Introduction

Gram-negative bacteria have lipopolysaccharide (LPS) in their cell walls. LPS is well known as a bacterial endotoxin and it is a key component for triggering sepsis or septic shock. These symptoms start when a large amount of LPS binds to cell surface receptors that then leads to the overexpression of inflammatory cytokines. CD14 is the common LPS receptor1 in such events.

We had previously established a convenient screening method for inhibitors of LPS binding to CD14, and reported about the novel inhibitors, pedopeptins.2, 3

In the course of continuous screening, we found other novel compounds, ogipeptins (formerly called B-5529s,4 Figure 1) from the cultured broth of Pseudoalteromonas sp. SANK 71903. Ogipeptins are cyclic lipopeptides with some basic amino acids and a fatty acid moiety, and in this way they are similar to pedopeptins and polymyxins.5 Here, we report the physicochemical properties and structural elucidations of ogipeptins. Details of the taxonomy of the producing organism, fermentation, isolation and biological activities of ogipeptins have been reported in a previous paper.6

Figure 1
figure 1

Structures of ogipeptins.

Abbreviations: Arg, arginine; β-OH Dab, β-hydroxydiaminobutylic acid; Dab, diaminobutylic acid; Dhb, dehydrobutyrine; Leu, leucine.

Results

Physicochemical properties of ogipeptins A (1), B (2), C (3) and D (4)

The physicochemical properties of ogipeptins are summarized in Table 1.

Table 1 Physicochemical properties of ogipeptins

Structural elucidation

The 1H NMR data of ogipeptins A (1), B (2), C (3) and D (4) closely resembled each other, except for some of the olefinic protons. The spectra seemed to suggest peptide moieties and fatty acid moieties. The amino acid analysis of the acid hydrolysate of each compound showed the same amino acids, such as one mole of leucine (Leu), one mole of arginine (Arg) and undentified amino acids. GC-MS analysis of the trimethylsilyl derivatives of each acid hydrolysate not only supported the amino acid analysis result, but also proved that one of the unidentified amino acids was diaminobutylic acid (Dab). According to the IR spectral data, they had no ester group, namely, they were not depsipeptides. These data seemed to indicate that they had the same peptide moiety, and the difference among them was a fatty acid moiety.

Structural elucidation of 2

The structural elucidation was mainly focused on 2 because 2 showed the best peak shapes in the NMR spectra among the four homologs. The molecular formula of 2 was determined as C44H80N14O11. The 1H and 13C NMR spectral data of 2 were obtained in 0.04N DCl/D2O and are summarized in Table 2. In the 1H NMR spectrum, signals from some methyl groups, α-methines of the amino acid residues (δ 4.25–4.48) and three olefinic protons (δ 5.42, 5.52 and 6.60) were observed. In the 13C NMR spectrum, all 44 carbons were observed and they were classified by DEPT and HSQC (heteronuclear single quantum coherence) spectra into 4 × methyl, 17 × methylene, 13 × methine and 10 × quarternary carbons.

Table 2 1H and 13C NMR signal assignments of ogipeptin B (2) in 0.04N DCl/D2O

As shown in Figure 2, by several NMR techniques, such as DQFCOSY (double quantum filtered COSY), HSQC, HMBC (heteronuclear multiple bond correlation) and band-selective HMBC,7, the partial structures, except for the three parts between N-1 of 2-amino-2-butenoic acid (dehydrobutyrine (Dhb)) and C-1 of β-hydroxydiaminobutylic acid (β-OH Dab)-3, between N-1 of arginine and C-1 of leucine, and between N-1 of β-OH Dab-1 and C-1 of arginine, were elucidated. The structure of the fatty acid moiety was also elucidated by these NMR data and found to be n-5-dodecenoic acid.

Figure 2
figure 2

Partial structures obtained from the NMR analyses of 2.

To determine the bond between C-1 of β-OH Dab-3 and N-1 of Dhb, 2 was hydrogenated by palladium black and the derivative (5), in which Dhb was hydrogenated into 2-aminobutanoic acid, was obtained. In an HMBC spectrum of 5, the C-H long-range correlation between H-2 of 2-aminobutanoic acid and C-1 of β-OH Dab-3 was observed. This indicated that the N-1 of Dhb was connected with C-1 of β-OH Dab-3.

In order to investigate uncertain bonds, partial acid hydrolysis of 2 with 3N HCl was performed that obtained the deacyl derivative (ogipeptin X, 6) shown in Figure 3. Further elucidation was carried out by using the compound 6.

Figure 3
figure 3

Band-selective HMBC correlations of 6 focused on the carbonyl region.

The molecular formula of 6 was determined to be C32H60N14O10 (m/z, found 799.4518 (M-H)-, calcd. 799.4539 for C32H59N14O10), consistent with the one for the deacylated 2. The 1H and 13C NMR spectral data of 6 obtained in D2O are summarized in Table 3. By several NMR techniques, such as DQFCOSY, HSQC, HMBC and band-selective HMBC, the structure was elucidated. Among them, band-selective HMBC gave useful information about inter- or intra-residual linkages of each constitutive amino acid, as shown in Figure 3. Thus, the planar structure of 2 was determined.

Table 3 1H and 13C NMR signal assignments of ogipeptin X (6) in D2O

The stereochemistry of 2 was elucidated as follows: The configuration of the olefin in the fatty acid moiety was determined by 1H homo-decoupling experiments. Each of the H-4 and H-7 allylic position-irradiated spectra gave the doublet signals at δ 5.42 (1H, d, 11.0 Hz) corresponding to H-5 and δ 5.52 (1H, d, 11.0 Hz) to H-6. These data showed that the double bond of the acyl chain had a cis configuration. Namely, the fatty acid of 2 was cis-5-dodecenoic acid.

The configuration of Dhb in 2 was determined by double-pulsed field gradient spin-echo (DPFGSE) NOE8, 9 experiments in D2O/H2O (1:9) containing 0.04N DCl for the purpose of observing NOEs between H-3/H-4 of Dhb and the amide proton in Dhb or the adjacent residue. At first, each NH signal observed in a WATERGATE spectrum was assigned by 1D-selective DPFGSE TOCSY spectra.10 According to this, NH signals of Dhb and Dab were assigned to δ 9.53 and 8.45, respectively. Selective excitation at H-3 or H-4 in DPFGSE NOE experiments exhibited NOEs between H-4 of Dhb and N-1 of Dhb and between H-3 of Dhb and N-2 of Dab. Thus, the configuration of Dhb was determined to be Z as shown in Figure 4.

Figure 4
figure 4

Configuration of dehydrobutyrine (Dhb).

Stereochemistry of each amino acid residue in 2 was determined by the advanced Marfey’s method.11 Acid hydrolysate of 2 was D- or L-FDLA (1-fluoro-2,4-dinitrophenyl-5-leucine amide) derived and analyzed by LC-MS. By comparing each authentic FDLA-amino acid standard, all the chiralities of arginine, Dab and leucine were determined to be L-configuration.

Structural elucidation of other homologs

The 1H and 13C NMR spectral data of 1, 3 and 4 were also obtained in 0.04N DCl/D2O. Their spectra were similar to 2 and their partial acid hydrolyses with 3N HCl obtained the same cyclic peptide (6). However, the olefinic protons located in the fatty acid moiety were observed in 4, but not in 1 and 3. This proved that ogipeptins had the same cyclic peptide skeleton and the difference among them was localized in a fatty acid moiety.

Fatty acid identification of 1. As described above, the 1H NMR data of 1 showed that 1 has a saturated acyl chain. Identification of the constitutive fatty acid of 1 was performed by GC-MS analysis using the methanolysis product of 1 with 5% HCl-MeOH. By comparing the data with that of the authentic standard, the fatty acid was found to be n-decanoic acid. Thus, the planar structure of 1 was elucidated, as shown in Figure 1.

Fatty acid identification of 3. The 1H NMR data of 3 showed that 3 has a saturated acyl chain similar to 1. Based on the molecular formula of 3, the chain length of 3 was considered to be two-methylene units longer than that of 1. Finally, 3 was analyzed by the same method as 1 and was found to be n-dodecanoic acid.

Fatty acid identification of 4. The 1H NMR data of 4 showed that 4 had a straight-acyl chain with an unsaturated bond. The chain length was considered to be two-methylene units longer than that of 2. The location of the double bond was elucidated by a linked-scan spectrum.12 To obtain the fatty acid, 4 was hydrolyzed with 3N HCl at 37 °C for 24 h and the reactant was extracted with ethyl acetate. Observed product ions at m/z 99, 153 and 154 in a linked-scan spectrum (negative-ion FAB-MS) revealed that the fatty acid was n-7-tetradecenoic acid, as shown in Figure 5.

Figure 5
figure 5

Linked-scan spectrum of the fatty acid of 4. A full color version of this figure is available at The Journal of Antibiotics journal online.

The configuration of the double bond was determined by the same 1H homo-decoupling experiments as 2. In the spectrum, decoupling of H-6 and H-9 allylic positions of the chain gave the same coupling constants (10.9 Hz) to the olefinic protons at δ 5.43 and δ 5.46, respectively. According to the coupling constant, the fatty acid of 4 was finally determined as n-cis-7-tetradecenoic acid.

Discussion

In the course of our screening for inhibitors of LPS binding to CD14, ogipeptins A (1), B (2), C (3) and D (4) were isolated from the cultured broth of Pseudoalteromonas sp. SANK 71903.

The structures of these compounds were elucidated by physicochemical data and spectral analyses, and they were determined to be new cyclic lipopeptides. The NMR analyses of ogipeptins were carried out with 0.04N DCl/D2O because the ogipeptins were soluble in acidic water, and 0.04N DCl/D2O provided the best peak shapes from among the other solvents we tested. In the case of the analyses of Dhb configuration, we used the mixed solvent of heavy and light water (heavy/light (1:9) containing 0.04N DCl) to observe amide protons.

In order to clarify the configuration of Dhb, we used the 1D-selective DPFGSE TOCSY and NOESY techniques that, with a good sensitivity, could clearly detect only objective signals. DPFGSE TOCSY in particular was a key method.

We considered that observation of the NOEs between H-3/H-4 of Dhb and the amide proton in Dhb or the adjacent residues were conclusive determinations. However, we were not able to assign the amide proton signals by using the long-range correlations from α-methine protons because these α-methine protons were overlapped with strong water signals. Therefore, we chose 1D-selective DPFGSE TOCSY to assign the intraresidual signals, including the water hidden signals, by exciting the proton of the β-position. By this method, we could assign the amide proton signals of Dab and that of Dhb that has no α-methine proton.

As results, we succeeded in determining the configuration of Dhb as Z by the observed NOEs to these amide protons. The result of the configuration was also supported by the chemical shift value of H-3 (δ 6.60) because H-3 signals of (Z)-Dhb appeared at a lower field (approx. δ 6.5) than that of (E)-Dhb (between δ 5.6 and 5.9) in the cases of the other Dhb-containing compounds.3, 13, 14, 15

In our assay system, pedopeptins, which we had previously obtained, or polymyxins also showed potent activities. Their structures were the similar cyclic lipopeptides with basic amino acids, the same as ogipeptins. Ogipeptins are composed of seven amino acid residues, and five of them are basic amino acids. It is characteristic that three of them are β-OH Dab. Although polymyxins also possess many basic amino acids, all of them are Dab, not β-OH Dab. Among known antibiotics, the compounds possessing β-OH Dab are rare and there are no compounds that are rich in β-OH Dab like ogipeptins and their related compounds.16

Experimental procedure

General experimental procedures

IR spectra were obtained on a JASCO FT/IR-610 spectrometer (JASCO, Tokyo, Japan). UV spectra were recorded on an UV-265FW spectrometer (Shimadzu, Kyoto, Japan). NMR spectra were recorded on AVC500 spectrometers equipped with C-H dual and TXI cryogenic probes (Bruker, Billerica, MA, USA). High-resolution mass spectra were recorded on a Micromass LCT spectrometer (Waters, Milford, MA, USA). GC-MS spectra were recorded on a HP6890 Series GC System equipped with a column (DB-5, 0.25 mm i.d. × 15 m, Agilent Technologies, Santa Clara, CA, USA) and a 5973 Mass Selective Detector. Linked-scan spectrum was obtained with a JMS-700 spectrometer (JEOL, Tokyo, Japan).

Amino acid analysis

Compound 2 (100 μg) was completely hydrolyzed with 6N HCl (400 μl) at 105 °C for 20 h. The hydrolysates were concentrated to dryness and redissolved in 0.02N HCl and applied to a HITACHI L-8800 Amino Acid Analyzer (Hitachi, Tokyo, Japan).

Analysis of amino acid by GC-MS

Compound 2 (1 mg) was completely hydrolyzed by the same method as described above. To the solution of the hydrolysate in pyridine (100 μl) was added bis-(trimethylsilyl)trifluoroacetamide (BS-TFA, Sigma-Aldrich, St Louis, MO, USA, 100 μl). The mixture was diluted with pyridine and applied to a GC-MS. Each authentic amino acid was also prepared and analyzed in the same way.

Reduction of 2 with palladium black

Compound 2 (20 mg) was dissolved in 0.1% aq. formic acid/dioxane (1:1) solution (8 ml). To the solution was added palladium black (6.0 mg), and it was then stirred at room temperature for 24 h under a hydrogen atmosphere. The reductant was purified by preparative HPLC using an ODS column (Senshu Pak Pegasil ODS, 20 i.d. × 150 mm, Senshu Scientific, Tokyo, Japan). The chromatography was performed with MeCN/5 mM aq. Na2SO4 in 0.5% phosphoric acid (Solution A) (30:70) at a flow rate of 8 ml min−1. For desalting, the solution was subjected to an HP-20 column (5 ml). After washing the column with water (15 ml), 5 was eluted with 60% aq. acetone-0.1% formic acid (25 ml). The eluate was concentrated to remove acetone and was lyophilized to afford 5 as a white powder (2.9 mg).

5: ESI-MS (m/z): 985 (M+H)+, 1007 (M+Na)+. 1H NMR (500 MHz, D2O, HDO as 4.80 ppm): δ 0.79 (3H, t, 7.0 Hz), 0.85 (3H, d, 5.5 Hz), 0.89 (3H, t, 7.5 Hz), 0.91 (3H, d, 5.5 Hz), 1.18–1.30 (16H, overlapped), 1.51–1.65 (6H, overlapped), 1.65–1.83 (4H, overlapped), 1.90 (1H, m), 2.00 (1H, m), 2.23 (1H, m), 2.34 (2H, m), 2.96–3.07 (5H, overlapped), 3.12–3.26 (4H, overlapped), 3.49 (1H, dd, 4.9, 13.9 Hz), 4.12–4.17 (2H, overlapped), 4.19–4.26 (2H, overlapped), 4.26–4.32 (4H, overlapped), 4.33 (1H, br. s), 4.40 (1H, dd, 4.7, 10.2 Hz). 13C NMR (125 MHz, D2O, formic acid as 169.5 ppm): δ 12.5, 16.2, 23.2, 24.9, 25.1, 26.9, 27.3, 27.5, 28.1, 30.6, 31.1–31.5 (7 carbons), 34.0, 38.3, 39.3, 42.0, 43.3, 44.5, 44.8, 45.2, 54.4, 55.6, 56.3, 58.9, 59.2, 59.5, 60.0, 69.5, 69.7, 71.6, 159.5, 172.7, 173.5, 175.4, 175.6, 176.6, 177.3, 177.6, 180.9.

Partial hydrolyzation of 2

Compound 2 (30 mg) was dissolved in 3N HCl (10 ml) and hydrolyzed at 37 °C for 24 h. The partial hydrolysate was applied to a Sep-Pak C18 cartridge (Waters). The elution was performed stepwise by using MeCN/Solution A in the ratios of 1:99, 1:19, 1:9 and 1:4 (each 5 ml). The ratio 1:9 eluate containing 6 was concentrated to remove MeCN. After desalting as described above, it was lyophilized to afford 6 as a white powder (7.0 mg).

Configuration of the constitutive amino acid

The configurations of the amino acids were determined by the advanced Marfey’s method.13 Each of the complete acid hydrolysate of 2 (1.0 mg) and authentic samples (0.2 mg, each) were dissolved in water (10 μl). To the solutions, 1 M NaHCO3 (5 μl) and 1% D- or L-FDLA/acetone solution (20 μl) were added and incubated at 37 °C for 1.5 h. After the incubation, the reactions were terminated by the addition of 1N HCl (5 μl). The reactants diluted with acetone were analyzed by LC-MS (LCQ spectrometer, Thermo Fisher Scientific, Waltham, MA, USA) with an ODS column (SYMMETRY C18, 4.6 i.d. × 150 mm, Waters).

Ogipeptin A (1)

1H NMR (500 MHz, 0.04N DCl/D2O, HDO as 4.80 ppm): δ 0.78 (3H, br. t), 0.82 (3H, d, 4.6 Hz), 0.87 (3H, d, 4.6 Hz), 1.14–1.31 (12H, overlapped), 1.46–1.62 (6H, overlapped), 1.70 (3H, d, 6.6 Hz), 1.66–1.82 (2H, overlapped), 1.89 (1H, m), 2.09 (1H, m), 2.20 (1H, m), 2.34 (2H, m), 2.94–3.17 (7H, overlapped), 3.17–3.26 (2H, overlapped), 3.53 (1H, br. dd), 4.13–4.29 (6H, overlapped), 4.32–4.40 (2H, overlapped), 4.42 (1H, br.), 6.56 (1H, q, 6.6 Hz). 13C NMR (125 MHz, 0.04N DCl/D2O, dioxane as 67.2 ppm): δ 13.3, 14.1, 21.0, 22.7, 22.9, 25.1, 25.2, 25.9, 28.2, 29.0–29.2 (5 carbons), 31.8, 36.1, 37.1, 39.5, 41.0, 42.3, 42.6, 43.1, 52.9, 53.6, 53.9, 56.8, 57.0, 58.1, 67.4, 67.5, 69.6, 128.8, 133.5, 157.3, 167.9, 170.7, 171.3, 172.8, 173.4, 174.6, 175.0, 179.0.

Ogipeptin C (3)

1H NMR (500 MHz, 0.04N DCl/D2O, HDO as 4.80 ppm) δ 0.84 (3H, t, 7.0 Hz), 0.88 (3H, d, 6.2 Hz), 0.93 (3H, d, 6.2 Hz), 1.20–1.35 (16H, overlapped), 1.50–1.70 (6H, overlapped), 1.76 (3H, d, 7.2 Hz), 1.72–1.88 (2H, overlapped), 1.98 (1H, m), 2.16 (1H, m), 2.27 (1H, m), 2.40 (2H, m), 3.01–3.16 (5H, overlapped), 3.20 (2H, m), 3.25–3.33 (2H, overlapped), 3.58 (1H, dd, 0.6, 13.8 Hz), 4.19–4.34 (5H, overlapped), 4.36 (1H, dd, 3.7, 10.8 Hz), 4.39–4.45 (2H, overlapped), 4.48 (1H, br. s), 6.61 (1H, q, 7.2 Hz). 13C NMR (125 MHz, 0.04N DCl/D2O, dioxane as 67.2 ppm): δ 13.2, 14.1, 21.0, 22.7, 22.9, 25.1 (2 carbons), 25.9, 28.4, 29.0–29.6 (7 carbons), 31.9, 36.0, 37.1, 39.5, 40.9, 42.4, 42.7, 43.1, 52.9, 53.7 (2 carbons), 56.9, 57.0, 58.3, 67.4, 67.5, 69.7, 128.9, 133.3, 157.3, 167.8, 170.7, 171.4, 172.7, 173.4, 174.6, 174.9, 178.7.

Ogipeptin D (4)

1H NMR (500 MHz, 0.04N DCl/D2O, HDO as 4.80 ppm) δ 0.84 (3H, t, 6.6 Hz), 0.88 (3H, d, 6.0 Hz), 0.93 (3H, d, 6.0 Hz), 1.22–1.40 (12H, overlapped), 1.56 (1H, overlapped), 1.60 (2H, overlapped), 1.61 (2H, overlapped), 1.63 (1H, overlapped), 1.75 (1H, overlapped), 1.76 (3H, d, 7.2 Hz), 1.79 (1H, overlapped), 1.94 (1H, m), 1.99-2.08 (4H, overlapped), 2.14 (1H, m), 2.26 (1H, m), 2.40 (2H, m), 3.03 (1H, overlapped), 3.10 (1H, overlapped), 3.10 (2H, overlapped), 3.12 (1H, overlapped), 3.19 (2H, m), 3.27 (1H, overlapped), 3.27 (1H, overlapped), 3.59 (1H, dd, 4.4, 14.3 Hz), 4.20 (1H, br. m), 4.23-4.30 (3H, overlapped), 4.31 (1H, overlapped), 4.33 (1H, overlapped), 4.40 (1H, overlapped), 4.43 (1H, overlapped), 4.47 (1H, br. s), 5.43 (1H, overlapped), 5.46 (1H, overlapped), 6.59 (1H, q, 7.2 Hz). 13C NMR (125 MHz, 0.04N DCl/D2O, dioxane as 67.2 p.p.m.) δ 13.2, 14.1, 21.0, 22.7, 22.9, 25.1 (2 carbons), 25.7, 27.0, 27.2, 28.4, 28.6, 28.8, 29.1, 29.2, 29.6, 31.6, 36.0, 37.1, 39.5, 40.9, 42.3, 42.7, 43.2, 52.9, 53.6, 53.7, 56.89, 56.93, 58.2, 67.4, 67.5, 69.7, 129.0, 131.0, 131.8, 133.2, 157.3, 168.0, 170.6, 171.4, 172.7, 173.4, 174.6, 175.0, 178.7.