Balticidins A–D, Antifungal Hassallidin-Like Lipopeptides from the

Article pubs.acs.org/jnp

Balticidins A−D, Antifungal Hassallidin-Like Lipopeptides from the Baltic Sea Cyanobacterium Anabaena cylindrica Bio33 Thanh-Huong Bui,† Victor Wray,‡ Manfred Nimtz,‡ Torgils Fossen,§ Michael Preisitsch,† Gudrun Schröder,¶ Kristian Wende,† Stefan E. Heiden,|| and Sabine Mundt*,† †

Institute of Pharmacy, Department of Pharmaceutical Biology, Ernst-Moritz-Arndt-University, Friedrich-Ludwig-Jahn-Straße 17, D-17489 Greifswald, Germany ‡ Helmholtz Centre for Infection Research, Inhoffenstraße 7, D-38214 Braunschweig, Germany § Department of Chemistry and Centre for Pharmacy, University of Bergen, Allegt 41, N-5007 Bergen, Norway ¶ Friedrich Loeffler Institute of Medical Microbiology, University Medicine Greifswald, Martin-Luther-Str. 6, D-17475 Greifswald, Germany || Institute of Pharmacy, Department of Pharmaceutical Biotechnology, Ernst-Moritz-Arndt-University, Felix-Hausdorff-Straße 3, D-17489 Greifswald, Germany S Supporting Information *

ABSTRACT: Balticidins A−D (1−4), four new antifungal lipopeptides, were isolated from the laboratory-cultivated cyanobacterium Anabaena cylindrica strain Bio33 isolated from a water sample collected from the Baltic Sea, Rügen Island, Germany. Fractionation of the 50% aqueous MeOH extract was performed by bioassay-guided silica gel column chromatography followed by SPE and repeated reversed-phase HPLC. The main fraction containing the compounds exhibited a strong and specific antifungal activity with inhibition zones in an agar-diffusion assay from 21 to 32 mm against Candida albicans, Candida krusei, Candida maltosa, Aspergillus fumigatus, Microsporum gypseum, Mucor sp., and Microsporum canis. The structures were elucidated by multidimensional 1H and 13C NMR spectroscopy, HRESIMS, amino acid analysis, and sugar analysis. Spectroscopic data analysis afforded an unambiguous sequence of R.CHO(S1).CHOH.CONH-Thr(1)-Thr(2)-Thr(3)-HOTyr(4)-Dhb(5)- D-Gln(6)-Gly(7)-NMeThr(8)(S2)-L-Gln COOH(9), in which Dhb is dehydroaminobutyric acid, S1 is D(−)-arabinose-(3-1)-D-(+)-galacturonic acid, S2 is D-(+)-mannose, and R is the aliphatic residue -C13H26Cl or -C13H27. Besides NMeThr, D-allo-Thr, D-Thr, and L-Thr were identified, but the position of the enantiomers in the sequence is not clear. The four balticidins differ in their cyclic (2, 4)/linear (1, 3) core and the presence (1, 2)/absence (3, 4) of chlorine in the aliphatic unit.

C

produced mainly by Nodularia and Aphanizomenon, the two dominant genera causing blooms in the Baltic Sea.18−23 Thus, the Anabaena spp. from the Baltic Sea probably have been overlooked in earlier cyanobacterial studies because these opensea species were presumed to be nontoxic. However, in the last 10 years, Anabaena spp. from the Baltic Sea have been identified as producers of hepatotoxic microcystins and cytotoxins.24−28 To date, no record has been published of antifungal compounds isolated from Baltic Sea cyanobacteria. In the current study, the 50% aqueous MeOH extract from the biomass of cultured Anabaena cylindrica Bio33 isolated from a sample collected from the Baltic Sea near Rügen, exhibited very specific inhibition activity against different fungi and yeasts. Purification of this extract afforded two linear lipopeptides,

yanobacteria are well-known for the production of toxins by species of the genera Microcystis, Nodularia, Lyngbya, Anabaena, Nostoc, and Oscillatoria.1−4 Nevertheless, it has been shown that cyanobacterial metabolites exhibit a diverse spectrum of biological activities including antibacterial, algicidal, antifungal, antiviral, anticancer, cytotoxic, and enzyme-inhibiting activities. Therefore, cyanobacteria are increasingly being accepted as important sources of potential pharmaceuticals.5−13 Due to increasing fungal infections and limitations of available drugs such as drug safety, resistance, drug−drug interactions, narrow spectrum of activity and effectiveness, the development of new antifungals with a broad fungicidal spectrum of action, good pharmacokinetic properties, and with fewer dose-limiting side effects becomes more and more important.14−17 Since the 1960s, when blooms were in clear evidence in the Baltic Sea, plankton research has been intensively developed. However, cyanobacterial reports focused on the toxins © XXXX American Chemical Society and American Society of Pharmacognosy

Received: December 8, 2013

A

dx.doi.org/10.1021/np401020a | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Chart 1. Structures of Balticidins 1−4

signals of eight amino acid residues were found (Figure S4 Supporting Information). From the characteristic chemical shifts, seven could be identified as Thr (×3), dehydroaminobutyric acid (Dhb), Glx (×2), and Gly (Table 1). One further residue (4 in Table 1) possessed an AA′BB′ spin system that was part of an amino acid system with NH at 8.12 ppm and additional signals at 5.14 and 4.71 ppm. This residue was shown to be β-hydroxytyrosine. These spin systems were confirmed from the Hα and high field region of the TOCSY spectrum. One further amino acid spin system (residue 8) lacking an amide proton was identified as NMeThr from the correlations of its Hα and Hβ at 4.94 and 4.45 ppm, respectively. Although there was insufficient material to record directly 1D 13 C NMR spectra, the availability of a high-field 600 MHz NMR instrument with a cryo-probehead allowed the measurement of 2D HSQC, 2D-edited HSQC, HSQC-TOCSY, and HMBC spectra. Hence, 13C chemical shifts of the protonated carbons of 1 were established directly from the HSQC spectrum (Figure S9). These data confirmed the nature of the amino acid systems present. Strong correlations in the HMBC spectrum (Figure S10) from the methyl groups confirmed the presence of four Thr residues. The nature of residue 4 was also confirmed by the same HMBC spectrum. The shifts of all aromatic carbons followed from the internal correlations with the aromatic protons. The proton (Hβ) at 5.18 ppm showed a direct correlation with the carbon (Cβ) at 74.8 ppm and long-range correlations with an aliphatic carbon (Cα) at 62.6 ppm and aromatic carbons C-1 and C-2/6. In addition, the reverse correlation of H-2/6 with the carbon at 74.8 ppm was also evident, thus confirming residue 4 is a ßhydroxytyrosine unit. Sequence-specific assignments were determined from the cross-peaks in the 2D NOESY spectrum on the basis of short observable distances between Hα and HN of amino acid i and HN of amino acid i + 1 (Figure S6). Confirmatory evidence was provided by the NH-NH correlations in the low field region (Figure S7). The full assignments and chemical shift data are presented in Table 1. The data in Table 2 afforded an unambiguous sequence of the nine residues Thr-Thr-ThrHOTyr-Dhb-Glx-Gly-NMeThr-Glx. The NMe substituent attached to the nitrogen of residue 8 was identified from intraresidue NOEs of the methyl group at 3.08 ppm with those of the Hα and Hβ of the same residue. Confirmatory sequence

balticidins A (1) and C (3), and their cyclic analogues B (2) and D (4). Here we describe the isolation and structure elucidation of these four compounds. Identifying and characterizing the taxonomy of cyanobacteria on the basis of morphological properties that usually vary especially when laboratory cultures are analyzed can be challenging, so genetic data (16S rRNA gene) together with morphology were used in this research to characterize the strain Bio33.

■

RESULTS AND DISCUSSION In a screening for antimicrobial activity, cyanobacteria isolated from water and soil samples collected in North Vietnam near Hanoi, from the Baltic Sea near Rügen and Hiddensee, and also from ponds and water reservoirs in the North of Germany were tested against four bacteria Bacillus subtilis ATCC 6051, Staphylococcus aureus ATCC 6538, Escherichia coli ATCC 11229, Pseudomonas aeruginosa ATCC 27853, and the yeast Candida maltosa SBUG 700 with an agar diffusion assay. Among the tested extracts, the 50% aqueous MeOH extract of strain Bio33 exhibited a strong and specific inhibition activity against the yeast (inhibition zone 22 mm). A bioassay-guided isolation of the extract on silica gel column yielded four fractions (FI, FII, FIII, and FIV) with antifungal activity concentrated in FIII. The purification of fraction FIII by SPE C18-E with 80% MeOH in H2O resulted in the fraction SPE80. Further separation of SPE80 by repeated semipreparative RP-HPLC yielded the balticidins (1−4) as white, amorphous powders after lyophilization (Chart 1). The isolated compounds were characterized by HRESIMS and NMR spectroscopy. Studies in various solvent mixtures indicated the best-resolved 1D 1H NMR spectra were produced in a 1:1 mixture of trifluoroethanol-d2 and H2O. This solvent was then used for the extensive NMR investigation. The positive HRESIMS spectrum of 1 showed the [M + H]+ ion at m/z 1786.764, which is compatible with a molecular formula of C75H120ClN11O36. The combination of 2D 1H NMR techniques with MS data allowed the identification of numerous amino acid residues, but there were also signals for a modified fatty acid and sugars. Initially, spin systems in the homonuclear COSY and TOCSY spectra were identified starting from the signals of the backbone amide protons in the region 9.2 to 6.4 ppm. The B


C

8-NMeThr

7-Gly

6-Gln

5-Dhb

4-β-HOTyrf

3-Thr

2-Thr

1-Thr

residue

NH CO CHα CHβ CH3γ NH CO CHα CHβ CHγ CONH2 NH CO CHα NMe CO CHα CHβ CH3γ

NH COc CHα CHβ CH3γ NH CO CHα CHβ(ring) CH3γ NH CO CHα CHβ CH3γ NH CO CHα CHβ arom

C/H no.

172.9c 43.4 32.8 171.5c 62.9 70.7 15.9

175.3c 55.3 29.4 33.2 180.1c

168.0c 129.4 138.0 14.1

173.5c 62.7 74.6 C1: 133.0, C2,6: 130.2 C3,5: 117.8 C4: 158.3

174.1 62.3 69.2 20.4

175.3 60.8e 69.7 20.8

173.7 62.1 69.4 21.0

δC

1

4.94 4.45 1.19

4.20, 4.13 3.08

4.54 2.26, 2.07 2.42 7.43, 6.58a 8.13

6.78 1.48 7.88

9.09

4.71 5.14 H2/6: 7.37 H3/5: 6.94

4.37 4.21 1.23 8.12

4.48 4.37 1.28 8.42

4.34 4.17 1.30 8.51

8.03

δH (J in Hz)

Table 1. 1H and 13C NMR Data of 1−3 in Trifluoroethanol-d2/H2O (1:1) at 300 K

174.8 43.7 32.2 169.4 62.5 69.9 15.3

175.2 53.3 32.7 32.7 nd

166.1 nd 141.1 14.6

nd 63.1 73.1 134.0 129.8 117.6 158.1

176.7 62.9 69.9 23.8

172.8 57.3 73.2 18.7

169.0 65.0 67.7 23.0

δC

δH

5.06 4.47 1.17

4.31, 3.87 3.03

4.87 2.07, 2.29 2.39, 2.17 7.60, 6.57 8.65

6.78 (7.10) 1.71 7.36

9.37

4.60 5.39 7.40 6.96

4.28 4.17 1.44 8.33

4.70 5.72 attached to 171.7 1.27 8.55

4.01 4.01 1.42 9.48

8.03

2

172.8 43.4 32.8 171.5 62.7 70.7 15.9

175.5 55.3 29.4 33.1 180.1

168.0 129.5 138.1 14.1

173.4 62.7 74.7 132.9 130.2 117.7 158.3

174.1 62.4 69.1 20.4

175.4 61.1 69.7 21.0

173.9 61.9 69.5 20.9

δC

3 δH

4.95 4.43 1.18

4.18, 4.12 3.07

4.53 2.26, 2.06 2.41 7.43, 6.66 8.16

6.77 1.46 7.88

9.11

4.71 5.12 7.36 6.93

4.36 4.20 1.22 8.14

4.48 4.38 1.26 8.42

4.34 4.16 1.30 8.53

8.03

Journal of Natural Products Article


D

C1 C2 C3 C4 C5 C6 C1 C2 C3 C4 C5 C6

C2 C3 C4 C5

NH CO CHα CHβ CHγ CONH2 C1 C2 C3 C4 C5 C6−11 C12 C13 C14 C15 C16 C1

C/H no.

98.4 73.1 72.8 69.1 75.0 nd 102.3 72.8 75.3 72.8 nd nd

68.9 77.4 68.9 64.7

nd 56.9 29.8 33.7 180.0 177.1c 74.6 76.4 29.3 27.3 30.8 39.9 66.8 41.9 20.9 14.1 96.9

δC

4.27 2.14, 1.97 2.29 7.42, 6.64a

7.85

δH (J in Hz)

3.95 4.26 4.05 3.43 (bd, J45A small), 3.34 (bd, J45A small, J5A5B 11.7) 5.03d 3.93 3.84 3.70 3.61 3.93 4.61 (d, J12 7.7) 3.68 3.75 4.29 4.04

4.28 4.07b 1.70−1.75 1.33−1.39 1.31 1.69 3.99 1.69 1.52, 1.42 0.91 5.12 (bs, J12 small)

1

98.7 73.3 72.8 68.5 74.7 62.8 101.3 (72.8) (75.5) 72.8 nd nd

68.8 76.9 69.0 64.7

nd 54.2 31.5 33.8 nd 177.3 74.6 75.5 28.0 27.7 30.8 39.9 66.9 41.9 21.5 14.1 95.6

δC

δH

3.99 (dd, J23 10.2) 4.40 (bd, J34 small) 4.23 3.55 (bd, J45A small), 3.42 (bd, J45B small, J5A5B 11.6) 5.02 (bs, J12 small) 3.81 (bs, J23 small) 3.48 (bd, J34 9.5) 3.75 (dd, J45 9.7) 3.54 3.85 4.70 (d, J12 7.8) 3.69 3.72 4.29e 4.12

4.27 4.14 1.80, 1.67 1.39 1.31 1.71 3.99 1.71 1.53, 1.43 0.92 5.15 (dd, J12 3.9)

4.70 1.87, 1.68 2.24, 2.11 7.48, 6.30

8.74

2 δH

98.4, 98.4 72.9 ×2 72.8 68.9 75.0 62.9 (t) 102.0 72.7 75.1 72.5 77.4 nd

68.8 77.5 68.9 64.7 (t)

3.97 4.25 4.04 3.49 (bd, J45A small), 3.38 (bd, J45A small, J5A5B 11.9) 5.04 3.93 3.84 3.75 3.61 3.93 4.62 3.69 3.76 4.30e 4.12

1.27 1.29 0.87 5.13 (bs, J12 small)

33.3 24.0 14.9 96.8

4.34 2.17, 1.99 2.31 7.43, 6.62

8.02

4.27 4.07 1.71 1.35 ∼1.3

3

nd 56.0 29.3 33.4 180.0 177.2 74.6 76.4 29.3 27.3 30.9

δC

a Interchangeable. bUnambiguous assignment from COSY data. cUnambiguous assignment from HMBC. dMannose unit assigned by comparison with 3. eThe intensities of the cross-peak in the TOCSY spectrum indicates J45 is small. fHydroxy groups are found attached to the ß carbon atom and C-4 of the aromatic system. nd: Not detected.

S1.2: galacturonic acid unit

S2: mannose unit attached to 8-NMeThr

S1.1:ß-arabinose attached to CβHO of 10-DhA

10-DhA

9-Gln

residue

Table 1. continued

Journal of Natural Products Article



Article

composition remains constant under the more extreme conditions, clearly racemization is not occurring; hence, the “inner” Gln must have the D-configuration. This hypothesis that no racemization occurs is further justified as only one enantiomer of NMeThr appeared under relatively mild conditions and is relatively easily cleaved in MS/MS. The chirality of NMeThr is unclear as reference material was not available. The data for Thr is more complex. Independent of hydrolysis conditions, both D- and L-forms of Thr were detected. In addition, a further single peak of the same intensity as the L-Thr peak was present and showed a similar fragmentation pattern to those of Thr. This peak was determined to be D-allo-Thr by comparison with a standard and with previous work.29 As four Thr residues were detected after hydrolysis although only three were observed in the NMR spectra a racemic mixture is formed under the nonenzymatic hydrolysis conditions and one residue each of D-Thr, L-Thr, and D-allo-Thr is present as in the hassallidins.29,30 The positions of the different enantiomers of Thr in the amino acid sequence have not been clarified. This could only be determined by X-ray crystallography or by total synthesis; however, crystallization of the molecule has not been successful to date. The absence of reference material prevented the enantioselective analysis of β-OHTyr. A literature search using partial sequences revealed two closely related cyclic peptides, hassallidins A and B,29,30 in which the same sequence was present apart from the replacement of the β-OHTyr unit by Tyr. In the hassallidins, ring closure occurred through the formation of a lactone bridged between the hydroxy group of the second Thr in the sequence and the carboxylic acid of the terminal Gln. This was identified by the downfield acylation shift of ∼0.8 ppm in the 1 H NMR spectrum of Hβ and ∼4−5 ppm upfield shift of Cγ of the substituted Thr compared to the other threonines in the molecule. This was not the case in our data of 1, so that a linear structure of the peptide core was proposed. The absence of the typical absorption for esters/lactones at 1732 cm−1 in the IR spectrum of 1 (Figure S12) but present for hassalidin A and B29,30 confirmed our assumption. The 1D and 2D NMR spectra of 1 showed a considerable number of spin systems that did not belong to the amino acids. A dihydroxy aliphatic acid derivative (10-DhA) with signals in both the region 5.5 to 3.0 ppm (Table 1) and high field aliphatic region was present. The linkage between the fatty acid moiety and the first 1-Thr was determined by the cross-peak at

Table 2. Sequence and Structural Information of the Peptide Moiety Deduced from the NOEs Found in the 2D NOESY Spectrum of 1 αH-NH (i, i + 1)

NH-NH (i, i + 1)

1−2 2−3 3−4 4−5 6−7a 8−9 10−1b

1−2 2−3 3−4 4−5 5−6 6−7

others

4H2/6-5NH 5Hß-6NH 8NMe-8Hα 8NMe-8Hß 8Hα-9Hα

a Not observed in the series of spectra due to irradiation of the αH under the H2O signal. It was observed under different conditions in other spectra. bThere are also correlations of 10ß with 1NH.

correlations exist between 5-NH and H-2/6 of residue 4 (βHOTyr) and 6-NH and Hβ of residue 5 (Dhb). The strong intraresidue NOE between Hγ (Me) and NH of the Dhb defines the Z configuration of the double bond of residue 5. The quality of the HMBC spectrum was sufficient to establish the assignment of the carbonyl groups. Hence, correlations from the Hα and Hβ protons allowed direct assignment of the intraresidue carbonyl carbons, whereas correlations of the NH protons afforded the assignment of the carbonyl carbons of the preceding residue in the sequence. These data again yielded independent sequence information that was compatible with the result from the NOE data. The amino acid sequence of 1 was confirmed from a detailed analysis of the b and y ions of the high-resolution ESIMS data of the peptide fragments from the MS2 of the [M+2H]2+ ion in Figure 1 (Table S2 Supporting Information). Finally, amino acid analysis verified the presence of Gly, Glx, Thr, NMeThr, and allo-Thr in the peptide unit. The NMR spectrum (Figure S11) showed that both Glx units are Gln. This could be confirmed by the strong correlations between the CONH2 protons and Hβ and Hγ observed in the ROESY spectrum (Figure S24). When 1 was hydrolyzed at 120 °C for 12 h and the enantiomeric labeled total hydrolysate was analyzed by chiral-phase GC-MS, D- and L-enantiomers of Gln in a ratio of 1:1 were found. Under mild conditions (80 °C for 2 h) only the L-enantiomer of Gln was present. As one Gln is a terminal residue, and according to MS/MS experiments readily cleaved, this residue must be the L-enantiomer. As the D- and L-

Figure 1. MSn fragmentation ions from the ESIMS spectra of 1/3 that confirm the peptide sequence. Denoted fragments a−o are specified in Table S2 in the Supporting Information. E



Article

δ 8.03/177.1 (NH 1-Thr/C1) and the weak cross-peak at δ 4.34/177.1 (Hα 1-Thr/C1) observed in the 2D 1H−13C HMBC spectrum. The downfield shift of H-2 (δ 4.28), C-2 (δ 74.6), H-3 (δ 4.07), and C-3 (δ 76.4) showed that these positions were hydroxylated. H-2 was identified by the crosspeaks at δ 4.28/177.1 (H-2/C-1), δ 4.28/76.4 (H-2/C-3), and δ 4.28/29.3 (H-2/C-4), also from the HMBC spectrum. H-3 was assigned by the cross-peak at δ 4.28/4.07 (H-2/H-3) observed in the 2D 1H COSY and the 2D 1H NOESY spectra, respectively. The terminal propyl group of the fatty acid moiety was established by the cross-peaks at δ 0.91/1.52 (H-16/H-15a), δ 0.91/1.42 (H3-16/H-15b), δ 1.52/1.69 (H-15a/H2-14), and δ 1.42/1.69 (H-15b/H2-14) observed in the 2D COSY spectrum and the cross-peaks at δ 0.91/20.9 (H3-16/C-15), δ 0.91/41.9 (H3-16/C-14), δ 1.52/14.1 (H-15a/C-16), δ 1.42/14.1 (H15b/C-16), δ 1.52/41.9 (H-15a/C-14), and δ 1.42/41.9 (H15b/C-14) observed in the HMBC spectrum. The multiple bond correlations at δ 1.52/66.8 (H-15a/C-13), δ 1.42/66.8 (H-15b/C-13), δ 1.69−1.70/66.8 (H2-14/C-13 and H2-12/C13), δ 3.99/41.9 (H-13/C-14), and δ 3.99/20.9 (H-13/C-15) observed in the HMBC spectrum showed that the terminal propyl group is attached to a methine carbon at δ 66.8 (C-13), which is consistent with substitution of a chlorine atom at this position. This identification of the fatty acid moiety was consistent with the fact that the mass spectrometric data supported the presence of a chlorinated dihydroxy C-16 fatty acid. In the 1H NMR spectra of 1, there were a number of systems in the region 5.2 to 3.3 ppm, with no corresponding high-field aliphatic signals, which clearly belonged to sugar-like residues (Figures S5a−c). This was in agreement with a sugar compositional analysis which indicated the presence of mannose, arabinose, and galacturonic acid. In the NMR spectra, two systems S1 and S2 were found in which the lowest field signals (anomeric protons at 5.12, 5.03, and 4.60 ppm, respectively) correlated with characteristic anomeric carbons in the region 96−103 ppm. The lowest field anomeric proton (S1, 5.12 ppm) was part of a five-membered spin system and showed strong NOE signals to the γCH2 group of the 10DhA unit, indicating this was substituted at C-3 of this unit. The magnitude of the coupling constants, particularly the presence of a geminal pair, indicated this belonged to a βarabinose system (Figure S5a). A further unit showed an anomeric proton at 4.60 ppm with a vicinal coupling of 7.7 Hz characteristic of a chair conformation of a β-pyranose system and a readily identified spin system in the TOCSY spectrum consisting of five protons (Figure S5c). This was only compatible with the galacturonic acid found in the sugar analysis. The position of this unit at C-3 of arabinose follows from the observation of NOEs from the anomeric proton (4.62 ppm) to H-3 and H-4 of the β-arabinose system (4.25 and 4.05, respectively) in the NOESY spectrum and of its correlation with C-3 of arabinose in the HMBC spectrum (Figure S32), so that S1 was identified as a disaccharide unit bound to the γCH2 group of the 10-DhA. The second sugar unit, S2, with an anomeric proton at 5.03 ppm was attached at Cβ of 8-NMeThr from its correlation in the NOESY spectrum with the signal at 4.45 ppm. The sugar analysis and NMR data (Figure S5b) indicated this unit belonged to a mannose system. The HRESI/MS/MS spectrum showed the [M + 2H]2+ ion underwent readily sequential loss of a hexose, pentose, and

hexuronic acid units and was compatible with the sugar compositional analysis, which indicated these units corresponded to mannose, arabinose, and galacturonic acid. Enantioselective carbohydrate analysis revealed the monosaccharide residues as D(+)-mannose, D(−)-arabinose, and D(+)-galacturonic acid. The NMR spectra, HRESIMS data, and sugar analysis indicated the peptide sequence and the glycosylation pattern were the same in each of the compounds 1−3, but in contrast to 1 the low field shift of the Hβ of 2-Thr in 2 indicated that ring closure had taken place with the carboxylic acid of the terminal Gln as in hassallidins A and B.29,30 Furthermore, the IR spectrum of 2 (Figure S21) showed absorption at 1737 cm−1 close to the absorption given for hassallidin A29 (1740 cm−1) and typical for the presence of an ester/lactone structure. As discussed above, the NMR spectra clearly showed that both Glx units are Gln, so no side chain carboxylate was present, and the lactone could only be formed with the C-1 acid. Variable temperature−time analysis of the enantiomeric labeled total hydrolysate of 2 by chiral-phase GC-MS indicated an identical ratio of D- and L-Gln as detected in 1. The aliphatic side chain was the same as identified in 1 from the comparison of the NMR and MS data, so that 2 represents the cyclic analogue of 1. The comparison of NMR spectra of 3 (Figures S23−S32) with those of 1 revealed that there were differences in the signals of the aliphatic chain. Although the carboxyl terminal region was identical in the HMBC spectrum, the signals indicated a heteroatom was absent. A detailed comparison of the HRESI/MS/MS data of [M + 2H]2+ ions for 1 and 3 (Table S2) indicated a hydrogen instead of chlorine in the side chain attached to the 1-Thr as the only difference between these two molecules. The absence of a low field shift of the Hβ of 2-Thr in the NMR spectrum (Figure S26) together with no absorption at 1740 cm−1 in the IR spectrum typical for esters/ lactones indicated the linear structure of 3. As limited amounts of 4 prevented the accumulation of NMR data of sufficient quality, the structure of 4 was deduced from the fragmentation pattern of the HRESIMS spectrum (Figures S38−39, Table S3) using the monoisotopic mass ion at m/z 867.903 [M + 2H]2+ as the precursor ion, which is compatible with a molecular formula of C75H119N11O35. The mass difference of 34 between 4 and 2 indicates an exchange of chlorine against a hydrogen atom at C-13 of the fatty acid residue as already described for compounds 1 and 3. Furthermore, the IR spectrum of 4 (Figure S35) showed an absorption at 1732 cm−1 in the range typical for esters/lactones and similar to the absorption of hassallidin A29 (1740 cm−1) and balticidin B (1737 cm−1), so that a cyclic structure is proposed and 4 represents the cyclic analogue of 2. The specific antifungal activity of 1−4 against Candida maltosa was confirmed in an agar diffusion test. Amounts of 10 μg/6 mm paper disc of 1, 2, 3, and 4 resulted in inhibition zones of 12, 15, 9, and 18 mm, respectively. No inhibition activity against Bacillus subtilis, Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa was detected (Table S1). The hassallidins are esterified eight-residue cyclic peptides linked with a fatty acid residue (dihydroxytetradecanoic acid). Hassalidin A contains the sugar mannose, and hassallidin B differs from A by an additional carbohydrate unit, rhamnose, attached at C-3 of the dihydroxy fatty acid. Both compounds exhibit a broad spectrum of antifungal activities but no antibacterial activities. The MICs of hassallidins A and B F



Article

Figure 2. Phylogenetic tree based on a secondary structure alignment containing 45 partial 16S rRNA gene sequences. The tree was inferred using a maximum likelihood method. Numbers given on the branches display bootstrap proportions as percentage of 1000 replicates for values greater than or equal to 50%. The accession numbers of sequences used for reconstruction are given in parentheses (for accession numbers that represent a whole genome, the genomic location is provided). Sequences used as outgroups (B. subtilis DSM 10 (AJ276351), E. coli ATCC 25922 (DQ360844)) are not shown.

against different Candida species were estimated at 4 and 8 μg/ mL and 8 and 16 μg/mL, respectively.29−31 The two cyclic lipopeptides balticidins B (2) and D (4) have the same core as the hassallidins with a mannose attached to NMeThr except for 4-β-OHTyr taking the place of Tyr. In contrast to the hassallidins, all of the balticidins contain a dihydroxyhexadecanoic acid side chain and a disaccharide formed by D(−)-arabinose and D(+)-galacturonic acid linked to the 3-OH of the fatty acid residue. In addition, the occurrence of chlorine in the fatty acid unit of balticidin B (2) and in the linear balticidin A (1) is a further difference between these compounds and the hassallidins. On the other hand, structural analysis revealed 1 and 3 as linear lipopeptide analogues of 2 and 4. To exclude the possibility that the linear compounds are hydrolysis artifacts derived from the cyclic forms during isolation by acidic silica column chromatography, the aqueous MeOH extract was analyzed in LCMS, and the masses of the compounds were detected (Figure S37). Furthermore, after isolation, the purity of 1−4 was checked by analytical HPLC

with 0.1% TFA/CH3CN, and no decomposition products have been observed, so that the cyclic forms seem to be stable under the acidic HPLC conditions that were used. Nevertheless, enzymatic hydrolysis during lyophilization of the biomass or extraction process cannot be fully ruled out. As yet, the balticidins are the first cyanobacterial lipopeptides containing an oligosaccharide moiety attached to a halogenated fatty acid. Hassallidins A and B were isolated from an epilithic cyanobacterium identified as a Tolypothrix species (basionym Hassallia) according to Geitler’s characterization and taxonomy scheme.29,30 However, no phylogenetic analysis of this Hassallia species has been published to date. For a correct taxonomic classification of strain Bio33, a planktic brackish-water cyanobacterium from the Baltic Sea, the morphology was investigated and was in accordance with the current description of Anabaena cylindrica. This species shows straight and solitary unbranched trichomes, the vegetative cells are cylindrical, 3−4 μm wide and 3−5 μm long, and the apical cells are rounded. The heterocysts are long and rounded, 5 μm wide and 6−8 μm G



Article

long, and intercalary in the filament.32−34 Comparing the nucleotide sequence obtained from sequencing the 16S rDNA of Bio33 with sequence information available in the National Center for Biotechnology Information database using BLAST (http://www.ncbi.nlm.nih.gov/BLAST) led to the sequence of Anabaena cylindrica as the best hit (99% max identity). Recently, a putative hassallidin biosynthetic gene cluster was identified in the genome of the freshwater Anabaena sp. strain 90, and in MS analysis, ions have been detected in the range between m/z 1200−1900, indicating the presence of hassallidins.35 In order to infer the phylogenetic position of our strain with the proposed producer genera of the hassalidins, Tolypothrix and Hassallia, and other Anabaena spp., a phylogenetic tree (Figure 2) was constructed on the basis of published partial 16S rRNA gene sequences of these genera using a maximum likelihood method. Using the taxonomic scheme of the cyanobacteria according to NCBI Taxonomy Browser (August 31, 2009), the order Nostocales is separated into four families Microchaetaceae (containing Hassallia, Tolypothrix), Nostocaceae (containing Anabaena), Rivulariaceae, and Scytonemataceae. From the phylogenetic tree, it is clear that Anabaena belongs to a separate cluster compared to Tolypothrix and Hassallia spp., respectively. It seems that Tolypothrix and Hassallia belong to a mixed cluster in the Microchaetaceae. It has already been indicated that Tolypothrix is a problematic taxon, as species of this genus are dispersed among other Microchaetacean genera such as Spirirestis, Hassallia, Rexia, and Coleodesmium.36 From the molecular genetic data, it appears that strain Bio33 belongs to the Anabaena cluster in the family Nostocaceae and is completely distinguished from the published Tolypothrix and Hassallia spp. No 16S rRNA gene nucleotide sequences of the hassallidinproducing strains have been published so that an incorrect morphological identification cannot be fully ruled out.

■

phases. Detection was performed under UV light at 254 nm or by spraying with anisaldehyde/sulfuric acid reagent and heating. The further purification was carried out with STRATA C18-E Giga tubes (55 μm, 70 Å, 5 g/20 mL, Phenomenex). Analytical and semipreparative HPLC were performed on a component system (Kontron Instruments), consisting of pumps 422 and 422 S, auto sampler 360, and diode array detector DAD 440. A Gemini C18 column (250 × 4.6 mm, 5 μm, 110 Å, Phenomenex) with a gradient of MeOH (HPLC grade, Merck) in deionized H2O (HPLC grade, Merck) plus 0.05% TFA and a LiChrospher RP-18 column (250 × 4 mm, 5 μm, 100 Å, Merck) with a gradient of CH3CN (HPLC grade, Merck) plus 0.05% HCOOH (Merck) in H2O plus 0.05% HCOOH were used for isolation of the pure compounds. HPLC runs were recorded using the Geminyx HPLC data system 1.91 SST version 1.6. All chemicals were used as received, and solvents were distilled before use except for HPLC. Culture Conditions. The cyanobacterium Anabaena cylindrica strain Bio33 was isolated from the Baltic Sea (2001, Rügen Island, Germany) and established as a laboratory culture. The strain is maintained in the culture collection of the Institute of Pharmacy, EMAU Greifswald, as a stock culture. The cyanobacterium was cultured in a glass column containing 35 L of BG 11 medium plus 0.5% NaCl under continuous fluorescent light (20 μmol m−2 s−1), the temperature was maintained at 22.5 °C, and the pH was adjusted to 8.5 using CO2 supplementation.37 After 35 days, the cells were harvested by centrifugation at 4000g, 10 °C in a continuous flow centrifuge (Stratos, Heraeus Instruments), lyophilized, and kept at −20 °C until used. The yield of lyophilized biomass was 0.7 g L−1. Extraction and Isolation. Lyophilized cells (7.5 g) of strain Bio33 were extracted in portions of 2.5 g three times with 250 mL of nhexane followed by MeOH/H2O (1/1) with stirring for 2 h, respectively. After centrifugation at 4000g at 10 °C for 10 min, the aqueous methanolic supernatants were pooled and evaporated to provide an extract of about 2.5 g. The aqueous methanolic extract was separated in portions of 500 mg on silica gel (open column, 3 × 400 mm, flow rate 0.5 mL min−1) using a stepwise gradient of EtOAc/ MeOH/H2O [200 mL EtOAc/MeOH/H2O (7/2/1) (FI), 150 mL EtOAc/MeOH/H2O (5/3.5/1.5) (FII), 500 mL EtOAc/MeOH/H2O (1/7/2) (FIII) and 400 mL MeOH/H2O (1/1) (FIV)]. Fraction FIII (yield 100 mg/500 mg extract) exhibited the highest antifungal activity. FIII (500 mg) was further purified by STRATA C18-E Giga tubes (55 μm, 70 Å, 10 g/50 mL) with a step gradient of 5% MeOH, 45% MeOH (yield 101.6 mg), 80% MeOH (yield 48.2 mg), and 100% MeOH in H2O. The analytical HPLC analysis of the SPE 80% MeOH fraction performed with a Gemini C18 column (250 × 4.6 mm, 5 μm, 110 Å, Phenomenex) and a gradient of MeOH in deionized H2O plus 0.05% TFA with a flow rate of 1.0 mL min−1 from 40% to 60% MeOH in 4 min, from 60% to 70% MeOH in 3 min and from 70% to 100% MeOH in 23 min yielded four main peaks FIII-4 (tR = 19.35 min), FIII-5 (tR = 20.00 min), FIII-6 (tR = 21.15 min), and FIII-7 (tR = 21.83 min). The collection of the four peaks was carried out by semipreparative HPLC (200 μg/injection) with a LiChrospher RP18e column (250 × 4 mm, 5 μm, 100 Å, Merck) and a flow rate of 1.0 mL min−1. A gradient of MeOH in H2O plus 0.05% TFA from 40% to 60% MeOH in 2.5 min, from 60% to 70% MeOH in 3 min, followed by 74% MeOH over 23 min was used. The separation was controlled by DAD at 226 nm. The pure compounds 2, 3, and 4 were collected directly as FIII-5, FIII-6, and FIII-7 eluted at tR = 17.45 min, tR = 21.86 min, and tR = 24.33 min, respectively. Fraction FIII-4 (10 μg/20 μL injection) was further purified with a LiChrospher RP-18e column (4 × 250 mm, 5 μm, 100 Å, Merck) and the same gradient of mobile phase as in analytical HPLC yielded the pure compound 1 at tR = 17.98 min. In parallel, the fraction FIII-4 was purified with a LiChrospher RP-18e column (250 × 4 mm, 5 μm, 100 Å, Merck) and a gradient of CH3CN (HPLC gradient grade, Merck) in H2O (both supplemented with 0.05% HCOOH) from 35% to 37% over 4.5 min, followed by 37% in 10 min and from 37% to 40% in 3 min at a flow rate of 1 mL min−1. Using these conditions, the pure compound 1 eluted at tR = 16.40 min. From 48.2 mg of the 80% MeOH SPE

EXPERIMENTAL SECTION

General Experimental Procedures. UV spectra were measured on a LC-20a prominence UFLC (SHIMADZU Corporation) with a diode array detector SPD-M20A. CD spectra were determined on a J810 CD spectropolarimeter (JASCO GmbH) in the wavelength-range from 200−300 nm at 20 °C and were analyzed with Spectra Manager Software (Version 1.53.01). The IR spectra were recorded using a NICOLET IR200 Fourier transform infrared spectrometer (FT-IR) (Thermo Electron Corporation). 1D and 2D NMR spectra were recorded at 300 K on a Bruker AVANCE DMX 600 NMR spectrometer (COSY, TOCSY, NOESY, HMQC, and HMBC) and a Bruker AVANCE 600 NMR spectrometer (COSY, TOCSY, NOESY, HSQC, 2D-edited HSQC, HSQC-TOCSY, and gHMBC) equipped with an UltraShield Plus magnet and a triple resonance cryoprobe with a gradient unit. Each spectrometer was locked to the deuterium resonance of the solvent, DMSO-d6 or trifluoroethanol-d2/ H2O (1:1). Chemical shifts were referenced to the residual signals of the respective solvents (DMSO 1H: 2.50/13C: 39.5 ppm and CF3CDHOH 1H: 3.95/13C: 61.5 ppm, respectively). For ESI MSn analysis, the compounds were analyzed on an Orbitrap Velos mass spectrometer (Thermo-Scientific) equipped with a nanospray ion source and an external high-energy collision cell. MS/MS spectra of the doubly protonated molecular ions were recorded at a resolution of 60 k and with an average mass error of under 2 ppm. The most probable isotopic compositions of all major fragment ions were determined, followed by their manual structural assignment. If necessary, MS3 and higher order experiments were performed for corroboration of structural assignments. Column chromatography was carried out on silica gel (Si 60, 0.040−0.063 mm, Merck). Fractions were monitored by TLC (Si 60 GF 254 nm, Merck) with EtOAc/ MeOH/H2O (5/3.5/1.5) or EtOAc/MeOH/H2O (1/7/2) as mobile H



Article

fraction 4.2 mg of 1, 1.2 mg of 2, 2.0 mg of 3, and 0.8 mg of 4 were collected as white, amorphous solids. Balticidin A (1). White, amorphous powder; UV (CH3OH/H2O 1/ 1) λmax 224 and 272 nm; CD (c 17.5 μM; TFE/H2O, 1/1) λmax (Δε) 220.5 (−23.97) nm; IR νmax (film) 3191, 2939, 2841, 1662, 1411, 1190, 1140, 1085, 1027, 968, 922, 882, 799, 722, 674, 640, 546, 471 cm−1; 1H and 13C NMR in CF3CD2OH/H2O (1/1), see Table 1; MSMS fragmentation, see Table S2; HRESIMS m/z [M + H]+ calcd for C75H121ClN11O36, 1786.766 [M + H]+; found 1786.764; m/z [M + 2H]2+ 893.885. Balticidin B (2). White, amorphous powder; UV (CH3OH/H2O 1/ 1) λmax 226 and 273 nm; CD (c 8.55 μM; TFE/H2O, 1/1) λmax (Δε) 220 (−28.65) nm; IR νmax (film) 3382, 3262, 3201, 2929, 2365, 1737, 1660, 1628, 1616, 1563, 1537, 1512, 1456, 1431, 1405, 1241, 1202, 1135, 1077, 1008, 973, 916, 820, 802, 773, 669, 645, 590, 520 cm−1; 1 H and 13C NMR in CF3CD2OH/H2O (1/1), see Table 1; MS-MS fragmentation, see Table S3; HRESIMS m/z [M + H]+ calcd for C75H119ClN11O35, 1768.755; found 1768.753; m/z 884.883 [M + 2H]2+. Balticidin C (3). White, amorphous powder; UV (CH3OH/H2O 1/ 1) λmax 227 and 274 nm; CD (c 28 μM; TFE/H2O, 1/1) λmax (Δε) 221 (−6.97) nm; IR νmax (film) 3389, 3309, 2926, 2841, 1654, 1612, 1518, 1454, 1407, 1203, 1135, 1073, 975, 922, 839, 800, 725, 669 cm−1; 1H and 13C NMR in CF3CD2OH/H2O (1/1), see Table 1; MSMS fragmentation, see Table S2; HRESIMS m/z calcd for C75H122N11O36, 1752.805 [M + H]+; found 1752.809; m/z [M + 2H]2+, 876.906. Balticidin D (4). White, amorphous powder; UV (CH3OH/H2O 1/ 1) λmax 224 and 273 nm; CD (c 22.4 μM; TFE/ H2O, 1/1) λmax (Δε) 220.5 (−9.43) nm; IR νmax (film) 3465, 3358, 3262, 2924, 2853, 1732, 1656, 1532, 1422, 1384, 1261, 1238, 1201, 1136, 1065, 1033, 1006, 975, 912, 840, 801, 721 cm−1; MS-MS fragmentation, see Table S3; HRESIMS m/z [M + H]+ calcd for C75H120N11O35, 1734.794; found 1734.798; m/z [M + 2H]2+, 867.903. Carbohydrate Compositional Analysis. Monosaccharides were analyzed as the corresponding methyl glycosides after methanolysis and trimethylsilylation by GC/MS.38 The absolute configurations of the monosaccharides were determined by separation of the trimethylsilylated S-(1)-but-2-yl glycosides.39 Enantioselective Analysis of Amino Acids. A sample of 1 and 2, respectively, were hydrolyzed using 4 N TFA under various temperature−time regimes (80 °C: 2 and 5 h; 100 °C: 3 h), conditions that resulted in the conversion of Gln to Glu. After drying, the resulting free amino acids were derivatized with 4 N HCl/propan-2-ol (1 h, 110 °C) and, after removal of reagents, the amino acid isopropyl esters were then acylated with pentafluoropropionic acid anhydride in CH2Cl2 (150 °C, 12 min). Excess reagents were again removed, and the amino acid derivatives were analyzed on a Chirasil Val column (50 m) connected to a GCQ ion trap mass spectrometer. The constituent amino acids were identified by their characteristic mass spectra, and their chirality was determined by comparison to standard D, L amino acids. The following retention times were estimated: D-Glu 18.57 min (sample), 18.56 min (standard); L-Glu 18.73 min (sample), 18.73 min (standard); D-Thr 12.64 min (sample), 12.64 min (standard); D-alloThr 10.37 min (sample), 10.38 (standard); L-allo-Thr 10.58 (standard). Analysis of D-allo-Thr/L-allo-Thr was run at different conditions, so that retention times cannot be compared with those given for the other amino acids. Antifungal Assay. The antifungal activity of the main fraction FIII (1.0 mg/6 mm paper disc) was determined against Candida albicans ATCC 90028, Candida krusei ATCC 90878, Candida maltosa SBUG 700, and clinical isolates of Aspergillus fumigatus, Microsporum gypseum, Microsporum canis, and Mucor sp. (culture collection of the Friedrich Loeffler Institute of Medical Microbiology, University Medicine Greifswald) with the agar diffusion method according to the Clinical and Laboratory Standards Institute (CLSI). The inhibition zones were measured including the diameter of the paper disc. The minimal inhibitory concentration (MIC) of the main fraction FIII was estimated against Candida maltosa SBUG 700 (Institute of Micro-

biology, EMAU Greifswald) by the broth dilution test following the guidelines of the European Pharmacopoeia.40 Morphological Characterization and Identification. Cell dimensions (length and width) of vegetative cells and heterocysts of different filaments of fresh cultures growing in late exponential phase were measured. Visual examination was carried out in N-free liquid BG 11 medium with 0.5% NaCl in an inverted research microscope (Axioskop 2 plus, Carl Zeiss) with coupled digital camera (AxioCam MRc camera, software Axiovision version 4). Taxonomic identification was based on standards in the literature.33,34 Amplification of l6S rRNA Genes. Amplification was carried out by PCR using universal PCR primers GM3F (5′-AGAGTTTGATCMTGGC-3′, corresponding to nucleotide positions 8−24 of 16S rRNA of E. coli) and GM4R (5′-TACCTTGTTACGACTT-3′, corresponding to nucleotide positions 1492−1507 of 16S rRNA of E. coli) annealing to the 5′ and 3′ end of the 16S rRNA gene, respectively.41 The PCR mixture contained 5 μL of 10× Polymerase Buffer C, 1 μL of dNTPS (10 mM), 0.25 μL of each primer, 0.25 μL of OptiTaq (5 U/μL) polymerase, and 0.5 μL of purified DNA (103.1 ng/μL). A volume of 42.75 μL nuclease-free-H2O was added to give a total reaction volume of 50 μL. PCR was conducted using MJ Mini Personal Thermal Cycler (Bio-Rad Laboratories). PCR conditions were as follows: an initial hold at 95 °C for 5 min, 34 cycles of denaturation at 95 °C for 30 s, annealing at 45 °C for 30 s, and elongation at 72 °C for 1 min 45 s. A final elongation of 72 °C for 7 min was used. PCR products were verified by electrophoresis on 1.5% agarose gels in TBE buffer. PCR products were purified with the QIAquick Gel Extraction Kit (QIAGEN GmbH) with solutions provided by the manufacturer. Gel-purified PCR products were used for sequencing. Sequence Alignment and Tree Construction. The amplicon was sequenced by Eurofins MWG Operon using primers GM1F42/ 907RM43 and their reverse complements GM1R/907FK. Single sequencing reads were assembled in Geneious version 6.0.3 created by Biomatters (available from http://www.geneious.com). The 16S rDNA sequence of Bio33 was automatically aligned according to the SILVA SSU ref NR99 r115 database (available from http://www.arbsilva.de)44 using the Silva INcremental Aligner (SINA) version 1.2.11.45 Related sequences missing from SSU ref NR99 r115 were obtained from SILVA Web release 117 and imported into the ARB software package version 5.5.46 The alignment was manually refined taking into account the secondary structure information on the rRNA and the bacterial Positional Variability by Parsimony (PVP) filter was applied while exporting from ARB. Phylogenetic reconstruction was performed with 45 sequences using a maximum likelihood method. The final tree was calculated with RAxML version 8.0.14 (GTRGAMMA model)47 and based on 558 distinct alignment patterns. It is the best tree out of 1000 independent inferences (Figure 2). The sequence of strain Bio33 is available from the INSDC databases (i.e. DDBJ, EMBL, and GenBank) under accession number KC333872.

■

ASSOCIATED CONTENT

S Supporting Information *

1D and 2D NMR spectra of 1−3; MS, IR, and CD spectra of 1−4; and more detailed information on biological tests. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: +49-3834-864885. Tel.: +49-3834-864869. Notes

The authors declare no competing financial interest. I



■

Article

(30) Neuhof, T.; Schmieder, P.; Seibold, M.; Preussel, K.; von Döhren, H. Bioorg. Med. Chem. Lett. 2006, 16, 4220−4222. (31) Neuhof, T.; Seibold, M.; Thewes, S.; Laue, M.; Han, Ch.-O.; Hube, B.; von Döhren, H. Biochem. Biophys. Res. Commun. 2006, 349, 740−749. (32) Galhano, V.; Figueiredo, D. R.; de Alves, A.; Correia, A.; Pereira, M. J.; Gomes-Laranjo, J.; Peixoto, F. Hydrobiologia 2011, 663, 187− 203. (33) Geitler, L. Cyanophyceae. In Rabenhorst’s Kryptogamenflora von Deutschland, Oesterreich und der Schweiz, 2nd ed.; Kolkwitz, R., Ed.; Akademische Verlagsgesellschaft: Leipzig, 1932. (34) Komarek, J.; Anagnostidis, K. Archiv Hydrobiol. Suppl. 1989, 82, 247−345. (35) Wang, H.; Sivonen, K.; Rouhiainen, L.; Fewer, D. P.; Lyra, C.; Rantala.Ylinen, A.; Vestola, J.; Jokela, J.; Rantasärkkä, K.; Li, Z.; Liu, B. BMC Genomics 2012, 13, 613−630. (36) Vaccarino, M. A.; Tillmaand, E. G.; Martin, M. P.; Johansen, J. R. Phylogenetic placement of the Microchaetaceae (Cyanobacteria); Presented at Plant Biology, Honolulu, Hawaii, 2009, July 18−22. (37) Mundt, S.; Kreitlow, S.; Nowotny, A.; Effmert, U. Int. J. Hyg. Environ. Health 2001, 203, 327−334. (38) Chaplin, M. F. Anal. Biochem. 1982, 123, 336−341. (39) Gerwig, G. J.; Kamerling, J. P.; Vliegenthart, J. F. G. Carbohydr. Res. 1978, 62, 349−357. (40) European Pharmacopoeia. Microbial Assay of Antibiotics, 5th ed., Council of Europe (COE), European Directorate for the Quality of Medicines (EDQM): Strasbourg, France, 2005; Vol. 5.3, p 188. (41) Muyzer, G.; Teske, A.; Wirsen, C. O.; Jannasch, H. W. Arch. Microbiol. 1995, 164, 165−172. (42) Muyzer, G.; De Waal, E. C.; Uitterlinden, A. G. Appl. Environ. Microbiol. 1993, 59, 695−700. (43) Muyzer, G.; Brinkhoff, T.; Nübel, U.; Santegoeds, C.; Schäfer, H.; Wawer, C. In Molecular Microbial Ecology Manual; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1998; pp 1−27. (44) Yilmaz, P.; Parfrey, L. W.; Yarza, P.; Gerken, J.; Pruesse, E.; Quast, C.; Schweer, T.; Peplies, J.; Ludwig, W.; Glöckner, F. O. Nucleic Acids Res. 2014, 42, D643−D648. (45) Pruesse, E.; Peplies, J.; Glöckner, F. O. Bioinformatics 2012, 28, 1823−1829. (46) Ludwig, W.; Strunk, O.; Westram, R.; Richter, L.; Meier, H.; Yadhukumar; Buchner, A.; Lai, T.; Steppi, S.; Jobb, G.; Förster, W.; Brettske, I.; Gerber, S.; Ginhart, A. W.; Gross, O.; Grumann, S.; Hermann, S.; Jost, R.; König, A.; Liss, T.; Lüßmann, R.; May, M.; Nonhoff, B.; Reichel, B.; Strehlow, R.; Stamatakis, A.; Stuckmann, N.; Vilbig, A.; Lenke, M.; Ludwig, T.; Bode, A.; Schleifer, K. Nucleic Acids Res. 2004, 32, 1363−1371. (47) Stamatakis, A. Bioinformatics 2014, 30, 1312−1313.

ACKNOWLEDGMENTS We thank the MOET of Vietnam for a scholarship enabling the stay of T.-H.B. at the Ernst-Moritz-Arndt University Greifswald for her Ph.D. studies, the DAAD, and the University of Greifswald for financial support. Further we thank B. Cuypers (EMAU Greifswald, Germany) for providing a laboratory culture of Anabaena cylindrica strain Bio33, H. Bartrow and M. Matthias (EMAU Greifswald, Germany), C. Kakoschke and A. Abrahamik (HZI Braunschweig, Germany) for their skillful technical assistance, and O. Morgenstern, Ph.D., and J. Technau for measuring IR spectra.

■

REFERENCES

(1) Aráoz, R.; Molgó, J.; de Marsac, N. T. Toxicon 2010, 56, 813− 828. (2) Van Apeldoorn, M. E.; Van Egmond, H. P.; Speijers, G. J. A.; Bakker, G. J. I. Mol. Nutr. Food Res. 2007, 51, 7−60. (3) Golubic, A.; Abed, R. M. M.; Palinska, K.; Pauillac, S.; Chinain, M.; Laurent, D. Toxicon 2010, 56, 836−841. (4) Pearson, L.; Mihali, T.; Moffitt, M.; Kellmann, R.; Neilan, B. Mar. Drugs 2010, 8, 1650−1680. (5) Burja, A. M.; Banaigs, B.; Abou-Mansour, E.; Burgess, J. G.; Wright, P. C. Tetrahedron 2001, 57, 9347−9377. (6) Rastogi, R. P.; Sinha, R. P. Biotechnol. Adv. 2009, 27, 521−539. (7) Tan, L. T. Phytochemistry 2007, 68, 954−979. (8) Tan, L. T. J. Appl. Phycol. 2010, 22, 659−676. (9) Nunnery, J. K.; Mevers, E.; Gerwick, W. H. Curr. Opinion Biotechnol. 2010, 21, 787−793. (10) Sharma, N.; Tiwari, S.; Tripathi, K.; Rai, A. J. Appl. Phycol. 2011, 23, 1059−1081. (11) Singh, R. K.; Tiwari, S. P.; Rai, A. K.; Mohapatra, T. M. J. Antibiot. 2011, 64, 401−412. (12) Mayer, A. M. S.; Rodríguez, A. D.; Berlinck, R. G. S.; Fusetani, N. Comp. Biochem. Physiol. 2011, 153, 191−222. (13) Gerwick, W. H.; Moore, B. S. Chem. Biol. 2012, 19, 85−98. (14) Kauffman, C. A.; Yoshikawa, T. T. CID 2001, 33, 550−555. (15) Carrillo-Muñoz, A. J.; Giusiano, G.; Ezkurra, P. A.; Quindós, G. Rev. Esp Quimioterap 2006, 19, 130−139. (16) Shalini, K.; Kumar, N.; Drabu, S.; Sharma, P. K. Beilstein J. Org. Chem. 2011, 7, 668−677. (17) Barrett, D. Biochim. Biophys. Acta 2002, 1587, 224−233. (18) Finni, T.; Kononen, K.; Olsonen, R.; Wallström, K. AMBIO 2001, 30, 172−178. (19) Lehtimaeki, J.; Lyra, C.; Suomalainen, S.; Sundman, P.; Rouhiainen, L.; Paulin, L.; Sivonen, K. Int. J. System. Evol. Microbiol. 2000, 50, 1043−1053. (20) Laamanen, M. J.; Gugger, M. F.; Lehtimäki, J. M.; Haukka, K.; Sivonen, K. Appl. Environ. Microbiol. 2001, 67, 4638−4647. (21) Stal, L. J.; Albertano, P.; Bergman, B.; Bröckel, K. v.; Gallon, J. R.; Hayes, P. K.; Walsby, A. E. Cont. Shelf Res. 2003, 23, 1695−1714. (22) Suikkanen, S.; Engström-Ö st, J.; Jokela, J.; Sivonen, K.; Viitasalo, M. J. Plankton Res. 2006, 28, 543−550. (23) Janson, S.; Granéli, E. Int. J. System. Evol. Microbiol. 2002, 52, 1397−1404. (24) Sivonen, K.; Halinen, K.; Sihvonen, L. M.; Koskenniemi, K.; Sinkko, K.; Rantasärkkä, K.; Moisander, P. H.; Lyra, C. AMBIO 2007, 36 (2), 180−185. (25) Halinen, K.; Jokela, J.; Fewer, D. P.; Wahlsten, M.; Sivonen, K. Appl. Environ. Microbiol. 2007, 73, 6543−6550. (26) Karlsson, K. M.; Kankaanpäa,̈ H.; Huttunen, M.; Meriluoto, J. Harmful Algae 2005, 4, 163−166. (27) Surakka, A.; Sihvonen, L. M.; Lehtimäki, J. M.; Wahlsten, M.; Vuorela, P.; Sivonen, K. Environ. Toxicol. 2005, 20, 285−292. (28) Engström-Ö st, J.; Repka, S.; Mikkonen, M. Harmful Algae 2011, 10, 530−535. (29) Neuhof, T.; Schmieder, P.; Preussel, K.; Dieckmann, R.; Pham, H.; Bartl, F.; Döhren, H. v. J. Nat. Prod. 2005, 68, 695−700. J


Balticidins A–D, Antifungal Hassallidin-Like Lipopeptides from the

Recommend Documents