5 Marine-derived Streptomyces sp
5.1 Marine-derived Streptomyces sp. Act8970
Based on chemical screening using TLC, the extract of the marine-derived Streptomyces sp.Act8970 showed a number of UV non-absorbing bands, which were stained blue on spraying with anisaldehyde/sulphuric acid. The strain exhibited mod-erate activity against the algae Chlorella vulgaris and Chlorella sorokiniana, and the Gram-negative bacteria Escherichia coli. The extract was weakly active against the Gram-positive bacteria, Bacillus subtilis and Staphylococcus aureus, the fungus Mu-cor miehei (Tü284), and the yeast Candida albicans.
Act8970 was therefore upscaled using M2+ medium (+50% artificial sea water) for 11 days to isolate its bioactive secondary metabolites. After usual work up, a brown crude extract was obtained which in turn was applied to flash silica gel col-umn chromatography. Three fractions were obtained by TLC-guided fractionation.
Purification of the fast moving fraction I using Sephadex LH-20 followed by silica gel led to attiamycin A (141) and attiamycin B (143a) as colourless oils. Purification of the middle polar fraction II delivered homononactic acid (145a), homononactic acid methyl ester (145b) and dinactin (werramycin) (148) as colourless oils. Finally, working up of fraction III led to nonactic acid (144a) and dinactin (148) as further colourless oils (Figure 74).
XAD-2 (MeOH/H2O), EtOAC Act8970
(20 l Shaker)
Biomass Filtrate
mixing with cilite and filtered by filterpress
3 x with EtOAC and 1x with acetone i.vac
(CH2N2) 2-[5-(2-Oxo-propyl)-tetrahydro-furan -2-yl]-propionic acid methyl ester Homononactic acid
Homononactic acid methyl ester Methylation (CH2N2)
Homononactic acid methyl ester
Nonactic acid
Nonactic acid methyl ester Methylation (CH2N2)
Figure 74: Work-up procedure of the marine-derived Streptomyces sp.Act8970.
5.1.1 Attiamycin A
Compound 141 was isolated as low polar colourless oil from fraction I, which was stained yellow and later to brown by anisaldehyde/sulphuric acid. The molecular weight of 141 was established as 172 Dalton using ESI and EI MS. ESI HRMS es-tablished the corresponding molecular formula as C9H16O3, entailing two double bond equivalents.
The 1H NMR spectrum displayed no aromatic or olefinic signals. Multiplet sig-nals of an oxymethine (δ 4.01) in addition to another three multiplets were observed.
The latter three signals between 2.61~2.46 (6H) and 1.73 (2H) were due to four me-thylene groups, the first three (2.61~2.46) being of sp2 systems. A methyl singlet (δ 2.18) was attached to a sp2 carbon and indicated an acetyl moiety. The remaining signal was a methyl triplet (δ 1.06) linked to a methylene group, which appeared as a quartet at δ 2.46.
The 13C/APT NMR spectra of 141 displayed 9 carbon signals, two among them were of ketone carbonyls (δ 211.7 and 209.7). The residual carbon signals indicated one oxy-methine (δ 66.8), four methylene carbons (δ 50.0, 38.1, 36.0 and 29.9) along with two methyl carbons (δ 30.6, 7.8).
Figure 75: 13C NMR spectrum (CDCl3, 75 MHz) of 4-Hydroxy-nonane-2,7-dione (141).
Based on the HMBC correlations (Figure 76) of 141, protons of the methyl trip-let (C-9) at δ 1.06 exhibited a 3J correlation towards the carbonyl (C-7) at δ 211.7, constructing a propionyl fragment (O=C-CH2-CH3). The methyl singlet (C-1, 2.18) showed a 2J correlation to the carbonyl C-2 at 209.7, and hence an acetyl group was supposed. The two methylene groups at 2.61 (C-3, C-6) displayed two 2J correlations to both carbonyls C-2 and C-7. The acetyl group (δ 2.18) showed a 3J correlation to one of the methylene carbons (δ 50.0, C-3), confirming the direct attachment be-tween C-2 and C-3. Moreover, the oxy-methine carbon C-4 showed 2J and 3J correla-tions with protons of the three methylene groups CH2-3/CH2-6 (δ 2.61), and CH2-5 (δ 1.73), while the latter methylene group (CH2-5) displayed three other relevant correlations with the two methylene carbons C-3 (δ 50.0, 3J) and C-6 (δ 38.1, 2J), and the oxy-methine carbon C-4 (δ 66.8, 2J). This resulted in the -CH2 -CH(OH)-CH2-CH2- chain, which was further confirmed by the H,H COSY data. Hence, the final structure of 141 was deduced as 4-hydroxy-nonane-2,7-dione.
C
A search in the current databases (AntiBase, DNP and CA), pointed to the nov-elty of 141. So, we have named it as attiamycin A. Attiamycin A (141) might be a precursor in the biosynthesis of decarboxy-oxo-nonactic acid (142) by ring closure, followed by an elimination of water and hydrogenation.
C Decar-boxy-oxo-nonactic acid (142).
5.1.2 Attiamycin B
Attiamycin B (143a) was obtained from fraction I as an additional low polar colourless oil. It was not UV absorbing and turned reddish-brown (later reddish vio-let) on spraying with anisaldehyde/sulphuric acid. The molecular weight of 143a was deduced from ESI MS to be 200 Dalton. ESI and ESI HRMS confirmed the molecu-lar formula as C10H16O4.
The 1H NMR spectrum displayed a similar signal pattern as 141. An acidic broad singlet (δ 6.30) might be of an acidic hydroxyl group, and two multiplets of
oxy-methines (δ 4.32 and 4.08) were observed. In the region of δ 2.80~1.50 with integration of 7H, a series of multiplets was present, corresponding to three methyl-ene groups and one methine proton (δ 2.54). The latter methine appeared as quartet, pointing to its direct connection with a methyl group, which gave a doublet at δ 1.18.
As in 141, compound 143a showed a methyl singlet at δ 2.20, corresponding to an acetyl group. On esterfication of 143a using diazomethane, component 143b was obtained, showing the singlet of a methyl ester (δ 3.69) and a molecular weight 14 amu higher than 143a. This indicated that 143a was a free carboxylic acid.
Figure 78: 1H NMR spectrum (CDCl3, 300 MHz) of Attiamycin B (143a).
The 13C/APT NMR spectra of 143a displayed 10 carbon signals, two of which at δ 207.5 and 179.7 were of ketone and carboxylic acid carbonyls, respectively. Addi-tionally, two oxycarbons (δ 80.3 and 75.5), and three methylene carbons (δ 49.6, 30.9 and 28.4) were visible with the methylene carbon at δ 49.6 possibly attached to an sp2 system. Finally, one methine signal ( 45.1) and two methyl carbons (δ 30.8 and 13.2) were exhibited. The first of the two methyl carbons was directly attached to the ketone carbonyl δ 207.5 to afford the expected acetyl group.
Based on the HMBC (Figure 79) correlations of 143a, the acetyl group and the keto carbonyl (C-2', 207.5) were further confirmed through the 2J cross-signal be-tween H3-1' (δ 2.20) and C-2' (δ 207.5). This acetyl group was directly attached to the methylene carbon C-3' (δ 49.6), as seen by the 3J coupling from H3-1', resulting
in a propanone partial structure. Towards the ketone carbonyl (C-2'), the oxy-methine H-5 (δ 4.32) showed a 3J correlation, confirming the direct attachment be-tween C-5 (δ 75.5) and C-3'. So, fragment A was deduced, which was further estab-lished by the H,H COSY correlation between H2-3' and H-5.
The methine quartet of proton H-2'' (δ 2.54) displayed a 2J coupling with the carboxylic carbonyl C-1'' (δ 179.7), and the latter showed in turn a 3J cross-signal with the methyl doublet of H3-3'' (1.18). This indicated the direct linkage between the carboxylic acid and CH-CH3. In the H,H COSY spectrum, a 3J correlation was found between H-2'' and the remaining oxy-methine H-2, confirming the direct attachment between C-2 (δ 80.3) and C-2'' (δ 45.1). So, fragment B was concluded.
CH3 O
O H H H
75.5
207.5 49.6 30.8
4.32
2.80, 260 2.20
O O
O H
CH3 H
H
179.7
13.2 45.1 80.3 6.3
4.08
2.54 1.18
A B
Based on H,H COSY, the remaining two methylene groups H2-3 (1.90/1.60) and H2-4 (2.10/1.50) were directly connected with each other. The methylene protons H2 -4 displayed a 3J correlation to the oxy-carbon C-2 (δ 80.3), while H2-3 displayed a 3J to the other oxy-carbon C-5 (δ 75.5). This confirmed the direct linkage between both fragments A and B through the assigned ethandiyl group (C2-C3). Finally, the two oxy-carbons at δ 75.5 and 80.3 must be connected via oxygen creating a ring closure to obtain a tetrahydrofuran ring, containing the isopropanoic acid and propanone groups at 2- and 5-positions. As a result, 2-[5-(2-oxo-propyl)-tetrahydrofuran-2-yl]-propionic acid (143a), a new oxy derivative of nonactic acid was fixed, which we named as attiamycin B.
CH3
Figure 79: H,H COSY ( ) and HMBC ( ) correlations of Attiamycin B (143a).
The relative stereochemistry of compound 143a was partially established on the basis of NOESY experiments (Figure 80). The proton at δ 4.32 (H-5) showed a cou-pling with one of the H2-4 protons (δ 2.10) and the latter showed in turn a coupling with one of the H2-3 protons (δ 1.90). The oxy-proton H-2 (δ 4.08) showed a cou-pling with the same H2-3 proton (δ 1.90) besides another coupling with the methyl H3-3'' (δ 1.17) in addition to a small coupling with H-2'' (δ 2.51). Further couplings
Figure 80: Diagnostic NOESY correlations of Attiamycin B (143a).
5.1.3 Nonactic acid
Compound 144a was obtained as colourless oil, showing similar chroma-tographic properties as 143a. The molecular weight (202 Dalton) and the correspond-ing formula (C10H18O4) of 144a obtained by ESI MS indicated two hydrogens more than in 143a.
The 1H/13C NMR spectra revealed 144a as a structural analogue of 143a, the sole difference was attributed to the reduction of the acetyl carbonyl (207.5), afford-ing an oxy-methine located at δ 66.1 (δH 3.91). The terminal carboxylic acid group was established by esterification, and the corresponding methyl ester singlet was found at δ 3.69.
Based on detailed spectroscopic data, compound 144a was elucidated as a hy-drogenated derivative of 143a, i.e. nonactic acid (144a), which was as further con-firmed by H,H COSY, HMQC and HMBC correlations (Figure 81) and literature data[233].
CH3 O
OH O
RO
CH3 1
2 5 1'
1'' 2'' 3'
3''
CH3 O
OH O
O H
CH3 1
2 5
1' 1'' 3'
2''
144a: R = H; 144b: R = CH3
Figure 81: H,H COSY ( ) H,H and HMBC (→) correlations of Nonactic acid (144a).
5.1.4 Homononactic acid
A further tetrahydrofuran derivative was was again obtained as colourless oil.
The molecular weight of 145a was deduced as 216 Dalton, and HRESI MS gave the corresponding molecular formula C11H20O4, containing one methylene group more than 144a. In the 1H NMR spectrum, the compound exhibited a identical pattern as for 144a, except that one of the methyl doublet present in the side chain of 144a was replaced by an ethyl group, which was responsible for the methyl triplet at δ 0.91.
Based on the discussed MS and NMR data and a search in AntiBase, (±)-homononactic acid (145a) was found as the sole coincident structure. The compound was further confirmed by comparing the spectroscopic data with the literature[233] as well as by H,H COSY correlations (Figure 82).
O
Figure 82: H,H COSY ( ) correlations of Homononactic acid (145a).
5.1.5 Homononactic acid methyl ester
Homononactic acid methyl ester (145b) was obtained as a further colourless oil.
Its molecular weight (230 Dalton), the H NMR pattern and the identity of the spec-troscopic data with those of the synthetic one[233] confirmed the structure. Compound 145b is reported here as a natural product for the first time.
5.1.6 Dinactin; Werramycin-B
The major product (>1.5 g) 148 was isolated from fractions I~III as colourless oil. It showed no UV absorbance or fluorescence during TLC, however, it was de-tected as brown zone after spraying with anisaldehyde/sulphuric acid and heating.
Based on its spectroscopic data (MS, 1H NMR, and 13C NMR), and their comparison with literature, 148 was identified as dinactin (werramycin-B).
O
The family of macrotetrolide antibiotics is commonly produced by Streptomyces sp., and is named collectively as nactins. They comprise a series of homologues
based upon a parent 32-membered ring[234,235]. The lowest homologue, nonactin (146), is constructed from four nonactic acid subunits (144a), linked in alternating enantiomeric sequence (+ - + -), so that the overall macrocycle possesses a meso-configuration. Monactin (147), dinactin (148), trinactin (149), tetranactin (150), as well as isodinactin (151) and isotrinactin (152), are examples of further macro-tetrolide antibiotics.
Nonactic acids, the building blocks[236] of macrotetorlides[237], have been chemi-cally synthesized[238-240]. They are currently discussed in the literature, however, were purified always as methyl esters[241]. Nonactic acid and homononactic acid were iso-lated as mixtures of pure (+)- and (-)-enantiomers, but also as racemates or with an excess of the (+)-enantiomers. 2-Epi-homononactic acid was described as the (-)-enantiomer[233].
It is important to mention that feigrisolides A-B (153-154) were first reported by Tang et al.[242], which might be building block of nonactic acid (144a) and homononactic acid (145a), respectively (Figure 83 and Figure 84) were recently ex-cluded by synthesis. The total synthesis of the proposed structure of feigrisolide A (153) was performed by Alvarez-Bercedo et al.[243], however, the synthesized one was structurally not identical with the natural one, indicating that the wrong structure was previously published. Therefore, the reported 153 and 154 were actually 144a and 145a, respectively (i. e feigirisolide A = (+)-nonactic acid and feigrisolide B = (+)-homonanoactic acid).
Figure 83: Correlation between Feigrisolide A (153) and Nonactic acid (144a).
O O C H3
O H
O
H CH3
O
OH O
RO
CH3
CH3
1
3 6
8 10
1
2 5 1'
1'' 2'' 3'
3''
154 145a
Figure 84: Correlation between Feigrisolide B (154) and Homononactic acid (145a).
Biosynthetically, homononactic acid (145a) and nonactic acid (144a) may be con-structed from acetate, propionate or succinate. As a result, a hydroxylated carboxylic acid chain is obtained, which in turn is cyclized to deliver homononactic acid deri-vates (Figure 85).
SCoA O R O
OH
R OH
OH OH
O
R SCoA
OH OH
Homononactic acid derivatives O acetate
propionate
succinate nonactate PKS
Figure 85: Biosynthesis of Nonactic and Homononactic acids (144a-145a).
5.1.7 Biological Activity
The cytotoxic activities of attiamycin A (141), attiamycin B (143a), nonactic acid (144a), homononactic acid (145a) and dinactin (148) are listed in Table 61.
Dinactin (148), attiamycin B (143a) and nonactic acid (144a) showed a highly selec-tive cytotoxicity against a range of human tumor cell lines with a mean IC50 of 0.002 µg/ml (mean IC70 = 0.011 µg/ml), IC50 of 1.160 µg/ml (mean IC70 = 3.423 µg/ml) and IC50 of 1.256 µg/ml (mean IC70 = 3.532 µg/ml), respectively.