 86% and 60% of the Kocuria and Micrococcus isolates possess at least one of these biosynthetic pathways classes. These results are remarkable because K. rhizophila DC2201 has been shown to contain one of the smallest actinomycetes genomes (2.7 Mb). Genome annotation has shown that it encodes only a limited number of secondary metabolite pathways, including a type III polyketide synthase and a nonribosomal peptide synthetase, but does not contain genes for typical bacterial type I or type II PKS [37].

The detection of genes associated to these biosynthetic clusters does not guarantee the expression of genes involved in the production of secondary metabolites. The absence of PCR products in some of the isolates may reflect the lack of biosynthetic genes or the presence of less conserved domains sequences with low homology with the primers. Furthermore, PCR products can reflect the presence of genes involved in the biosynthesis of other types of metabolites, such as pigments or structural components of the microbial cell such as fatty acids [38], and an involvement in quorum sensing or iron metabolism has also been proposed for these gene products [31,32,39,40].

2.3. Evaluation of Antimicrobial Activities

To evaluate the production of bioactive secondary metabolites the 44 marine isolates were cultivated in nutritional arrays of 12 different liquid media using Duetz 96-deep well plates. Applying this micro-cultivation system, all steps in the screening procedure, including the fermentation, solvent extraction of fermentation broths, storage of extracts and bioassays, could be efficiently performed in a standard 96-well format. In total, 528 crude extracts were analyzed for the production of antibacterial and antifungal activities in whole-cell agar-based growth inhibition assays against Bacillus subtilis MB964, methicillin-resistant Staphylococcus aureus (MRSA) MB5393, Acinetobacter baumannii MB5973 and the yeast Candida albicans MY1055.

The frequency of microbiological activities against bacterial human pathogens was considered to be an indicator of their ability to produce anti-infective molecules with potential therapeutic properties [41]. Three out of the 528 extracts that were tested in these bioassays inhibited the growth of MRSA. The producer actinobacteria strains correspond to two Kocuria spp. and one Micrococcus sp., that have been identified after phylogenetic analysis as the strains Kocuria marina F-276,310, Kocuria palustris F-276,345, and Micrococcus yunnanensis F-256,446 (Figure 1).

The preliminary inhibitory activity observed with 10 μL extracts in the agar well-based assay format was confirmed in a second agar assay directly using 10 μL and 20 μL extracts spotted on Nunc plates (24 × 24 cm) containing the reporter strains. The bioactivity was confirmed in both formats and the diameters of the zones of inhibition recorded, observing a 2 mm increase of the inhibition zone upon duplication of the extract volume used (Table 2).

The anti-MRSA activity was confirmed in new 10 mL fermentations and a preliminary LC-MS analysis showed similar profiles in the three extracts, with a principal component (UV absorption maxima at 218, 307 and 349 (sh) nm) of 1514.366 Da and others minor components. The strain Kocuria sp. F-276,345 was then selected for further studies on the production and characterization of this antibacterial compound that was later identified by Martín et al. (2012) [42] as kocurin, a new member of the thiazolyl peptide family of antibiotics (Figure 2).

Open in a separate window

Figure 2

Chemical structure of thiazolyl peptide kocurin.

2.4. Production Conditions of the Thiazolyl Peptide Kocurin

The production of the new thiazolyl peptide kocurin by the strains of Kocuria and Micrococcus was detected in short fermentations in only one of the 12 production conditions tested, the medium R358, previously described by Jensen et al. (2007) [43] as providing enough amounts of bromide and iron, elements frequently observed in halogenated marine-derived compounds or of limited access in marine environments [39,44]. As previously observed with other examples of the family Microccocaceae, the compound was only produced in this medium in the absence of marine salts, conditions that were used to scale up the production for the isolation of the molecule. In fact, our group had already described two related strains of Microccocaceae corresponding to the species Arthrobacter, R-7513 and R-7914 (Figure 1), which were isolated in a previous study from Antarctic microbial mats. Both strains produced a potent anti-MRSA compound characterized by a major LCMS peak of 1514 Da proposed at that time as a new thiazolyl peptide [21] that we confirmed to correspond to kocurin (data not shown). In contrast to the Kocuria and Micrococcus strains, both strains of Arthrobacter produce kocurin in the medium CGY [21], a medium also used in our study without success.

The production conditions of kocurin by the Kocuria strain F-276,345 were analyzed with a time course study monitoring the growth and kocurin production in the medium R358 during 4 days. Given that our preliminary experiments using standard seed ratios of 5% had shown a rapid exponential growth and production of the compound in concentrations of 0.2 μg/mL within the first 24 h of cultivation, the experiment was established using diluted seed conditions ensuring initial optical densities as low as OD₆₀₀ 0.015. Using this approach we have confirmed that whereas the strain reached the stationary phase within the first 48 h, kocurin production, estimated as the intensity of the ion M + 2Na⁺ with m/z = 780, was initiated during the exponential growth reaching its maximal production 24 h after the establishment of the stationary phase to fall afterwards as the incubation proceeds up to 4 days (Figure 3).

Open in a separate window

Figure 3

Time course production of kocurin by Kocuria palustris F-276,345. Growth levels (OD₆₀₀) (▲) and kocurin production (■) (estimated as the m/z = 780 intensity) were followed during 100 h of cultivation at 28 °C in medium R358; all samples points were analyzed in triplicate.

3.2. Strain Culture and DNA Extraction

All strains were grown on R2A agar medium (BD) and MY liquid medium (10 g/L glucose, 3 g/L yeast extract, 5 g/L proteose-peptone, 3 g/L malt extract) both supplemented with 3% (w/v) sea salts (Sigma Aldrich). Plates were incubated in a humidified chamber at 28 °C for 2–4 weeks. Following incubation, selected cultures were incubated in MY liquid medium at 28 °C in an orbital shaker at 220 rpm for 4 days. For long-term storage, pure cultures were frozen in 20% glycerol at −80 °C. Total genomic DNA from marine actinomycetes was recovered and purified as described elsewhere [53].

3.3. Analysis of Fatty Acid by Gas Chromatography

Bacterial cultures were grown on TSB (Trypticase Soy Broth BBL 30 g/L, agar 15 g/L and 30 g/L sea salts) at 28 °C for 24 h. Vegetative growth was scraped and fatty acid methyl esters (FAMEs) were prepared using a modified sample preparation [54]. Analysis of FAMEs was carried out by capillary gas chromatography using a Hewlett-Packard Model 5890 gas chromatograph/MIDI system (Microbial ID, Inc., Newark, DE, USA) equipped with phenyl-methyl silicon column (0.2 mm × 25 m). Chromatography conditions were performed as recommended by the manufacturer. Individual FAMEs were identified using the Microbial Identification Software (MIS).

3.4. Phylogenetic Analysis

The 16S rRNA gene (approximately 900 bp) was partially PCR-amplified from the purified genomic DNA obtained from bacterial cultures [11]. Nucleotide sequences were analyzed and edited using MEGA (5.1 version) [55]. The partial 16S rRNA gene sequences from cultivable isolates were compared to 16S rRNA genes in the Eztaxon database [56] to determine the closest phylogenetic positions. Sequences (averaging 900 nucleotides) were aligned using ClustalW [57] including representative actinomycete 16S rRNA gene sequences from the family Micrococcaeae. Phylogenetic tree was constructed generating a complete alignment of 16S rRNA gene sequences of selected members of each genus within the family Micrococcaceae. Trees were generated using the Neighbor-Joining and Jukes-Cantor algorithms [24,25]. Bootstrap values were calculated from 1000 resampled datasets.

3.5. PCR Amplification

Ketosynthase (KS) domains of type I polyketide synthase (PKS) gene were PCR amplified from genomic DNA using the primers K1F (5′-TSAAGTCSAACATCGGBCA-3′) and M6R (5′-CGCAGGTTSCSGTACCAGTA-3′). Type II PKS sequences were amplified specifically using KSαF (5′-TSGRCTACRTCAACGCSCACGG-3′) and KSβR (5′-TACSAGTCSWTCGCCTGGTTC-3′). Degenerate PCR primers targeting specifically NRPS adenylation domains were A3F (5′-GCSTACSYSATS TACACSTCSGG-3′) and A7R (5′-SASGTCVCCSGTSCGGTAS-3′). The bands correspond to KS domains (≈1250–1400 bp), conserved sequences KSα and KSβ (≈800–900 bp) and adenylation domains (≈700 bp). These degenerated primers were designed by Ayuso and Genilloud (2005), and Ayuso et al. (2005) [13,38]. PCR amplifications were performed in a iCycler Bio-Rad thermal cycler in a final volume of 50 μL PCR mixture contained 0.5 μM of each primer, 0.2 mM of each of the four dNTPS (Applied Biosystems), 10% dimethyl sulfoxide (DMSO), 1 unit of Taq DNA polymerase (5 U/μL) (with its recommended reaction buffer) (MP) and 5 μL of template DNA. The PCR protocol for PKS-I amplifications consisted of a 5 min denaturation step at 95 °C and 35 cycles of 30 s at 95 °C, 2 min at 57 °C for K1F/M6R, 58 °C for KSαF/KSβR and 61 °C for A3F/A7R and 4 min extension at 72 °C; followed by 10 min at 72 °C. The amplified products were analyzed by electrophoresis in 2% (w/v) agarose gels stained with ethidium bromide (Invitrogen).

3.6. Microfermentation and Extraction of Marine Microbial Secondary Metabolites

Marine actinomycetes were cultivated for the production of secondary metabolites in 96-deep well plate fermentations using different nutritional arrays in the presence and absence of sea salts [58]. Seeds were prepared inoculating 0.5 mL of a frozen inoculum stock of each strain in 12 mL of MY seed medium supplemented with 3% Sigma Aldrich sea salts. Fermentations in 96-deep well plates were performed in 800 μL of production medium and incubated for four days at 28 °C in a rotary shaker at 330 rpm and 70% humidity. Production conditions included the following six media prepared in parallel with 3% (w/v) sea salts or without sea salts: Antibiotic Assay Medium No. 3 (Assay Broth) (FLUKA); medium CGY (5 g/L bacto casitone, 5 g/L glycerol, 1 g/L bacto yeast extract, adjusted to pH 7) [21]; medium DEF-15 (40 g/L sucrose, 2 g/L NH₄Cl, 2 g/L Na₂SO₄, 1 g/L K₂HPO₄, 1 g/L MgCl₂·6H₂O, 1 g/L NaCl, 2 g/L CaCO₃, 1 mL trace elements solution (100 mg/L MnCl₂·4H₂O, 100 mg/L ZnCl₂, 100 mg/L FeCl₂·4H₂O, 50 mg/L NaI, adjusted to pH 7); medium IN (2 g/L dl-serine, 2 g/L dl-alanine, 8.6 g/L K₂SO₄, 1.4 g/L KCl, 1.4 g/L MgSO₄·7H₂O, 10 g/L sucrose, 30 g/L yeast extract, adjusted to pH 7) [59]; Marine Broth (BD); medium R358 (10 g/L starch from potato, 4 g/L yeast extract, 2 g/L peptone, 5 mL of a 20 g/L stock solution KBr and 5 mL of a stock solution of 8 g/L FeSO₄·7H₂O, adjusted to pH 7) [43]. Plates were incubated for four days at 28 °C in a rotary shaker at 330 rpm and 70% humidity. After four days of incubation, broths were extracted with 800 μL acetone and 40 μL DMSO. Each extract was shaken at 200 rpm for 1 h, the acetone evaporated and the sample concentrated under vacuum, to be finally reconstituted in 200 μL of water. The resulting crude extracts were transferred to a 96-well plate for storage at −20 °C.

3.7. Production of Thiazolyl Peptides

Scale-up fermentation conditions to reproduce the production of the bioactive compounds in microfermentations were initially performed in 40 mL EPA vials. Each vial containing 10 mL of production medium R358 was seeded with 5% inoculum in MY prepared as described previously and incubated 24 to 96 h at 28 °C in a rotary shaker at 220 rpm and 70% humidity. Acetone extraction and preparation of crude extracts was performed as described previously.

The production conditions of the active compound were studied on the basis of the initial results obtained with the Kocuria strain F-276,345. A first seed culture of strain F-276,345 was prepared in MY medium with 3% artificial sea salts (3 days at 28 °C and in a rotary shaker at 220 rpm and 70% humidity) and 2 L volume of medium R358 were inoculated to obtain a final OD₆₀₀ of 0.015 measured with a Eppendorf BioPhotometer. 10 mL aliquots of the seeded medium were then distributed in 40-mL EPA vials, and incubated during 4 days at 28 °C in a rotary shaker at 220 rpm and 70% humidity. Samples were taken by triplicate to evaluate the growth rate as derived from the DO₆₀₀ estimation and the production of the compound determined by mass spectrometry as the intensity of the kocurin fragmentation ion with m/z = 780.

The authors thank the assistance of Catalina Moreno in the preparation of the extracts. Some of these results were presented by S.P. to obtain her Diploma of Advanced Studies from the Universidad Autónoma de Madrid, Spain. This is contribution no. 13105 from the Institute of Marine and Environmental Technology and contribution number 4746 from the University of Maryland Center for Environmental Science.