Antifungal cyclic lipopeptides from Bacillus amyloliquefaciens strain BO5A

Article abstract of DOI:10.1021/np400119n

A bioassay-guided fractionation of Bacillus amyloliquefaciens strain BO5A afforded the isolation of two new cyclic lipopeptides (1 and 2) as the major lipid constituents (>60%) of the CHCl₃-MeOH (2:1) extract. The chemical structures of the isolated metabolites were elucidated by spectroscopic methods, including 1D and 2D NMR spectroscopy, mass spectrometry (MS), secondary ion mass spectrometry (MS1, MS2), and chemical degradation. The compounds are members of the surfactins family and are based on a heptapeptide chain composed by Glu-Val-Leu-Val-Asp-Leu-Leu. Its N-terminal end is N-acylated by an (R)-3-hydroxy fatty acid with linear alkyl chains of 16:0 and 15:0 (1 and 2, respectively). The 3-hydroxyl group closes a 25-membered lactone ring with the carboxylic group of the C-terminal amino acid. The isolated compounds were tested for their inhibitory activity against the four pathogenic fungi Fusarium oxysporum, Aspergillus niger, Botrytis cinerea, and Penicillium italicum and the biocontrol fungus Trichoderma harzianum. Compound 2 displayed activity against all tested pathogens.

Full text of DOI:10.1021/np400119n

Article
pubs.acs.org/jnp
Antifungal Cyclic Lipopeptides from Bacillus amyloliquefaciens
Strain BO5A
Adriana Romano,† Domenico Vitullo,‡ Mauro Senatore,† Giuseppe Lima,‡ and Virginia Lanzotti*,†
†Dipartimento di Agraria, Universita
̀
di Napoli Federico II, Via Universita
̀
100, 80055 Portici (NA), Italy
‡Dipartimento Agricoltura, Ambiente e Alimenti, Universita
̀
del Molise, Via F. De Sanctis snc, 86100 Campobasso, Italy
S
* Supporting Information
ABSTRACT: A bioassay-guided fractionation of Bacillus
amyloliquefaciens strain BO5A afforded the isolation of two
new cyclic lipopeptides (1 and 2) as the major lipid con-
stituents (>60%) of the CHCl3−MeOH (2:1) extract. The
chemical structures of the isolated metabolites were elucidated
by spectroscopic methods, including 1D and 2D NMR
spectroscopy, mass spectrometry (MS), secondary ion mass
spectrometry (MS1, MS2), and chemical degradation. The
compounds are members of the surfactins family and are based
on a heptapeptide chain composed by Glu-Val-Leu-Val-Asp-
Leu-Leu. Its N-terminal end is N-acylated by an (R)-3-hydroxy
fatty acid with linear alkyl chains of 16:0 and 15:0 (1 and 2, respectively). The 3-hydroxyl group closes a 25-membered lactone
ring with the carboxylic group of the C-terminal amino acid. The isolated compounds were tested for their inhibitory activity
against the four pathogenic fungi Fusarium oxysporum, Aspergillus niger, Botrytis cinerea, and Penicillium italicum and the biocontrol
fungus Trichoderma harzianum. Compound 2 displayed activity against all tested pathogens.
against the four pathogenic fungi F. oxysporum, Aspergillus niger,
Botrytis cinerea, and Penicillium italicum and the biocontrol
fungus Trichoderma harzianum.
yclic lipopeptides are a class of bacterial secondary meta-
Cbolites based on a polypeptide chain, generally composed
of seven amino acids and a 3-hydroxy or 3-amino fatty acid,1
connected in two places, giving rise to a macrocyclic structure.
The N-terminal end of the polypeptide is involved in one
amide bond with the carboxylic group of the fatty acid, while
the 3-hydroxy or 3-amino group of the fatty acid gives rise
to an ester or amide bond, respectively, with the C-terminal
amino acid.2
Cyclic lipopeptides have received increasing attention as key
compounds in biocontrol of plant disease3,4 due to their strong
antimicrobial activity, low toxicity, and low stability, i.e., easily
degradable under natural conditions, compared to other chemical
pesticides. In this context, many members of the Bacillus genus
have been studied chemically because they produce lipopeptides
that inhibit plant pathogens as well as having a key role in
stimulating the host defense mechanism.5−9 Known lipopeptides
from Bacillus spp. include basic structuresnamed iturins,
surfactins, and fengycinsdiffering in the amino acid
composition and sequence and in the structure of the alkyl chain.
In the course of characterizing bioactive compounds from
Bacillus spp., we recently isolated three cyclic lipopeptides
from Bacillus amyloliquefaciens strain BO7 that showed strong
and dose-dependent activity against Fusarium oxysporum.10,11
We have now extended our analysis to strain BO5A of
B. amyloliquefaciens, isolating and characterizing two new
lipopeptides (1 and 2), whose stereostructures have been
elucidated by NMR, MS techniques, and chemical methods.
The isolated compounds were subjected to antifungal tests
RESULTS AND DISCUSSION
■
The cell-free culture filtrate of B. amyloliquefaciens strain BO5A
was found to contain approximately 0.0015% w/v of CHCl3−
MeOH (2:1) extractable lipopeptides. Purification of these
compounds was carried out by reversed-phase HPLC and
isocratic elution, affording the new compounds 1 (8.4 mg/L)
and 2 (5.8 mg/L). The purity of the isolated compounds,
1
eluted as single peaks, was determined by H NMR spectra
(Figures S1 and S7).
Compound 1, isolated as an amorphous solid as the major
component, showed a molecular formula of C53H93N7O13,
deduced by high-resolution FAB-MS measurements and
consistent with 13C NMR data (Figure 1 and Table 1). Low-
resolution positive ion ESI-MS spectra (Figure S2) gave the
pseudomolecular ion peak at m/z 1037 [M + H]+ and, in
addition, showed correlated peaks of Na and K adducts at m/z
1059 [M + Na]+ and 1075 [M + K]+, respectively. Preliminary
1H NMR analysis of compound 1 indicated its lipopeptide
nature (Figure 1 and Table 1), thus showing aminic protons
(δ 8.60−7.50), α-amino acid protons (δ 5.30−4.00), and a long
alkyl chain (δ 0.80−1.40). In addition, the 13C NMR spectrum
contained diagnostic signals of carboxyl groups (δ 175.5−172.1)
Received: February 22, 2013
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
After permethylation with ethereal diazomethane and further
purification (see Experimental Section), the lipophilic fraction
was shown to contain 3-methoxy hexadecanoate, as identified
by GC-MS. The aqueous phase was analyzed by TLC
(CH3CN−EtOAc−CH3CO2H−H2O, 20:6:4:3) and GC-MS
(as N-trifluoroacetyl derivatives of the corresponding butyl
esters), allowing the identification of valine, leucine, and
glutamic and aspartic acids and the estimation of their ratios
(2:3:1:1, respectively), according to peak areas. To confirm the
stereochemistry of single amino acids, an aliquot of the aqueous
phase from the hydrolysate of 1 was subjected to enzymatic
oxidation by L- and D-amino acid oxidases,12 showing reaction
with the L-enzyme and remaining unchanged with the D-enzyme.
This procedure allowed us to unequivocally determine the
L-series for all amino acid residues.
Further information of the chemical structure of 1 was
obtained by tandem ESI-MS analysis of the Na-ionized
molecular ion (Figure S3). Taking the MS1 peak at m/z
1059 [M + Na]+ as precursor, obtained by simple cleavage
(Figure 2A), the MS2 fragments at m/z 946 (−113) and 833
(−113) obtained by ESI-MS1 and -MS2 (Figure 2B) showed
the loss of amino acid residues after the cleavage of the ester
bond (Figure 2A), thus indicating the existence of a cyclic
structure.13 The fragment at 113 amu corresponds to either
leucine or iso-leucine, but according to the NMR data and
the chemical degradation study must be leucine. In particular,
these peaks were indicative of the consecutive losses of two Leu
Figure 1. Structure of compounds 1 and 2 purified from Bacillus
subtilis strain BO5A.
that, together with seven nitrogen-bearing carbon signals
(δ 50.5−59.8), indicated the presence of seven amide bonds
(Figure 1 and Table 1).
An aliquot of 1 (1.5 mg) was treated with 0.30 mL of 6 M
HCl for 40 h at 110 °C. The reaction mixture was diluted with
water, and the lipophilic products were extracted with CHCl3.
1
Table 1. H and 13C NMR Data of 1 and 2 at 500 and 100 MHz, Respectively, in CD3OD
1
2
1
2
position
δC mult. δH mult. (J in Hz) δC mult. δH mult. (J in Hz)
position
δC mult. δH mult. (J in Hz) δC mult. δH mult. (J in Hz)
Glu-1
NH
CO
α
Asp-5
7.57 d (2.5)
7.57 d (2.5)
α
β
53.9 d
35.2 t
4.18 t (4.5)
53.9 d
35.5 t
4.18 t (4.5)
172.4 s
59.7 d
26.9 t
172.4 s
59.7 d
26.9 t
2.98, 2.85 dd
(15.0, 4.5)
2.98, 2.85 dd
(15.0, 4.5)
4.04 bt (5.5)
2.11 m
4.03 bt (5.5)
2.11 m
γ
173.2 s
173.3 s
β
Leu-6
γ
30.1 t
2.22 t
30.1 t
2.21 t
N−H
8. 43 d (2.5)
8.43 d (2.5)
δ
173.3 s
173.3 s
CO
172.0 s
53.0 d
41.1 t
24.4 t
21.6 q
172.0 s
53.0 d
41.4 t
24.3 t
21.5 q
Val-2
NH
CO
α
α
4.38 bt (6.5)
4.39 bt (6.5)
8.20 d (2.5)
8.20 d (2.5)
β
1.70 dd (6.5, 7.5)
1.59 dq (5.5, 6.0)
0.91 d (6.0)
1.70 dd (6.5, 7.5)
1.59 dq (5.5, 6.0)
0.90 d (6.0)
172.7 s
52.2 d
24.5 t
172.7 s
52.2 d
24.4 t
γ
4.42 bd (6.0)
1.47 dq (6.0)
0.93 d (6.0)
4.43 bd (6.0)
1.46 dq (6.0)
0.93 d (6.0)
δ
β
Leu-7
γ
21.5 q
21.6 q
N−H
8. 51 d (2.5)
8. 50 d (2.5)
Leu-3
NH
CO
α
CO
172.1 s
52.0 d
41.0 t
24.8 t
21.8 q
172.1 s
52.0 d
41.1 t
24.9 t
21.8 q
8. 40 d (2.5)
8. 40 d (2.5)
α
4.24 bt (6.5)
4.25 bt (6.5)
172.6 s
52.1 d
41.4 t
24.4 t
21.3 q
172.6 s
52.2 d
41.4 t
24.5 t
21.6 q
β
1.71 dd (6.5, 7.5)
1.59 dq (5.5, 6.0)
0.92 d (6.0)
1.72 dd (6.5, 7.5)
1.59 dq (5.5, 6.0)
0.92 d (6.0)
4.43 bt (6.5)
4.44 bt (6.5)
γ
β
1.78 dd (6.5, 7.5)
1.49 dq (5.5, 6.0)
0.92 d (6.0)
1.78 dd (6.5, 7.5)
1.49 dq (5.5, 6.0)
0.92 d (6.0)
δ
γ
β-OH acid
CO2H
α
δ
175.0 s
39.5 t
175.0 s
39.5 t
Val-4
NH
CO
α
2.52, 2.46 dd
(15.0, 5.0)
2.50, 2.47 dd
(15.0, 5.0)
7.91 d (2.5)
7.91 d (2.5)
175.5 s
50.5 d
24.4 t
175.5 s
50.5 d
24.4 t
β
72.2 d
33.5 t
28.9 t
29.6 t
31.9 t
22.7 t
16.0 q
5.30 m
72.2 d
33.4 t
28.9 t
29.6 t
31.9 t
22.7 t
16.0 q
5.30 m
4.46 bd (6.0)
1.49 dq (6.0)
0.90 (d (4.0)
4.47 bd (6.0)
1.49 dq (6.0)
0.89 d (4.0)
γ
2.50, 1.75 m
2.63, 2.01 m
1.30 m
2.50, 1.74 m
2.64, 2.00 m
1.30 m
β
δ
γ
17.9 q
17.8 q
ε−ω4
ω2
ω2
ω
a
a
Asp-5
N−H
CO
1.34 m
1.33 m
a
a
7.93 d (2.5)
7.93 d (2.5)
1.34 m
1.33 m
172.3 s
172.3 s
0.92 t
0.92 t
a
Overlapped with other signals.
B
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. (A) Simple cleavage of 1 and 2 in ESI-MS. (B) Set of fragmentation ions obtained in tandem ESI-MS of compound 1.
residues from the C-terminal end, while further peaks at m/z
718 (−115), 619 (−99), and 506 (−113) indicated the
consecutive losses of Asp, Val, and Leu, thus showing the
peptide chain sequence (Figure 2B). Figure 3A shows a double
hydrogen transfer mechanism in the ESI-MS1, -MS2 of cyclic
lipopeptides.14 Starting from the molecular ion peak at m/z
1059 (Figure 3B), the difference from the ion peak at m/z 815
(−244) was attributed to the loss of the dipeptide residue Leu-
Leu-H2O. This confirmed that a Leu residue at the C-terminal
end is involved in the ester bond with the 3-hydroxyl group,
affording the typical 25-membered lactone ring of the cyclo-
peptide skeleton. Figure 3B shows the set of fragment ions
produced from compound 1 after cleaving the cyclic structure
by a double hydrogen transfer mechanism that confirmed the
amino acid sequence found after a simple cleavage mechanism
(Figure 2A). Thus, the remaining glutamic acid residue had to
be located at the N-terminal end.
FAB-MS measurements and consistent with the 13C NMR data
(Table 1). Low-resolution ESI-MS spectra (Figure S8) showed
the pseudomolecular ion peak at m/z 1023 [M + H]+, together
with the peaks for Na and K adducts, at m/z 1045 [M + Na]+
and 1061 [M + K]+, respectively. Preliminary NMR analysis
showed a close similarity between compounds 1 and 2,
indicating the lipopeptide nature of compound 2. The latter
compound had a molecular weight 14 amu lower, thus
indicating the absence of a methylene group in compound 2
that could be located either on the alkyl chain or on the peptide
chain. A detailed investigation of compound 2 was undertaken
by obtaining 1D and 2D NMR spectra and by ESI-MS1, -MS2
experiments. Both NMR and tandem MS data indicated that 1
and 2 had the same amino acid sequence.
Hydrolysis of compound 2, followed by GC-MS analysis of
the water phase of the hydrolysate, showed the same amino
acid composition for both compounds 1 and 2, suggesting
the presence of an alkyl chain one carbon shorter in 2. This
was confirmed by GC-MS analysis of the CHCl3 phase of
the hydrolysate after permethylation, which identified the
lipophilic part of the compound as 3-hydroxypentadecanoic
acid.
Thus, the chemical structure of compound 2 is based on
the same heptapeptide chain, Glu-Val-Leu-Val-Asp-Leu-Leu,
N-acylated to the N-terminal amino acid, Glu, by a 3-hydroxy
C15 fatty acid with a saturated linear chain (15:0) (Figure 1).
The configuration at C-3 of the fatty acids in 1 and 2 was
determined using the derived 3-hydroxy fatty acids 3 and 4
obtained by acid hydrolysis, after derivatization with chiral
(α-methoxy-α-trifluoromethyl)phenylacetic acid (MTPA,
Mosher’s reagent)15,16 with S- and R-configuration (Scheme 1),
affording the (S)- and (R)-MTPA esters (3a/3b, 4a/4b),
respectively. Analysis of the Δδ(S−R) values according to the
Mosher model (Scheme 1) for each pair of diastereomeric MTPA
esters (3a/3b, 4a/4b) was undertaken. The S-configuration at
C-3 of the compounds should give positive Δδ(S−R) values
With these data in hand, we performed a detailed NMR
analysis, running 1D and 2D COSY (Figure S4) and HOHAHA
spectra, that allowed us to sequence all the proton multiplets in
eight spin systems belonging to the seven amino acid residues
and to the 3-hydroxy fatty acid residue. Association of all proton
signals with those of the directly linked carbons was obtained by
analysis of the 2D HSQC spectrum (Figure S5). Finally, study
2,3
of the
J
correlation peaks from the 2D HMBC spectrum
H−C
(Figure S6) allowed the spin systems to be connected, pointing
to a lipopeptide structure composed of seven amino acids
with the following sequence starting from the N-terminal end:
Glu-Val-Leu-Val-Asp-Leu-Leu. The N- and C-ends are attached
to the acidic and alcoholic functional groups, respectively,
of a 16-carbon 3-hydroxy acid (3-hydroxyhexadecanoic acid)
(1, Figure 1). 1H and 13C NMR assignments of compound 1 are
reported in Table 1. These data support the chemical structure
of compound 1 as depicted in Figure 1.
Compound 2, isolated as an amorphous solid, showed a
molecular formula of C52H91N7O13, deduced by high-resolution
C
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 3. (A) Double hydrogen transfer cleavage of 1 and 2 in tandem ESI-MS. (B) Molecular and fragmentation ions obtained in tandem ESI-MS
of compound 1. These ion peaks are 18 amu higher than corresponding peaks after simple cleavage. (C) Set of fragmentation ions obtained in
tandem ESI-MS of compound 1.
Scheme 1. (S)- and (R)-MTPA Esters (3a/3b, 4a/4b) of
Previous studies revealed a broad antimicrobial activity
spectrum of lipopeptides from Bacillus species such as B. subtilis7
and B. thuringiensis.17 Antifungal strains of B. amyloliquefaciens
were shown to produce iturins.18,19 Only in a one case have
surfactins been isolated from strain BO7 of this species, as
reported from our previous work,10 in which we isolated three
surfactin derivatives that showed significant antifungal activity
against Fusarium oxysporum f.sp. lycopersici (Fol). Thus, this is
the second report of surfactin compounds from B. amylolique-
faciens. Following the same methods,10 we tested the inhibitory
activity of the surfactins 1 and 2 against the same fungus and
against three of the most important fungal plant pathogens,
namely, the fruit rot agent Penicillium italicum, the air-borne
pathogens Aspergillus niger and Botrytis cinerea, as well as the
biocontrol agent Trichoderma harzianum.
The data obtained, reported in Figure 4, showed that
compound 2 exhibits significant inhibitory activity at three
concentrations tested (10, 50, and 100 ppm), with an average
mycelial growth reduction of 57%, 43%, 38%, and 54% for Fol,
P. italicum, A. niger, and T. harzianum, respectively. Compound
1 was active against Fol and A. niger with an average inhibition
of 59% and 36%, respectively. Neither compound showed
activity against B. cinerea. The mixture of both compounds
3-Hydroxy Fatty Acids 3 and 4 Obtained by Acid Hydrolysis
a
of 1 and 2
a
Δδ(S−R) values (in ppm) reported are those of 3a/3b. Analogous
values have been found for 4a/4b.
for H-2 and negative Δδ(S−R) for H-4/H-5, but the measured
Δδ(S−R) values were −0.03 and −0.05 (H-2), +0.05 (H-4), and
+0.02 (H-5), which corresponds to the R-configuration. There-
fore, according to the Mosher model, the R-configuration at C-3
is indicated for compounds 3 and 4, and consequently also for 1
and 2.
D
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 4. Antifungal activity of compounds 1 and 2 against five common fungal species. Each compound, either alone or in combination, was tested
at the final concentrations of 10, 50, and 100 ppm, using a bioassay in microtiter plates. Fungal growth was measured spectrophotometrically at 595
nm after 72 h incubation, and inhibitory activity (IA) was calculated as the percentage of fungal biomass reduction compared to the control (EtOH =
0%). Bars indicate the standard deviation from four replicates.
did not supply higher activity than compounds 1 and 2 applied
alone on all tested fungi (Figure 4).
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a Perkin-Elmer 192 polarimeter equipped with a sodium
lamp (589 nm) and a 10 cm microcell. FAB-MS (recorded in a glycerol
matrix) were measured on a Prospec Fisons mass spectrometer. ESI-
MS experiments were performed on an Applied Biosystem API 2000
triple-quadrupole mass spectrometer. The spectra were recorded by
infusion into the ESI source using MeOH as solvent. ESI-MS1, -MS2
spectra were recorded on an API 2000 instrument. GC-MS analysis was
performed on a Carlo Erba instrument by an Agilent Technologies
There have been reports on antifungal activity of bacterial
surfactins,10,11 but these compounds are better known for their
antiviral and antibacterial activities as well as their surfactant
properties.4 Although the cellular mechanism of the antifungal
mode of action of such compounds is still unknown, their
efficacy and availability at low cost make the isolated surfactins
attractive potential natural fungicides.
E
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
6890N Network gas 236 using a HP-5 capillary column (30 m ×
0.25 mm i.d.) packed with 5% polyphenyl siloxane, helium carrier flow
10 mL min−1, and an FID detector operated at 2608. The column
temperature was maintained at 70 °C for 3 min and then gradually
equimolecular ratio to produce a final additive concentration of 10, 50,
or 100 ppm. The solvent ethanol was used as a negative control.
Bioassays were conducted in a 96-well plate against conidia of the five
chosen fungal strains (F. oxysporum f.sp. lycopersici, P. italicum,
A. niger, T. harzianum, and B. cinerea). Samples contained a mixture of
purified LPs (or ethanol as a control) mixed with PDB and 105 fungal
conidia mL−1. Plates were incubated at 28 °C on a rotary shaker
(200 rpm) for different times, and fungal growth was determined
spectrophotometrically by measuring absorption at 595 nm, using a
microplate reader (model 550, Bio-Rad). Data from bioassays were
submitted to variance analysis (ANOVA) using the SPSS software
(version 16.0 for Windows, SPSS Inc., Chicago, IL, USA), and means
were compared by Tukey’s test. The effective lipopeptide concen-
tration required to obtain 50% growth inhibition of tested fungi
(EC50) was calculated by regressing inhibition of fungal growth against
the logarithm of LP concentration. Commercial surfactin (C15, Sigma
Aldrich) was used as a control (data not show).
1
increased (10 °C/min) to 300 °C. H and 13C NMR spectra were
recorded on a Varian Unity Inova spectrometer at 500.13 and
125.77 MHz, respectively. Chemical shifts were referred to the residual
solvent signal (CD3OD: δH 3.31, δC 49.0).20 The multiplicities of 13C
1
NMR resonances were determined by DEPT experiments. H con-
nectivities were determined by using COSY and HOHAHA experi-
ments; the 2D HOHAHA experiments were performed in the phase-
sensitive mode (TPPI) using the MLEV-17 (mixing time 125 ms)
1
sequence for mixing. One-bond heteronuclear H−13C connectivities
were determined with a 2D HSQC pulse sequence with an interpulse
1
delay set for a JCH of 130 Hz. Two- and three-bond heteronuclear
1H−13C connectivities were determined with 2D HMBC experiments,
2−3
optimized for a
J
of 8 Hz. HPLC in isocratic mode was performed
CH
on a Varian 940-LC apparatus equipped with a refractive index detector
by using a μ-Bondapack C18 analytical column, 3.9 mm × 300 mm, i.d.
(Merck, USA). TLC on SiO2 with BuOH−H2O−CH3CO2H, 60:25:15
(BAW), for development was used. Spots were visualized first by UV
lamp and then with cerium sulfate in 2 N H2SO4 after heating as brown
spots (Rf = 0.6 for compound 1 and 0.5 for compound 2).
Compound 1. Yield: 8.4 mg; colorless, amorphous solid; [α]25
D
−26.5 (c 0.1 MeOH); IR (KBr) νmax 3413, 2928, 1151, 1045 cm−1; 1H
NMR data, see Table 1; 13C NMR data, see Table 1; HRFAB-MS
(positive ion) found m/z 1036.6928 [M + H]+, calcd for C53H94N7O13
m/z 1036.6914; FAB-MS (positive ion) m/z 1037 [M + H]+, m/z
1059 [M + Na]+, m/z 1075 [M + K]+; ESI-MS1/MS2 m/z 1059, 946,
933, 815, 718, 700, 693, 580, 467, 396, 281.
Production of 1 and 2 in Bacillus Cell Suspension: Extraction
and Purification Procedures. Samples of Bacillus amyloliquefaciens
strain BO5A were isolated from an orchard soil, collected in Larino
(Campobasso, Italy), and in particular from the rhizosphere of the
olive plant, using the methodology described by Boulter et al.21 and
characterized by routine bacteriological tests (Biolog and API test) and
by sequencing of the small 16S ribosomal subunit (CBS: Centraalbureau
voor Schimmelcultures, The Netherlands). Strain BO5A was routinely
cultivated in minimum salt liquid medium (MSLM: K2HPO4 2.5 g/L;
KH2PO4 2.5 g/L; (NH4)2HPO4 2.5 g/L; MgSO4 × 7H2O 0.2 g/L;
FeSO4 × 7H2O 0.01 g/L; MnSO4 × 7H2O 0.007 g/L; sucrose 10 g/L;
pH 7.5) at 28 °C and shaken at 120 rpm. Crude culture filtrate was
obtained by centrifugation at 14 000 rpm for 5 min and sterilized by
filtration (0.22 μm pore size). Bacterial cells were resuspended in sterile
PBS pH 7.4,22 and cell concentration was adjusted spectrophotometri-
cally (610 nm). Bacterial strains were stored at −80 °C with 30%
glycerol.
Bacteria cells were removed from the culture by centrifugation at
5000 rpm for 30 min. The lipopeptide fraction was obtained from
an aliquot (1 L) of cell-free supernatant that has been subjected to
acid precipitation, by the addition of 6 N HCl to a final pH of 2.0, and
incubated overnight at 4 °C. The acid precipitate was recovered by
centrifugation at 10.000 rpm for 20 min and lyophilized overnight.
The lipopeptides were extracted from the powder by using a CHCl3−
MeOH mixture (2:1, v/v) and passed through a polytetrafluoroethylene
(PTFE) filter. The extract was concentrated under vacuum to obtain a
crude lipopeptide mixture. The final step in the purification employed
high-performance liquid chromatography (HPLC) in isocratic mode on
a Varian 940-LC apparatus equipped with a refractive index detector.
The lipid mixture was chromatographed on an analytical C18 reversed-
phase column (Waters μBondapak C18 3.9 × 300 mm) eluted with 0.1%
of trifluoroacetic acid (TFA) in CH3CN−H2O (8:2, v/v), obtaining
pure compounds 1 (8.4 mg, tR = 5.4 min) and 2 (5.8 mg, tR = 4.2 min).
Antifungal Assays: Growth Conditions and Bioassays with
Purified Lipopeptides. Fusarium oxysporum f.sp. lycopersici (Fol)
strain 4287 (FGSC 9935) was isolated from infected tomato plants.23
Penicillium italicum (PI), Aspergillus niger (AN), Botrytis cinerea (BC),
and Trichoderma harzianum (TH) strain T-22 were obtained from the
Dipartimento di Agraria, Naples, Italy. These fungal strains were
cultured on solid potato dextrose agar (PDA, Liofilchem S.p.a., Italy)
or in submerged culture at 150 rpm in potato dextrose broth (PDB,
prepared from fresh potatoes) for 4 days at 28 °C on a rotary shaker
(200 rpm). Conidial suspensions for antifungal assays were prepared
as previously reported.24
Compound 2. Yield: 5.8 mg; colorless, amorphous solid; [α]25
D
−25.7 (c 0.1 MeOH); IR (KBr) νmax 3410, 2930, 1150, 1045 cm−1; 1H
NMR data, see Figure 1 and Table 1; 13C NMR data, see Figure 1 and
Table 1; HRFAB-MS (positive ion) found m/z 1022.6769 [M + H]+,
calcd for C52H91N7O13 m/z 1022.6757; FAB-MS (positive ion) m/z
1023 [M + H]+, m/z 1045 [M + Na]+, m/z 1061 [M + K]+; ESI-MS1/
MS2 m/z 1045, 932, 919, 801, 704, 686, 679, 566, 453, 382, 267.
Hydrolysis of Compounds 1 and 2. A 1.5 mg amount of
compounds was dissolved in 6 M HCl (0.30 mL) and stirred at 110 °C
for 40 h. After cooling, the solution was diluted with water, and the
lipophilic products were extracted with chloroform. Both phases
were concentrated under a stream of N2 and analyzed separately. The
lipophilic fraction, containing 3-hydroxy fatty acids 3 and 4 (Scheme 1),
respectively, was methylated with ethereal diazomethane and chromato-
graphed with a silica gel column, eluting with a gradient solvent system
from hexane with increasing amounts of ethyl acetate (from 20% to
70%). The fraction eluted with 50% hexane−ethyl acetate was shown
to contain a single compound identified by GC-MS and NMR (for 1,
methyl 3-methoxyhexadecanoate; for 2, methyl 3-methoxypentadeca-
noate). The water phase residue was analyzed by TLC (solvent
system: CH3CN−EtOAc−CH3CO2H−H2O, 20:6:4:3) and GC-MS as
N-trifluoroacetyl derivatives of the corresponding butyl esters following
the procedure developed by Gelpi et al.25 A mixture of L-valine,
L-leucine, L-glutamic acid, and L-aspartic acid taken in an equimolecular
ratio served as a standard. Using this procedure L-valine, L-leucine,
L-glutamic acid, and L-aspartic acid were identified and their 2:3:1:1
ratios was estimated according to the peak areas. To confirm the stereo-
chemistry of the amino acids, an aliquot of the aqueous phase from
the hydrolysate of 1 was subjected to enzymatic oxidation catalyzed
by L- and D-amino acid oxidases (Sigma Aldrich, USA) following the
procedure described by Debono et al.,12 showing reaction with the
former enzyme and remaining unchanged with the latter enzyme. By
this procedure, the amino acids were identified to belong to the L-series.
Synthesis of the (S)- and (R)-MTPA Esters of 3-Hydroxy
Acids 3 and 4. Two aliquots (1 mg) of each 3-hydroxy fatty acid
(3, 4; Scheme 1), obtained by acid hydrolysis (6 M HCl) of com-
pounds 1 and 2, were dissolved in dry pyridine and allowed to react
overnight with (S)- and (R)-MTPA chloride, affording the (S)- and
(R)-MTPA esters (3a/3b, 4a/4b; Scheme 1), respectively. Analysis of
the Δδ(S−R) values obtained by the 1H NMR spectrum for each couple
of diastereomeric MTPA esters (3a/3b, 4a/4b), according to the
Mosher model (Scheme 1), pointed to an R-configuration at C-3.
The two purified LPs (1 and 2) were dissolved in ethanol, alone or
in mixture, and added to the medium at a final concentration of 10, 50,
and 100 ppm. In mixed applications, each compound was used at
F
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(19) Yu, G. Y.; Sinclair, J. B.; Hartman, G. L.; Bertagnoli, B. L. Soil
Biol. Biochem. 2002, 34, 955−963.
ASSOCIATED CONTENT
■
S
* Supporting Information
(20) Barile, E.; Bonanomi, G.; Antignani, V.; Zolfaghari, B.; Sajjadi, S.
E.; Scala, F.; Lanzotti, V. Phytochemistry 2007, 68, 596−603.
(21) Boulter, J. I.; Boland, G. J.; Trevons, J. T. World J. Microbiol.
Biotechnol. 2000, 16, 115−134.
NMR and MS data of compounds 1 and 2 (Figures S1−S8) are
available free of charge via the Internet at http://pubs.acs.org.
(22) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A
Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory: Cold
Spring Harbor, NY, 1989.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +39 081 2539459. Fax: +39 081 7754942. E-mail:
lanzotti@unina.it.
(23) Di Pietro, A.; Garcia-Maceira, F. I.; Meglecz, E.; Roncero, M. I.
G. Mol. Microbiol. 2001, 39, 1140−1152.
(24) Di Pietro, A.; Roncero, M. I. G. Mol. Plant-Microbe Interact.
1998, 11, 91−98.
Notes
The authors declare no competing financial interest.
(25) Gelpi, E.; Koenig, W. A.; Gilbert, J.; Oro,
1969, 7, 604−613.
̀
J. J. Chromatogr. Sci.
ACKNOWLEDGMENTS
■
The authors are grateful to CERMANU and CSIAS, University
of Naples Federico II, for NMR and MS spectra. We thank
Prof. A. Di Pietro, Departamento de Genetica, University of
́
Cordoba, Spain, for providing the Fol strain. We thank Prof.
F. Scala and Dr. S. Wo, Dipartimento di Agraria, University of
Naples Federico II, Italy, for free access to the laboratories for
antifungal assays on PI, TH, AN, and BC. We acknowledge
financial support of bilateral project “Molecular signalling
between biocontrol agents and fungal pathogens in the
rhizosphere of agronomically important crops” by Spanish
MEC (HI2007-0011) and Italian MIUR (IT08F0A8E8), and
PRIN project “Studies on biocontrol agents and their
biomolecules to optimize the suppressive activity of natural
amendments against soilborne pathogens of horticultural crops”
by Italian MIUR (20089LSZ2A_003).
REFERENCES
■
(1) Katz, E.; Demain, A. C. Bacteriol. Rev. 1977, 41, 449−474.
(2) Bonmatin, J. M.; Laprevote, O.; Peypoux, F. Comb. Chem. High
Throughput Screening 2003, 6, 541−556.
́
(3) Toure, Y.; Ongena, M.; Jacques, P.; Guiro, A.; Thonart, P. J. Appl.
Microbiol. 2004, 96, 1151−1160.
(4) Ongena, M.; Jacques, P. Trends Microbiol. 2008, 16, 115−125.
(5) Gilardi, G.; Garibaldi, A.; Gullino, M. L. Phytoparasitica 2007, 35,
457−465.
(6) Asaka, O.; Shoda, M. Appl. Environ. Microbiol. 1996, 62, 4081−
4085.
(7) Romero, D.; de Vicente, A.; Rakotoaly, R. H.; Dufour, S. E.;
Veening, J.-W.; Arrebola, E.; Cazorla, F. M.; Kuipers, O. P.; Paquot,
M.; Perez-Garcia, A. Mol. Plant-Microbe Interact. 2007, 20, 430−440.
(8) Zhao, Z.; Wang, Q.; Wang, K.; Brian, K.; Liu, C.; Gu, Y. Biores.
Technol. 2010, 101, 292−297.
(9) Jacques, P. In Biosurfactants; Soberon-Chavez, G., Ed.; Springer:
Berlin, 2011; Vol. 20, pp 57−91.
(10) Romano, A.; Vitullo, D.; Di Pietro, A.; Lima, G.; Lanzotti, V. J.
Nat. Prod. 2011, 74, 145−151.
(11) Vitullo, D.; Di Pietro, A.; Romano, A.; Lanzotti, V.; Lima, G.
Plant Pathol. 2012, 61, 689−699.
(12) Debono, M.; Barnhart, M.; Carrell, C. B.; Hofmann, J. A.;
Occolowitz, J. L.; Abbott, B. J.; Fukuda, D. S.; Hamill, R. J. J. Antibiot.
1987, 40, 761−777.
(13) Liu, X. Y.; Yang, S. Z.; Mu, B. Z. J. Pept. Sci. 2008, 14, 864−875.
(14) Yang, S. Z.; Wei, D. Z.; Mu, B. Z. J. Biochem. Biophys. Methods
2006, 68, 69−74.
(15) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512−515.
(16) Obtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem.
Soc. 1991, 113, 4092−4096.
(17) Kim, P. I.; Bai, H.; Bai, D.; Chae, H.; Chung, S.; Kim, Y.; Park,
R.; Chi, Y.-T. J. Appl. Microbiol. 2004, 97, 942−949.
(18) Arrebola, E.; Jacobs, R.; Korsten, L. J. Appl. Microbiol. 2010, 108,
386−395.
G
dx.doi.org/10.1021/np400119n | J. Nat. Prod. XXXX, XXX, XXX−XXX