The marine environment is a rich source for unique actinomycete bacteria of significant biological and chemical diversity.1,2 Marine-obligate and marine-derived actinomycetes have been shown to be the source of structurally diverse secondary metabolites, many of which have been shown to possess interesting anticancer properties.3 The discovery of antibiotics from this source is far less developed, but promising antibiotics such as anthracimycin provide significant motivation to explore this source further.4 Earlier, we showed that the marine-derived Streptomyces sp., strain CNB-091, produces a suite of at least five structurally unprecedented depsipeptides, salinamides A−E, that show significant antibacterial and anti-inflammatory properties.5,6 Subsequently, salinamide A was shown to possess significant inhibitory activity against RNA polymerase (RNAP) from Gram-positive and Gram-negative bacteria.7,8 As part of a collaborative program to further explore the chemical biology of the RNAP-inhibitory activity of the salinamides, we examined in more detail the extract and semi-purified fractions from the recultivation of Streptomyces sp. CNB-091 and report here the isolation of a new salinamide analog, salinamide F, which, like salinamide A, also possesses significant RNAP-inhibitory and antibacterial activity.

Materials and methods

Culture conditions and extraction

Streptomyces sp., strain CNB091, was isolated from a surface swab of a jellyfish (C. xamachana) collected in the Florida Keys in 1987 as previously described.5 The producing organism was cultured in forty 2.8 l Fernbach flasks each containing 1 l of a seawater-based medium (10 g starch, 4 g yeast extract, 2 g peptone, 1 g CaCO3, 40 mg Fe2(SO4)3-4H2O, 100 mg KBr) by shaking at 230 r.p.m. at 27 °C for 7 days. At the end of the cultivation, sterilized XAD-16 resin (20 g l−1) was added to adsorb the organic metabolites, and the culture and resin were shaken at 215 r.p.m. for 2 h. The resin and cell debris were filtered through cheesecloth, washed with deionized water and eluted with acetone. The acetone filtrate was concentrated under reduced pressure, and the resulting aqueous layer was extracted with ethyl acetate (3 × 500 ml). The ethyl acetate-soluble fractions were combined and reduced in vacuo to yield 3.5 g of extract from a 40 × 1 l culture.

Isolation and purification of salinamide F

The ethyl acetate extract (3.5 g), from cultivated strain CNB091, was subjected to silica vacuum flash chromatography using sequential mixtures of dichloromethane and methanol as eluents (elution order: 1, 2, 3, 4, 5, 10, 15, 20 and 30% methanol in dichloromethane, and 100% methanol). Guided by the results of bioassays and 1H NMR analyses, the fractions containing bioactive metabolites were combined (210.0 mg) and purified using a C8 reversed-phase semi-preparative HPLC column (YMC ODS-A column, 1 × 25 cm, 50% aqueous MeCN) to yield 13.2 mg of salinamide F (1) as pale yellow amorphous solid along with the previously described salinamides A and B and minor amounts of salinamides C–E.

Preparation of the acetonide 4

Salinamide F (1, 2 mg) was dissolved in dry acetone (1 ml), and BF3·Et2O (20 μl) was added at 0 °C. The reaction was allowed to stir for 3 h at 0 °C, and then quenched with saturated aqueous NaHCO3. Next, the aqueous phase was extracted twice with ethyl acetate. The combined ethyl acetate solutions were concentrated under reduced pressure, and the residue was purified by C8 reversed-phase semi-preparative HPLC (YMC ODS-A column, 1 × 25 cm, 50% aqueous acetonitrile) to yield 1.4 mg of the acetonide 4 as pale yellow solid.

RNAP-inhibitory activity

Fluorescence-detected RNAP-inhibition assays were performed by a modification of the procedure previously described.9 Reaction mixtures contained (20 μl): 0–100 μM test compound, 75 nM bacterial RNAP holoenzyme (75 nM Escherichia coli RNAP holoenzyme or 75 nM Staphylococcus aureus RNAP core enzyme and 300 nM S. aureus σA; (prepared as described in a previous paper10), 20 nM DNA fragment containing bacteriophage T4 N25 promoter (positions −72 to +367; prepared by PCR from plasmid pARTaqN25-340-tR211), 100 μM ATP, 100 μM GTP, 100 μM UTP and 100 μM CTP in transcription buffer (50 mM Tris–HCl, pH 8.0, 100 mM KCl, 10 mM MgCl2, 1 mM DTT, 10 μg ml−1 bovine serum albumin, 5% methanol and 5.5% glycerol). Components other than DNA and NTPs were pre-incubated for 10 min at 37 °C. Reactions were carried out by addition of DNA and incubation for 15 min at 37 °C, followed by addition of NTPs and incubation for 60 min at 37 °C. DNA was removed by addition of 1 μl 5 mM CaCl2 and 2 U DNase I (Ambion), followed by incubation for 90 min at 37 °C. RNA was quantified by addition of 100 μl Quant-iT RiboGreen RNA Reagent (Life Technologies, Carlsbad, CA, USA; 1:500 dilution in 10 mM Tris-HCl, pH 8.0, 1 mM EDTA), followed by incubation for 10 min at 22 °C, and measurement of fluorescence intensity (excitation wavelength=485 nm and emission wavelength=535 nm; GENios Pro microplate reader (Tecan, Männedorf, Switzerland)).

Antibacterial activity

Minimum inhibitory concentrations (MICs) were quantified using broth microdilution assays;12 using a starting cell density of 2 × 105 c.f.u. ml−1, LB broth13 and an air atmosphere for E. D21f2tolC (tolC:Tn10 rfa lac28 proA23 trp30 his51 rpsL173 ampC tsx81; strain with cell-envelope defects resulting in increased susceptibility to hydrophobic agents, including salinamides8,14), S. aureus (ATCC 12600), Enterococcus faecalis (ATCC 19433) and Enterobacter cloacae (ATCC 13047); and using a starting cell density of 2 × 105 c.f.u. ml−1, Haemophilus Test Medium broth,15 and a 7% CO2, 6% O2, 4% H2, 83% N2 atmosphere for Haemophilus influenzae (ATCC 49247) and Neisseria gonorrhoeae (ATCC 19424).

Salinamide F (1), a new bicyclic depsipeptide, was isolated in addition to the known salinamides A (3) and B (2) (Figure 1), as well as salinamides C–E, which were produced in minor amounts but not purified. Analysis of salinamide F by HRTOFMS showed quasi-molecular ions at m/z 1038.51940 [M+H]+ and 1060.50454 [M+Na]+, which analyzed for the true molecular formula C51H71N7O16. The molecular weight of 1 was larger than salinamide A (3) by 18 mass units, which suggested the addition of one molecule of water. The structure could be fully defined by comprehensive analysis of 1D and 2D NMR data, including 1H,13C NMR, COSY, HSQC and HMBC experiments (Table 1). A loss of the C-40 signals in both the 1H and 13C NMR spectra at δH 2.44 (d, 5.4), 2.95 (d, 5.4) and δC 55.4, as well as the appearance of new signals δH 3.47 (m) and δC 66.0, in addition to the downfield shift of C-8 by δΔ +20 p.p.m. suggested that the epoxide ring had been opened (C-7-O-41-C-40) (Table 1). The HMBC NMR spectrum showed a 2J correlation of H-6 (δH 6.19, d, J=15.0) and H-8 (δH 4.61, m) with C-8 (δC 80.7), as well as 3J correlation between H-40 (δH 3.47, m) and C-6 (δC 147.7) as well as C-8 (δC 79.6) (Figure 2) supporting this suggestion. The remaining 1H and 13C NMR signals for 1 were virtually identical to those of salinamide A (3).5,6

Figure 1
figure 1

Structures of salinamides F (1), B (2) and A (3).

Full size image
Table 1 1H and 13C NMR data for salinamide F (1) in CDCl3
Full size table
Figure 2
figure 2

NMR 1H-1H COSY and HMBC correlations for salinamide F (1). COSY correlations are labeled by bold bonds; HMBC correlations shown as arrows.

Full size image

The relative configuration at C40 was assigned by analysis of 2D ROESY NMR data derived from 1 and its acetonide derivative 4 (Figure 3). For salinamide F (1), NOE correlations between H2-40 (δH 3.47, m), H-6 (δH 6.19, d, J=15.0) and H-8 (δH 4.61, m) also suggested the opening of the epoxide ring (C-7-O-41-C-40) with retention of configuration at the quaternary center. To confirm this, the acetonide derivative 4 was prepared and its relative configuration evaluated by ROESY NMR experiments using methods alrerady applied in similar systems.16 Strong NOE correlations were observed between the H2-40 (δH 3.64, m), and both H-66 methylene protons [δH 3.29, dd, J=15.0, 10.0), 3.62, dd, J=15.0, 5.0] and an acetonide methyl (δH 1.26, s), as well as NOE correlation between H-8 (δH 4.61,m) and H-66 (δH 3.62, dd, J=15.0, 5.0). Slight differences in the bond angles of derivative 4, apparently derived by formation of the semi-planar ketal ring decreases the spatial distance between the H2-40 protons and the benzyl protons at C-66, thus confirming the spatial proximity of these protons.

Figure 3
figure 3

Illustration of the ROESY correlations observed for salinamide F (1) and the acetonide product 4.

Full size image

Salinamide F showed potent inhibition of Gram-positive and Gram-negative bacterial RNAP, with IC50=4 μM for S. aureus RNAP and 2 μM for E. coli RNAP. Salinamide F exhibited significant antibacterial activity against Gram-positive and Gram-negative bacteria, showing MIC50=12.5 μg ml−1 for Enterococcus faecalis, 100 μg ml−1 for S. aureus, 12.5 μg ml−1 for H. influenzae, 25 μg ml−1 for Neisseria gonorrhoeae, 50 μg ml−1 for Enterobacter cloacae and 0.20 μg ml−1 for E. coli D21f2tolC. The comparable RNAP-inhibitory activities of salinamide F (1) and salinamides A (3) and B (2)8 are consistent with the conclusion that the epoxide functionality in salinamide A and the corresponding clorohydrin functionality in salinamide B are not essential for RNAP inhibition.8 Substitutions of the RNAP β and β' subunits that confer high-level (⩾8-fold) resistance to salinamides A (3) and (2)8 also confer high-level resistance to salinamide F (1) (Table 2). We infer that salinamide F inhibits RNAP through the same binding site on RNAP as salinamides A and B (i.e., the ‘Sal target,’ comprising residues of the ‘F-loop’ and ‘link region’ within RNAP β subunit and the ‘bridge-helix N-terminal hinge’ within RNAP β' subunit8). Substitutions of the RNAP β subunit that confer high-level resistance to the structurally unrelated RNAP inhibitor rifampin do not confer resistance to salinamide F (Table 3). We infer that salinamide F, like salinamides A and B,8 does not interact with the rifampin binding site on RNAP.

Table 2 Sal-resistant mutants: cross-resistance to SalF
Full size table
Table 3 Rif-resistant mutants: absence of cross-resistance to SalF
Full size table