ABBS 2005,38(04): Study of the Antifungal Ability of Bacillus subtilis Strain PY-1 in vitro and Identification of its Antifungal Substance (Iturin A)

Received: December 25, 2005

Accepted: February 16, 2006

This work was supported by a grant from the National Natural Science
Foundation of China (No. 39880002)

*Corresponding author: Tel, 86-28-85164029; Fax, 86-28-85164029; E-mail, [email protected]

Abstract        A Bacillus strain,
denoted as PY-1, was isolated from the vascular bundle of cotton. Biochemical,
physiological and 16S rDNA sequence analysis proved
that it should belong to Bacillus subtilis.
The PY-1 strain showed strong ability against many common plant fungal
pathogens in vitro. The antibiotics produced by this strain were stable
in neutral and basic conditions, and not sensitive to high temperature. From
the culture broth of PY-1 strain, five antifungal compounds were isolated by
acidic precipitation, methanol extraction, gel filtration and reverse-phase
HPLC. Advanced identification was performed by mass spectrometry and nuclear
magnetic resonance spectroscopy. These five antifungal compounds were proved to
be the isomers of iturin A: A2, A3, A4, A6 and A7. In
fast atom bombardment mass spectrometry/mass spectrometry collision-induced
dissociation spectra, fragmentation ions from two prior linear acylium ions were observed, and the prior ion, Tyr-Asn-Gln-Pro-Asn-Ser-bAA-Asn-CO+, was first reported.

Key words        Bacillus subtilis;
antifungal compounds; FAB MS/MS CID spectrum; NMR spectroscopy; iturin A

Fusarium wilt causes
huge economic losses in a wide variety of crops [1]. The pathogen,
Fusarium oxysporum,
infects plants through the roots by direct penetration or wounds, colonizes
the vascular tissue and causes plant death [2]. Chemical soil
fumigation is the main treatment of Fusarium wilt.
Broad-spectrum biocides, particularly methyl bromide, can be used to fumigate
the soil, but they cause serious environmental damage [3]. Safer and more
efficient methods are not available at present.

Recently, scientists
have paid attention to biological methods of defense against plant diseases.
Control of pathogens by antagonistic microorganisms or their antibiotic
products is now considered a viable disease control technology [4–6]. A Bacillus strain with an effective
ability against the Fusarium wilt pathogen was
isolated from the vascular tissue of a cotton Fusarium
wilt-resistant strain, and named PY-1.

In our study, we
identified this strain to be a Bacillus subtilis strain
by biochemical, physiological and 16S rDNA sequence
analysis. In vitro antagonism experiments showed that B. subtilis PY-1 was effective against not only F. oxysporum,
but also many other agricultural fungal pathogens, implying it has great
potential as an agent for biological control of many fungal diseases. To
explore the antifungal mechanism, we analyzed the culture broth of B. subtilis PY-1. Five compounds with high antifungal
activity were separated from the culture filtrate by reverse-phase HPLC
(RP-HPLC) and proved to be a series of isoforms of iturin A by mass spectrometry and nuclear magnetic
resonance (NMR)
spectroscopy. A new fragmentation of iturin
A was also observed in the fast atom bombardment mass spectrometry/mass
spectrometry collision-induced dissociation (FAB MS/MS CID) experiment.

Materials and Methods

Microbial strains and
culture conditions

B. subtilis
PY-1 strain was separated from cotton vascular bundle by the Key
Laboratory of Biotechnology and Crop Quality Improvement at the Southwest
Agricultural University of China (

Identification of B. subtilis PY-1 by 16S rDNA
analysis

The biochemical and
physiological identification of the PY-1 strain was performed using the BD
Phoenix 100 Automated Microbiology System (BD Diagnostic Systems,

In vitro antagonism
experiments

The ability of the PY-1
strain to inhibit the growth of various plant fungal pathogens was tested in
Petri dishes containing PDA medium. The mycelial
plugs of each fungus were deposited in the center of the plates, and the
bacterium was inoculated on the edge, approximately

Production of antifungal
compounds

The PY-1 strain was
cultured in 100 ml KMB liquid medium (2% tryptone, 1%
glycerin, 1.5% K2HPO4, 1.5% MgSO4∙7H2O, pH 7.0) on
a constant temperature shaker (30 ºC, 200 rpm) for 12 h. The broth was
transferred to 900 ml fresh medium and incubated at 30 ºC, 200 rpm for 72 h.
During incubation, a 1-ml sample was taken every 12 h for antifungal assay. In
this study, all antifungal assays were performed by the paper disk assay (40 ml sample per paper disk) against Aspergillus

Effects of pH and temperature
on antimicrobial activity

In the pH stability
test, the filter-sterilized crude supernatant was adjusted to pH 1.0–14.0, and maintained for 24 h at 4 ºC. The
antifungal activity was assayed after the solution had been readjusted to pH
7.0. To test the effect of temperature, the samples of the culture broth were
exposed at 60 ºC, 80 ºC and 100 ºC for 30 min and 121 ºC for 15 min, then the
remaining activity was assayed after the samples were cooled to room
temperature.

Isolation of the
antifungal compounds

After centrifugation at

Structure
characterization by mass spectrometry and NMR spectroscopy

The electrospray
ionization (ESI) time of flight (TOF) mass spectrometry carried out by a BioTOF Q Bruker (

Results

Relationship between B.
subtilis PY-1 and other Bacillus strains

Biochemical and
physiological identification showed that the PY-1 strain should be a B. subtilis. By 16S rDNA
analysis, B. subtilis PY-1 showed
approximately 99% similarity to B. subtilis
P45B. A phylogenetic tree showing the relationship
between B. subtilis PY-1 and other Bacillus
strains is shown in Fig. 1.

In vitro inhibition
of fungal growth

In this experiment, PY-1
strain showed strong inhibition ability against all tested fungal plant
pathogens (Table 1). F. oxysporum appeared
to be one of the most sensitive species with 32% relative inhibition of mycelial
growth.

Activity of the culture
broth and its stability

As shown in Fig. 2,
the culture broth had the strongest antifungal activity after incubation for 72
h and 10 ml of the
filter-sterilized crude supernatant was sufficient to clearly inhibit mycelium
expansion. The result of a pH stability test is shown in Fig. 3(A). In a
neutral condition, the antifungal activity of the culture broth was the
highest, and it was also stable, when the pH ranged from 5.0 to 13.0 with only moderate
reduction. The antibiotics had weak solubility and stability in acidic
conditions. When the culture broth was adjusted to pH 3.0, with large
precipitation, the supernatant lost activity. If the precipitation was
dissolved immediately in neutral phosphate-buffered saline, more than 80%
activity could be recovered. But long-term treatment in acidic conditions
significantly weakened activity. Only approximately 30% activity remained after
the broth had been at pH 3.0 for 24 h, and the activity was completely
destroyed at pH 1.0 for 24 h. However, in alkaline pH, activity was more stable
than in acidic pH. A thermal stability test showed that the antifungal activity
was not sensitive to high temperature [Fig. 3(B)]. The culture
broth even maintained approximately 60% activity after being exposed to 121 ºC
for 15 min.

Isolation of the active
compounds

The result of a pH
stability test demonstrated that the antifungal compounds were insoluble in
acidic aqueous solution, and could be precipitated completely at pH 3.0.
Methanol could effectively dissolve the antifungal compounds. The methanol extract
of the precipitation at pH 3.0 showed dramatic activity against fungi
(approximately 30 times stronger than that of culture broth, assessed by two-fold dilution), which also
implied that the antifungal compounds should have higher hydrophobicity.
As shown in Fig. 4, using gel filtration and RP-HPLC, five pure
antifungal compounds (called Compounds 1–5) were obtained. Compound 1 was the component with the
largest abundance. The ratio of the abundance of Compounds 1–5 was approximately 4.4:1:1.4:1.5:1.

Mass spectrometry
analysis

ESI-TOF MS analysis
showed that these five compounds had three different molecular weights: (M+H)+ ions at m/z
1043 (Compound 1), 1057 (Compounds 2 and 3) and 1071 (Compounds 4 and 5). It
was implied that Compounds 1–5 might be
five isoforms of iturin A
[10,11]. For further identification, their sequences were analyzed by FAB-MS/MS
CID spectrometry. The CID spectrum of Compound 1 is shown in Fig. 5.
From the spectrum, five compositive a-amino acids could be identified from their
special fragments: Asn (m/z 87.06); Pro
(m/z 70.07); Ser (m/z 60.04); Gln
(m/z 101.07); and Tyr (m/z
136.08). An immonium ion of the b-amino acid belonging to the iturin
family (H2N+=CH-C11H23) could also be found at m/z 184.21
(198.23 for Compounds 2 and 3; 212.24 for Compounds 4 and 5). For a cyclopeptide, the peptide ring should be opened at a
certain peptide bond and first form a linear acylium
ion, then other peptide bonds could be broken to form fragment ions. Because of
a random ring-opening reaction, it was difficult to sequence the cyclopeptide without the presence of proline
residue. Easier breakage on the peptidyl-prolyl (Xaa-Pro) bond led to the formation of the major linear acylium ions [12]. For the presence of proline
residue in iturin A molecules, the main linear acylium ion should be Pro-Asn-Ser-bAA-Asn-Tyr-Asn-Gln-CO+ (bAA denotes b-amino acid) [13]. The formation of b-type and y-type
ions is shown in Fig. 6(A). In the CID spectrum, most of these fragment
ions as well as some related a- or c-type ions were found, as listed in Table
2. This result confirmed that the compounds were isoforms
of iturin A. However, another series of ions with
high abundance were also observed (Table 2). It was implied that the
prior ion, Tyr-Asn-Gln-Pro-Asn-Ser-bAA-Asn-CO+ [Fig.
6(B)], should be also formed in a high ratio.

Other major ions
included the fragment ions from molecular ions that lost NH3 or H2O groups: m/z 1026.53 (M+H–NH3)+, m/z 1025.55 (M+H–H2O)+, m/z 1009.50 (M+H–2NH3)+, m/z 1008.54 (M+H–H2O–NH3)+, m/z 992.48 (M+H–3NH3)+, m/z 991.52 (M+H-H2O–2NH3)+, m/z
975.47 (M+H–4NH3)+, m/z 974.51 (M+H–H2O–3NH3)+; ions with double positive charges: m/z
522.31 (M+2H)2+, m/z
505.29 (M+2H–2NH3)2+; and some
internal fragment ions: m/z 392.16 (AsnTyrAsn),
m/z 375.15 (AsnTyrAsn–H2O).

NMR spectroscopy
analysis

The entire spin systems
of amino acid residues were identified through 1H-1H COSY and
TOCSY experiments [14]. Amide groups of asparagine
and glutamine and the phenolic group of tyrosine
could be clarified from 1H-1H COSY and NOESY spectra. Corresponding with the
mass spectrometry result, the NOESY experiment could also be used to induce the
amino acid sequence (cyclo-Asn1-Tyr2-Asn3-Gln4-Pro5-Asn6-Ser7-bAA8) [15] (Fig. 7). A strong NOE between
Gln4 and Ser7 suggested that Gln4-Pro5-Asn6-Ser7 should form a b-turn structure. The trans-amide bond of
Gln4-Pro5 was indicated by the strong NOE correlation between the a proton of Gln4 and d protons of Pro5 [12,15]. -amino acid could be discriminated
by 1H-1H COSY experiment because its aminomethenyl
(NH2-CH) group connected
with two methylene groups (-CH2-). Their diolefine
tails were identified by 1D 1H-NMR,

Discussion

Antagonism is ubiquitous
in nature among different species. For a long time, people have been interested
in rationally making use of it in the areas of agricultural defense or therapy
of diseases. Plant fungal diseases are difficult to control and can cause huge
damage to economic crops. Environmental pollution, caused by abusing chemical
biocides, is another serious problem. Using antibiotic production bacteria to
control plant fungal diseases is a popular topic and has been studied
extensively [17]. Compared with
chemical biocides, many antibiotics produced by antagonistic strains have the
advantage of being easily decomposed, leaving no harmful residues [18].
According to the pathogenesis mechanism, we know that, as well as soil
fumigation, inhibiting F. oxysporum from
invading the vascular tissue of plants might be another valid method to control
Fusarium wilt. B. subtilis
PY-1 was isolated from the vascular tissue of a cotton Fusarium
wilt resistant strain, and showed strong inhibitory ability against many plant
fungal pathogens in vitro, especially F. oxysporum
and Exserohilum turcicum,
which may be why this cotton strain can resist Fusarium
wilt.

Analysis of the B. subtilis PY-1 broth showed that the main antifungal
compounds were five isomers of iturin A. Iturins are a group of similar cyclic lipopeptides
with high antifungal activity, which can modify the membrane permeability and
lipid composition, and inhibit the mycelium growth and sporulation
of fungi [19]. The structure of iturins is
characterized by an amphiphilic peptide ring, which
is composed of seven chiral amino acids including an
invariable D-Tyr2 with the constant sequence LDDLLDL, and a rare b-amino acid with a long hydrophobic diolefine tail [20]. There are some isomeric compounds in iturins, such as iturin A6 (bAA:C16) and mycosubtilin (bAA:C16):C48H74N12O14 (MW 1070 D) [21,22]; iturin
A2 (bAA:C14) and mixirin A [23]:C48H74N12O14 (MW 1042
D). It is not enough to identify these compounds just by their molecular
weight, and sequence analysis is necessary. FAB-TOF/TOF CID MS is a valid
method. Because of the presence of proline residue,
the main linear acylium ion would be formed by the
ring-opening reaction at the peptidyl-prolyl bond. As
for iturin A, major fragment ions were derived from
the prior ion: Pro-Asn-Ser-bAA-Asn-Tyr-Asn-Gln-CO+. However, in our research, by CID spectra data
analysis, two series of ions from different prior linear acylium
ions were observed, and the second prior ion, Tyr-Asn-Gln-Pro-Asn-Ser-bAA-Asn-CO+, was first reported in FAB-MS/MS CID spectrometry
experiment of iturin A. CID spectrum can be used as a
“fingerprint” to identify iturin A,
especially in a mixture [13,24]. This finding will be very useful to analyze
the spectrum and discriminate some special fragment ions, so as to identify iturin A quickly and correctly. NMR spectroscopy is
necessary to determine the structure of the b-amino acid residue and gives us more information about
the space conformation of the molecule. But NMR spectroscopy needs an amount of
pure sample and is inappropriate for fast identification.

With the
high production of iturin A and its
endogenic character, B. subtilis
PY-1 is a promising agent in the biocontrol of
fungal diseases in agriculture. The application in farmland and the biologic
security of B. subtilis PY-1 needs advanced
testing.

Acknowledgements

We are grateful to Ms.
Lei TENG of

References

 1   Inoue I, Namiki F, Tsuge T. Plant colonization by the vascular wilt fungus Fusarium oxysporum
requires FOW1, a gene encoding a mitochondrial protein. Plant Cell 2002, 14:
1869–1883

 2   Simons G, Groenendijk
J, Wijbrandi J, Reijans M, Groenen J, Diergaarde P, Van der Lee T et al. Dissection of the fusarium I2 gene cluster in tomato reveals six homologs and one active gene copy. Plant Cell 1998, 10: 1055–1068

 3   Fravel DR, Olivain C, Alabouvette C. Fusarium oxysporum
and its biocontrol. New Phytologist
2003, 157: 493–502

 4  Han JS, Cheng JH, Yoon TM, Song J, Rajkarnikar A, Kim WG, Yoo ID et
al. Biological control agent of common scab disease by antagonistic strain Bacillus
sp. sunhua. J Appl Microbiol
2005, 99: 213–221

 5   Cook RJ, Thomashow LS,
Weller DM, Fujimoto D, Mazzola M, Bangera
G, Kim DS. Molecular mechanisms of defense by rhizobacteria
against root disease. Proc Natl Acad
Sci  6   Klich MA, Arthur
KS, Lax AR, Bland JM. Iturin A: A potential new
fungicide for stored grains. Mycopathologia 1994,
127: 123–127

 7   Toure Y, Ongena M, Jacques P, Guiro A, Thonart
P. Role of lipopeptides produced by Bacillus subtilis GA 8   Roepstorff P, Fohlman J. Proposal for a common nomenclature for sequence
ions in mass spectra of peptides. Biomed Mass Spectrom 1984, 11: 601

 9   Biemann K. Mass
spectrometry of peptides and proteins. Annu Rev Biochem 1992, 61: 977–1010

10  Winkelmann G, Allgaier
H, Lupp R, Jung G. Iturin
AL¾a new long
chain iturin A possessing an unusual high content of
C16-b-amino
acids. J Antibiot (11  Peypoux F, Guinand
M, Michel G, Delcambe L, Das
BC, Lederer E. Structure of iturine
A, a peptidolipid antibiotic from Bacillus subtilis. Biochemistry 1978, 17: 3992–3996

12  Alassane W, Céline
L, Henri L, Francoise V, Yanjun Z, Bernard B.
Sequence and solution structure of cherimolacyclopeptides
A and B, novel cyclooctapeptides from the seeds of Annona cherimola.
Tetrahedron 2004, 60: 405–414

13  Yu GY, Sinclair JB, Hartman GL, Bertagnolli
BL. Production of iturin A by Bacillus amyloliquefaciens suppressing Rhizoctonia
solani. Soil Biol Biochem 2002, 34: 955–963

14  Wagner G, Kumar A, Wuthrich K. Systematic
application of two-dimensional 1H nuclear-magnetic-resonance techniques for
studies of proteins. 2. Combined use of correlated spectroscopy and nuclear overhauser spectroscopy for sequential assignments of
backbone resonances and elucidation of polypeptide secondary structures. Eur J Biochem 1981, 114: 375–384

15  Macura S, Ernst RR. Elucidation of cross
relaxation in liquids by two-dimensional NMR spectroscopy. Mol Phys 1980, 41:
95–117

16  Besson F, Peypoux
F, Michel G, Delcambe L. Identification of
antibiotics of iturin group in various strains of Bacillus
subtilis. J Antibiot (17  Raaijmakers JM, Vlami
M, de Souza JT. Antibiotic production by bacterial biocontrol
agents. Antonie Van Leeuwenhoek 2002, 81: 537–547

18  Cook RJ. Making greater use of introduced microorganisms for
biological control of plant pathogens. Annu Rev Phytopathol 1993, 31: 53–80

19  Latoud C, Peypoux
F, Michel G. Action of iturin A, an antifungal
antibiotic from Bacillus subtilis, on the yeast
Saccharomyces cerevisiae:
Modifications of membrane permeability and lipid composition. J Antibiot (20  Maget-Dana R, Peypoux
F. Iturins, a special class of pore-forming lipopeptides: Biological and physicochemical properties. Toxicology
1994, 87: 151–174

21  Isogai A, Takayama
S, Murakoshi S, Suzuki A. Structures of b-amino acids in antibiotics iturin A. Tetrahedron Lett 1982,
23: 3065–3068.

22  Peypoux F, Pommier
MT, Marion D, Ptak M, Das
BC, Michel G. Revised structure of mycosubtilin, a peptidolipid antibiotic from Bacillus subtilis. J Antibiot (23  Zhang HL, Hua HM, Pei
YH, 24  Leenders F, Stein TH, Kablitz
B, Franke P, Vater J. Rapid
typing of Bacillus subtilis strains by their
secondary metabolites using matrix-assisted laser desorption
ionization mass spectrometry of intact cells. Rapid Comm
Mass Spectrom 1999, 13: 943–949