Download Journal of Biotechnology

Author's personal copy
Journal of Biotechnology 140 (2009) 27–37
Contents lists available at ScienceDirect
Journal of Biotechnology
journal homepage: www.elsevier.com/locate/jbiotec
Genome analysis of Bacillus amyloliquefaciens FZB42 reveals its potential for
biocontrol of plant pathogens
X.H. Chen a,1 , A. Koumoutsi a,1,2 , R. Scholz a , K. Schneider b , J. Vater b , R. Süssmuth b , J. Piel c , R. Borriss a,∗
a
Institut für Biologie, Humboldt Universität Berlin, Chausseestrasse 117, D10115 Berlin, Germany
Institut für Chemie, Technische Universität Berlin, Franklinstr. 29, D-10587 Berlin, Germany
c
Kekulé-Institut für Organische Chemie und Biochemie, Rheinische Friedrich-Wilhelms-Universität Bonn, Gerhard-Domagk-Str. 1, D53121 Bonn, Germany
b
a r t i c l e
i n f o
Article history:
Received 1 August 2008
Received in revised form
23 September 2008
Accepted 21 October 2008
Keywords:
Bacillus amyloliquefaciens FZB42
Genome
Lipopeptides
Polyketides
a b s t r a c t
The genome of plant-associated Bacillus amyloliquefaciens FZB42 harbors an array of giant gene clusters
involved in synthesis of lipopeptides and polyketides with antifungal, antibacterial and nematocidal activity. Five gene clusters, srf, bmy, fen, nrs, dhb, covering altogether 137 kb, were shown to direct synthesis of
the cyclic lipopeptides surfactin, bacillomycin, fengycin, an unknown peptide, and the iron-siderophore
bacillibactin. In addition, one gene cluster encoding enzymes involved in synthesis and export of the
antibacterial dipeptide bacilysin is also functional in FZB42. Three gene clusters, mln, bae, and dfn, with
a total size of 199 kb were shown to direct synthesis of the antibacterial acting polyketides macrolactin,
bacillaene, and difficidin. In total, FZB42 dedicates about 340 kb, corresponding to 8.5% of its total genetic
capacity, to synthesis of secondary metabolites. On the contrary, genes involved in ribosome-dependent
synthesis of lantibiotics and other peptides are scarce. Apart from two incomplete gene clusters directing
immunity against mersacidin and subtilin, only one peptide-like compound has been detected in the
culture fluid that inhibits the growth of B. subtilis lacking the alternative sigma factor W.
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
2. Genome and genome analysis
The plant root-colonizing Bacillus amyloliquefaciens strain FZB42
is an environmental strain which is distinguished from the domesticated model organism Bacillus subtilis 168 by its ability to stimulate
plant growth and to suppress plant pathogenic organisms (Idriss
et al., 2002). FZB42 genome analysis revealed the presence of
numerous gene clusters involved in synthesis of non-ribosomally
synthesized cyclic lipopeptides (Koumoutsi et al., 2004) and
polyketides (Chen et al., 2006; Schneider et al., 2007) with distinguished antimicrobial action.
In light of the whole genome sequence information, we will
summarize here the impressive capability of B. amyloliquefaciens
FZB42 to produce a vast array of secondary metabolites aimed to
suppress competitive bacteria and fungi within the plant rhizosphere.
The circular chromosome consists of 3,918,589 bps with predicted 3695 protein-coding sequences (Chen et al., 2007). Software
package Mauve (Darling et al., 2004) was used to compare the
genomes of B. amyloliquefaciens FZB42, B. subtilis 168, B. licheniformis and B. pumilus. The four species belong to a closely related
taxonomic unit vernacularly called the B. subtilis group (Fritze,
2004). A core genome formed by the four members of the B. subtilis group consisted of 2139 genes sharing more than 50% identity to
each other. 310 genes with less than 50% amino acid sequence identity to any other Bacillus gene were defined as being unique for B.
amyloliquefaciens. The majority of the FZB42 genes were found conserved in B. subtilis (3271), while B. amyloliquefaciens and B. pumilus
have only 2378 genes in common (Fig. 1). Significant portions of the
FZB42 genome were also found conserved in B. cereus ZK (2342), B.
anthracis Sterne (2338), B. thuringiensis (2335), B. clausii (2162), B.
halodurans (2105), and Geobacillus kaustophilus (1995).
Notably, the genome of FZB42 harbors nine giant gene clusters
directing synthesis of bioactive peptides and polyketides by modularly organized mega-enzymes named non-ribosomal peptide
synthetases, NRPS and polyketide synthases, PKS. Synthesis of both
cyclic lipopeptides and polyketides is dependent on the presence of
a functional sfp gene product coding for a 4 -phosphopantetheinyl
transferase (Mootz et al., 2001). No counterparts for the gene clus-
∗ Corresponding author. Tel.: +49 30 20 93 81 37; fax: +49 30 20 93 81 27.
E-mail address: [email protected] (R. Borriss).
1
These authors contributed equally.
2
Present address: University of California, San Francisco, CA 94143-0868, USA.
0168-1656/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jbiotec.2008.10.011
Author's personal copy
28
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
3. Non-ribosomally synthesized peptides bacillomycin D
and fengycin
Fig. 1. Genome comparison of FZB42 with the genomes of the other members of the
B. subtilis group. Alignment of the FZB42 genome (BA) with the genomes of B. subtilis
(BS), B. licheniformis (BL) and B. pumilus (BP) was performed using the progressive
Mauve algorithm (Darling et al., 2004). Number of the genes shared by the pairs of
species compared is indicated at the lower part of the three central columns. Number
of FZB42 genes without orthologues in the respective alignment is indicated in the
upper part of the column. The left-hand column summarizes the number of genes
with orthologues detected in any of the three alignments (bottom) and the number
of genes unique for FZB42 (top). The right-hand column informs about the number
of “core” genes present in the four species.
ters responsible for synthesis of bacillomycin D, difficidin and
macrolactin and a hitherto unidentified hybrid peptide/polyketide
were detected in other Bacillus genomes. However, only one of
them, the nrs gene cluster, is localized within a region of deviating oligonucleotide pattern, which is indicative for a DNA island
acquired by horizontal gene transfer (Reva and Tümmler, 2005).
Five gene clusters, srf, bmy, fen, nrs, dhb, covering altogether 137 kb,
were shown to direct synthesis of the cyclic lipopeptides surfactin,
bacillomycin, fengycin (Koumoutsi et al., 2004), an unknown peptide, and the iron-siderophore bacillibactin. In addition, a gene
cluster encoding enzymes involved in synthesis and export of the
antibacterial dipeptide bacilysin is also functional in FZB42. Three
gene clusters, mln, bae, and dfn, with a total length of 199 kb were
shown to direct synthesis of the antibacterial acting polyketides
macrolactin, bacillaene, and difficidin (Chen et al., 2006; Schneider
et al., 2007). In total, FZB42 dedicates about 340 kb, corresponding to 8.5% of its total genetic capacity, to synthesis of secondary
metabolites (Fig. 2). That value exceeds those recently estimated
for an average Bacillus genome including B. subtilis 168 by two
times (Stein, 2005). By comparison, Streptomyces avermitilis devotes
6.4% of its genome secondary metabolite production (Omura et al.,
2001).
Despite their structural heterogeneity, non-ribosomal peptide
antibiotics share a common mode of synthesis, the multicarrier
thiotemplate mechanism (Stein et al., 1996). They are biosynthesized by multi-modular proteins termed non-ribosomal peptide
synthetases. Each elongation cycle in non-ribosomal peptide
biosynthesis needs the cooperation of three basic domains. (1) The
A domain (adenylation domain) selects its cognate amino acid and
generates an enzymatically stabilized aminoacyl adenylate. This
mechanism resembles the aminoacylation of tRNA synthetases during ribosomal peptide biosynthesis. (2) The PCP domain (peptidyl
carrier domain) is equipped with a 4 -phosphopantetheine (PPan)
prosthetic group to which the adenylated amino acid substrate
is transferred and bound as thioester. A 4 -phosphopantetheinetransferase (PPTase) converts the apo-form of the PCP into its
holo-form by loading the Ppan-cofactor to an active serine. Sfp is
a prototype of PPTases with wide substrate tolerance. Sfp-PPTase
plays an essential role in priming non-ribosomal peptide synthetases, siderophore synthetases (e.g. bacillibactin) and polyketide
synthases. It can covalently convert specific serine residues in
peptidyl- as well as in acyl- and aryl-carrier-proteins/domains (PCP,
ACP, ArCP), – to generate their active holo-forms. This occurs by
tethering the phosphopantetheinyl moiety of the cosubstrate coenzyme A (CoA) in phosphodiester linkage to the hydroxymethyl side
chain of the conserved active serine residue in the CP domains
(Walsh et al., 1997). (3) The formation of a new peptide bond
is catalyzed by condensation domains (C domains). The linear
organization of such core units (1–3) ensures the coordinated elongation of the peptide product. The assembly of the multifunctional
proteins of the peptide synthetases is reflected in its genetic organization following the colinearity rule (Duitman et al., 1999). Cyclic
lipopeptides as surfactin, fengycin, and iturin-like antibiotics are
widely spread in B. subtilis and related strains (Stein, 2005). The
domesticated B. subtilis 168 contains three gene clusters devoted
to non-ribosomal synthesis of surfactin, fengycin (synonymous to
plipastatin) and the siderophore bacillibactin, but is deficient in
their syntheses due to a frame shift mutation on the sfp gene which
converts PCP domains to their active form (Mootz et al., 2001).
In the genome of B. amyloliquefaciens FZB42 three gene clusters for the non-ribosomal biosynthesis of lipopeptides were
detected (Fig. 3, Koumoutsi et al., 2004; Chen et al., 2007). Two
of them are also present in the reference organism B. subtilis
168. Matrix-assisted laser desorption/ionization-time of flightmass spectrometry (MALDI-TOF-MS) revealed the expression of
surfactin, fengycin and bacillomycin D both in whole cells as well
as in surface extracts and crude culture supernatants (Koumoutsi et
Fig. 2. Overview about distribution of structural and regulatory genes (gene clusters) involved in non-ribosomal and ribosomal synthesis of antibiotics (lipopeptides, polyketides and peptides) in B. amyloliquefaciens FZB42. The alignment with the genomes of B. subtilis, B. licheniformis and B. pumilus was performed with the Mauve algorithm
(Darling et al., 2004). Position of the DNA islands 1–17 are shown at top. Vertical lines indicate degree of similarity with the genes of the other Bacillus species. Regulatory genes
are indicated in light grey. Abbreviations: srf, surfactin gene cluster; sfp, 4 -phosphopantetheine transferase (PPTase), necessary for non-ribosomal synthesis of lipopeptides
and polyketides; yczE, necessary for synthesis of polyketides and bacillomycinD, pksA, positive regulator of polyketides synthesis, mln, macrolactin gene cluster; bae, bacillaene
gene cluster; bmy, bacillomycinD gene cluster; fen, fengycin gene cluster; dfn, difficidin gene cluster; nrs, gene cluster for synthesis of unknown peptide; comA, regulator
of surfactin biosynthesis and competence genes; dhb, bacillibactin gene cluster; spa, immunity against subtilin, degUS, two-component system involved in regulation of
bacillomycinD synthesis; bac, bacilysin gene cluster, mrs, immunity against mersacidin.
Author's personal copy
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
29
Fig. 3. Presence of gene clusters involved in lipopeptide synthesis in FZB42. Progressive Mauve alignment was performed in comparison with the genomes of B. subtilis, B.
licheniformis and B. pumilus. The light grey vertical lines indicate that the genes are conserved in more than one Bacillus species. The dark grey lines indicate that they are
conserved in Bacillus subtilis only. Absence of vertical lines means that these genes are unique in FZB42.
al., 2004). The gene clusters of FZB42 for the non-ribosomal biosynthesis of lipopeptides were attributed to the biosynthesis of these
lipopeptides by generation of gene inactivation mutants using cassette mutagenesis which led to the loss of one of the lipopeptide
products. The surfactin (srf) and fengycin (fen) operons are organized and located as in the reference organism B. subtilis 168. A
large 37.2 kb DNA sequence was attributed to the biosynthesis of
bacillomycin D (Koumoutsi et al., 2004). The bmy operon was found
adjacent to the fen gene cluster which corresponds to the pps operon
in B. subtilis 168.
A functional dbh gene cluster shown to direct synthesis of the
bacillibactin iron-siderophore and a further gene cluster probably
directing non-ribosomal synthesis of a hitherto unidentified product was also present in the genome (Chen et al., 2007).
3.1. Surfactin
Surfactin is a heptapeptide with an LLDLLDL chiral sequence
linked by a ␤-hydroxy fatty acid consisting of 13–15 carbon atoms
to form a cyclic lactone structure. Surfactin is a potent biosurfactant and shows hemolytic, antimicrobial and antiviral activities
by altering membrane integrity (Peypoux et al., 1999). It is commercially used for efficient inactivation of mycoplasma in tissue
cultures and other pharmaceutical and biotechnological products
(Vollenbroich et al., 1997). Mutants of B. amyloliquefaciens, blocked
in surfactin biosynthesis, were shown to be severely impaired in
biofilm formation (Borriss, unpublished observation) as previously
reported for B. subtilis (Branda et al., 2001; Hofemeister et al., 2004).
While surfactin is required for formation of aerial structures on the
surface of colonies of B. subtilis, it simultaneously inhibited development of aerial hyphae and spores by cocultivated Streptomyces
coelicolor (Straight et al., 2006). Furthermore, it plays an important role in the swarming process of Bacillus subtilis (Kearns and
Losick, 2003; Julkowska et al., 2004, 2005). It is likely that surfactin synthesized by B. amyloliquefaciens FZB42 protects it against
other bacteria and enables it to form biofilms, thus equipping the
bacterium with powerful antagonistic advantages during surface
colonization of plant roots. Colonization of plant roots by B. subtilis
is associated with surfactin production and biofilm formation, and
strikingly, surfactin protects the plant against the infection by the
pathogen Pseudomonas syringae (Bais et al., 2004).
The 26.5 kb srf operon present in B. amyloliquefaciens FZB42
genome is organized in a similar manner as in B. subtilis 168.
It contains four open reading frames encoding three multifunctional proteins SrfA-C and an external thioesterase/acyltransferase
enzyme SrfD (Fig. 3). SrfA and B comprise three amino acidactivating modules, while SrfC is an one-module enzyme. In
the initiation process of surfactin formation acyltransferase SrfD
mediates the transfer of the ␤-hydroxy fatty acid substrate from
coenzyme A to the first module of SrfA. Here the ␤-hydroxy acyl
moiety is coupled with the starter amino acid l-Glu forming the first
peptide bond. SrfA, B and C interact in a coordinated manner and
provide the assembly line for elongation of the growing surfactin
chain up to the lipoheptapeptide intermediate which in the termination reaction is cyclized by macrolactone formation catalyzed by
the internal thiosterase-domain of SrfC. The surfactin biosynthetic
genes of B. subtilis 168 and B. amyloliquefaciens FZB42 exhibit identities between 72% (srfAA) and 83% (srfAC) on the amino acid level.
Author's personal copy
30
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
The genes present at the upstream flanking region of the operon
are hxlBAR, as in the case of B. subtilis 168. In FZB42 significant
modifications were detected in the region downstream of the srf
gene cluster (Koumoutsi et al., 2004). The B. subtilis genes ycxAB
are substituted by two ORFs with unknown function. One of them,
aat, located adjacent to srfD displays homology to aspartate aminotransferase and was also detected in the biocontrol strain B. subtilis
B3 (Yao et al., 2003). No dramatic change in the production of surfactin or of other lipopeptides was observed when aat was deleted
in strain FZB42 (Koumoutsi, 2006).
3.2. Bacillomycin D
Bacillomycin D is a member of the iturin family that comprises
iturin A, C, D and E, bacillomycin F and L, bacillopeptin and mycosubtilin (Moyne et al., 2004). Iturins contain one ␤-amino fatty acid
and seven ␣-amino acids with a d-tyrosine as the second amino
acid and two additional d-amino acids at positions 3 and 6. These
agents exhibit strong antifungal and hemolytic activities and limited antibacterial activity (Thimon et al., 1995).
The 37.2 kb bmy gene cluster is an insertion within the FZB42
genome, which is separated by only 25 kb from the fengycin gene
cluster. It comprises four genes (bmyD, bmyA, bmyB and bmyC)
without orthologues in B. subtilis 168 (Fig. 3). The first gene bmyD
encodes a putative malonyl coenzyme A transacylase with a molecular mass of 45.2 kDa which is similar to FabD in fatty acid synthesis
and is nearly identical to FenF of B. subtilis ATCC 6633 and B. subtilis RB14 operons (Duitman et al., 1999; Tsuge et al., 2001). It has
been shown that this enzyme is indispensable for iturin production
(Tsuge et al., 2001) and corresponds to the FenF-protein encoded
by a gene located upstream of the fengycin synthetase gene cluster
of B. subtilis F29-3 (Chen et al., 1995).
BmyA–C are organized like their counterparts in the iturin A and
mycosubtilin operons (Tsuge et al., 2001; Duitman et al., 1999). They
encode the peptide forming subunits of bacillomycin D synthetase.
Seven amino acid-activating modules were distinguished, one in
BmyA, four in BmyB and two in BmyC. Modules B1, B2 and C1 contain epimerization domains directing the conversion of amino acids
2, 3 and 6 into the d-configuration which is in accordance with the
LDDLLDL-chiral sequence of lipopeptides of the iturin family. At the
C-terminal end of BmyC a thioesterase domain is located which is
responsible for the cyclization and release of the lipoheptapeptide
intermediate.
Of particular importance for understanding the routes of
lipopeptide biosynthesis is the architecture of BmyA which catalyzes the initiation reactions in bacillomycin D formation. It is
a hybrid of a peptide synthetase, a fatty acid synthase and an
aminotransferase. In contrast to the biosynthesis of surfactin the
ultimate steps in the synthesis of the ␤-amino fatty acid residue
as well as its transfer and coupling with the starter amino acid laspartate are directly managed by BmyA. The amino acid-activating
module for the starter amino acid l-Asn is preceeded by four
domains with homologies to proteins involved in fatty acid synthesis. The first one resembles long-chain fatty acid CoA-ligases
containing a putative ATP-binding box. The second and fourth
domains show high similarities to acyl-carrier-proteins/domains
(ACP) with the characteristic 4 -phosphopantetheine motif. The
third domain shows striking homologies to ␤-ketoacyl synthetases.
These fatty acid synthase domains are followed by a domain with
a high homology to glutamate-1-semialdehyde aminotransferases
which contain pyridoxalphosphate (PLP) as cofactor. The fatty acid
and aminotransferase domains are connected to the first amino
acid-activating module via an extra condensation and thiolation
domain. From this architecture the following suite of events in the
initiation of the biosynthesis of bacillomycin D and related compounds can be inferred: In the first step the precursor of the fatty
acid component of bacillomycin is coupled with coenzyme A by the
acyl-CoA-ligase domain in an ATP dependent reaction. The activated fatty acid is then transferred to the 4 -PPan-cofactor of the
first acyl carrier domain. Simultaneously malonyl-CoA is loaded to
the 4 -Ppan-cofactor of the second ACP which is catalyzed by the
malonyl-CoA transacylase encoded by bmyD followed by condensation of the malonyl residue to the fatty acid precursor catalyzed by
the ␤-ketoacyl synthase. The formed ␤-keto-fatty acid intermediate
is converted into the ␤-amino fatty acid constituent of bacillomycin
D by the aminotransferase in a transamination reaction. It is coupled
with the starter amino acid l-Asn attached to the 4 -Ppan-cofactor
of BmyA by a thioester bond. According to this route bacillomycin
D biosynthesis is initiated.
Regions flanking the 37 kb bmy operon are characterized by
DNA rearrangements joining the gene cluster with sequences originally present in different regions of the B. subtilis chromosome.
In particular, on its downstream side two rearranged gene clusters
are situated: yxjCDEF and bioIBDFAW, located in B. subtilis at positions 4000 and 3088–3094 kb, respectively (Fig. 3). On the upstream
flanking side, genes located in B. subtilis at positions 1910–1943 kb
were detected (Koumoutsi et al., 2004). Notably, the bmy operon is
inserted at exactly the same position as the iturin A gene cluster in
B. subtilis RB14 (Tsuge et al., 2001) and the bmyL gene cluster in B.
subtilis A1/3 (Hofemeister et al., 2004).
3.3. Fengycin
Fengycin and the closely related plipastatin are cyclic lipodecapeptides containing a ␤-hydroxy fatty acid with 16–19 carbon
atoms. Four d-amino acids have been identified in its peptide portion. Fengycin A is composed of 1 d-Ala, 1 l-Ile, 1 l-Pro, 1 d-allo-Thr,
3 l-Glx, 1 d-Tyr, 1 l-Tyr, 1 d-Orn, whereas in fengycin B d-Ala is
replaced by d-Val. It is specifically active against filamentous fungi
and inhibits phospholipase A2 (Nishikori et al., 1986).
The fen five-gene cluster (fenA–E) present in FZB42 is about 25 kb
distant from bmy (Fig. 3). It is related to the pps operon in B. subtilis
168 and is situated at the corresponding locus. Because of its similarity to the fen gene cluster of the fengycin producers B. subtilis
F29-3 (Chen et al., 1995) and B. subtilis A1/3 (Steller et al., 1999),
the pps operon was assigned to fengycin biosynthesis, although
B. subtilis 168 does not produce this lipopeptide. Interestingly, in
the genome of B. subtilis ATCC 6633, the gene cluster coding the
biosynthesis of mycosubtilin, an iturin-like compound, is situated
at the same location (Duitman et al., 1999), suggesting that additional NRPS operons could be integrated in different ways in this
area either as an insertion or as a substitution of existing NRPS
operons.
The five open reading frames in the fen and pps operons code for
the corresponding multifunctional subunits FenA–E of the fengycin
multienzyme system comprising ten amino acid-activating modules in total (Steller et al., 1999). The proteins FenA–C are of similar
size and contain two of such modules that in turn activate and
couple the first six amino acid components of fengycin. FenD is
a three-module enzyme that activates l-Pro, l-Gln and l-Tyr in
positions 7–9 of fengycin. FenE contributes the C-terminal l-Ile
and is involved by its internal thioesterase domain in closing the
fengycin ring by macrolactone formation between the carboxyl
group of l-Ile and the free hydoxy group of l-Tyr in position 3.
Fengycin biosynthesis presumably is initiated in a similar way
as surfactin formation, because both lipopeptides contain a ␤hydroxy fatty acid and FenA uses the same starter amino acid
l-Glu as SrfA. However, in the fen operon no gene for an external thioesterase/acyltransferase enzyme has been found. Possibly,
SrfD may function as acyltransferase in fengycin biosynthesis,
too.
Author's personal copy
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
3.4. The nrs gene product
Besides the srf, bmy and fen operons, a fourth gene cluster, nrs, encoding a putative peptide/polyketide product was
detected within the genome of B. amyloliquefaciens FZB42
(Fig. 2). However, until now, no metabolite could be assigned
to this sequence. The flanking genes ytzC (77%) and ytqA
(92%) are homologous to the respective genes of B. subtilis.
Low GC content (26–34%) and local deviation of oligonucleotide usage pattern (Reva and Tümmler, 2005) corroborate
the island-specific character of this insertion (DNA island
12, Fig. 2). The nrsABCDEF gene cluster contains a putative
peptide synthetase, NrsD, consisting of two modules with Cysspecific A domains. Other members of the cluster possess
domains with homology to acyl-CoA-synthetases (nrsA), acylcarrier-proteins (nrsC), mcbC-like-oxidoreductases (nrsD), and
thioesterases (nrsE).
3.5. Bacillibactin
In FZB42 a functional dhb gene cluster was shown to direct
the synthesis of the catecholate iron-siderophore bacillibactin,
with the monomer unit 2,3-dihydroxybenzoyl-Gly-Thr (Chen et
al., 2007), which is a cyclic trimeric lactone. It is part of a
specific transport system enabling Bacillus cells to accumulate
and take up iron ions from their natural environment under
iron limitation. Like other non-ribosomally produced peptides its
synthesis is dependent on a functional Ppant-transferase (Sfp).
Therefore bacillibactin is not produced in B. subtilis 168, despite
the presence of the respective gene cluster (May et al., 2001). The
production of bacillibactin in FZB42 was evidenced by MALDI-TOF
MS analysis of lyophilized culture supernatants (Vater, unpublished).
The dhbACEBF gene cluster within the FZB42 genome was found
to be collinear to the dhb gene cluster of Bacillus subtilis 168.
The respective genes show identities of about 60% for dhbAC and
between 70 to 80% for dhbEBF.
3.6. Bacilysin
Bacilysin [l-alanyl-[2,3-epoxycyclohexanone-4]-l-alanine) is a
dipeptide antibiotic which contains a l-Ala residue at the Nterminus and a non-proteinogenic amino acid, l-anticapsin, at
the C-terminus. Bacilysin formation is catalyzed by an amino
acid ligase. Bacilysin is active against a wide range of bacteria
and against yeast Candida albicans due to the anticapsin moiety, which becomes released after uptake into susceptible cells.
Biosynthesis of bacilysin is encoded by the bacABCDE (former
ywfBCDEF) gene cluster (Inaoka et al., 2003; Steinborn et al.,
2005). The respective gene cluster in FZB42 was found to be
collinear to that in the genome of B. subtilis 168 (Fig. 2). Similarity to B. subtilis is in the range of 84–93%. Knock-out of
the bacA gene resulted in a complete block of bacilysin production (Chen et al., 2009). In contrast to the other non-ribosomal
peptides listed above, synthesis of bacilysin is Sfp independent.
3.7. Regulation of non-ribosomal lipopeptide synthesis
While the chemical principles for the non-ribosomal biosynthesis of lipopeptides are largely known (Stachelhaus et al., 2002),
little is known about the regulation mechanisms that control the
synthesis of cyclic lipopeptides in Bacilli. An exception is the case
of surfactin, for which studying the regulation of gene expression
received increased attention due its connection with the development of genetic competence.
31
The synthesis of surfactin is growth-phase-dependent and is
induced during transition to stationary phase. Transcription of the
srf operon is driven by a ␴A -dependent promoter and regulated
via a complex network, including the two-component regulatory system, ComAP (Nakano et al., 1991; Roggiani and Dubnau,
1993). Phosphorylated ComA binds upstream of the srf operon
and induces its expression. Therefore, systems involved in the
phosphorylation/DNA binding ability of ComA (ComXQ, RapC-CSF,
RapF) modify indirectly the formation of the antibiotic (Bongiorni
et al., 2005). PerR, a general repressor of the peroxide stress
regulon, was shown to positively regulate surfactin in a direct
manner, independently of ComA (Hayashi et al., 2005). In contrast
CodY, a GTP-activated global regulator, acts as a direct repressor under casamino acids rich conditions (Serror and Sonenshein,
1996).
Knowledge on transcriptional regulation of the other lipopeptides in B. subtilis is rather limited. The promoters of fengycin and
iturin operons have been successfully identified and show similarity to a housekeeping ␴A promoter (Lin et al., 1999; Tsuge et al.,
2001). Furthermore, even in the presence of a functional copy of
the sfp gene, synthesis of fengycin was found to be severely reduced
in Bacillus subtilis 168. The degQ promoter of the strain contains a
mutation suggesting that synthesis of fengycin and iturin-like compounds is also dependent on the DegQ regulator protein (Tsuge et
al., 1999, 2005).
Synthesis of bacillomycin D in B. amyloliquefaciens FZB42 is
driven by a stationary-phase-induced ␴A promoter. Three global
regulators, DegU, DegQ, ComA and the minor sigma factors ␴B
and ␴H that positively influence the transcriptional activation of
the bmy promoter (Pbmy ) have been identified. Direct control on
Pbmy is exerted by DegU, which occupies two distinct sites at
the bmy promoter (Koumoutsi et al., 2007). In contrast ComA
controls bmy expression indirectly via DegQ, which serves as an
auxiliary factor to DegU. Taking into consideration the strong similarity of the upstream regions between the promoters of itu and
bmy, it would be not surprising if the same global regulators
control the expression of iturin A. In addition to the transcriptional control described above, a second level of regulation seems
to exist. Deletion of both degU and yczE specifically abolished
the production of bacillomycin D. While DegU has an additional
effect on Pbmy activity, membrane spanning YczE protein exerts
its effect only in a posttranscriptional manner. Fig. 4 summarizes
our recent knowledge about bacillomycin D expression regulation.
3.8. Fungicidal activity is due to synergistic action of bacillomycin
D and fengycin
FZB42 inhibited growth of plant pathogenic fungi such as
Fusarium spp. including F. oxysporum, Gaeumannomyces graminis, Rhizoctonia solani, Alternaria alternatae, Botrytis cinerea and
Pythium aphanidermatum (Grosch et al., 2001). Mutants deficient
in the synthesis of defined lipopeptides were generated by a gene
replacement strategy via double cross-over homologous recombination. Biological activity of FZB42 and the respective mutants were
assayed in direct growth tests and by bioautography (Koumoutsi et
al., 2004). Mutant strains deficient in bacillomycin D production
were severely impaired in their antifungal activity suggesting that
bacillomycin D contributes significantly to the antifungal action of
FZB42. Mutants bearing knock-out mutations within the srf and
fen gene clusters still retained their antifungal character. Interestingly, fungi grew unaffected in presence of the double mutant AK3
bearing bmyA::EmR and fenA::CmR mutations indicating synergistic action of bacillomycin D and fengycin against the target
microorganism, (Koumoutsi et al., 2004).
Author's personal copy
32
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
Fig. 4. Regulatory network directing synthesis of bacillomycin D in FZB42. DegU transcriptionally and posttranscriptionally controls bacillomycin D synthesis, while YczE
affects bacillomycin D in a posttranscriptional manner. ComA, DegQ and possibly alternative sigma factors indirectly affect expression of the bmy gene cluster. For further
details see text.
4. Antibacterial polyketides: bacillaene, difficidin and
macrolactin
Polyketides are a large family of secondary metabolites
that include many bioactive compounds with antibacterial,
immunosuppressive, antitumor, or other physiologically relevant
bioactivities. Bacterial polyketides are synthesized by type I polyketide synthases (PKSs), modularly organized assembly lines starting
from acyl-CoA precursors by decarboxylative Claisen condensations
(Khosla et al., 2007). In general their biosynthetic pathway follows the same logic as in non-ribosomally synthesized peptides
and requires at least three domains (Walsh, 2004): an acyltransferase (AT), a ketosynthase (KS) and an acyl-carrier-protein (ACP)
which – like PCPs in NRPSs – needs to be activated by a PPTase. Usually the order of the modules dictates the sequence of biosynthetic
events. While gene clusters encoding type I polyketide synthases
are widespread in different taxonomic groups, such as Pseudomonas
spp., actino-, cyano-, and myxobacteria, the knowledge about PKS
gene clusters in bacilli was until recently restricted to the cryptic pksX gene cluster of B. subtilis 168 (Albertini et al., 1995) and
zwittermicin biosynthesis in B. cereus (Emmert et al., 2004). Polyketides difficidin/oxydifficidin, bacillaene (Hofemeister et al., 2004)
and macrolactin (Jaruchoktaweechai et al., 2000) were shown to
be produced by several wild type strains belonging to B. subtilis
and B. amyloliquefaciens. We could demonstrate that the genome
of B. amyloliquefaciens FZB42 harbors three giant gene clusters
involved in polyketide synthesis together spanning nearly 200 kb
(Koumoutsi et al., 2004). Notably, none of the PKSs encoded by
the three pks gene clusters contained modules with cognate AT
domains necessary to catalyze the selection of polyketide building blocks. Instead, one or more genes encoding discrete ATs were
detected within upstream regions preceding the megasynthases
encoding genes in all the three pks gene clusters of FZB42 (Chen
et al., 2006). This variant of type I polyketide synthase architecture was described as “trans-AT” type I PKS or “AT-less” in gene
clusters involved in pederin (Piel, 2002) and leinamycin (Cheng et
al., 2003) biosynthesis and seems to be more abundant than previously anticipated (Shen, 2003). It was shown that the discrete ATs
act iteratively by loading malonyl CoA onto all PKS modules during
polyketide synthesis (Cheng et al., 2003). Using a combined genetic
and chemical approach we assigned the FZB42 pks gene clusters to
all the polyketides presently known to be produced by different
Bacillus strains (Chen et al., 2006; Schneider et al., 2007).
4.1. Bacillaene
The bae gene cluster exhibits an almost identical architecture to
the pksX gene cluster of B. subtilis 168. The only differences are a
missing pksA-like gene encoding a putative transcriptional regulator (however, the close homologue pksR was detected as a discrete
gene in a region far upstream from the cluster), and the absence of
a module containing a KS and an ACP domain in BaeM (Fig. 5). The
bae gene cluster contains two NRPS modules that suggested the
presence of two amino acid building blocks in the structure. One
anticipated amino acid was Gly, as deduced from the non-ribosomal
code of the adenylation domain located within BaeJ, the other was
unknown, since the corresponding code could not be assigned. The
presence of three discrete acyltransferases BaeC, D, and E in the
upstream region of the bae operon is another distinguished trait
shared with the pksX system.
The pksX cluster has been known from B. subtilis for over a
decade, but its metabolic product remained elusive. Recently it
was shown by cryoelectron microscopy of immunolabelled B. subtilis cells that the PKS proteins are formed in vivo and assemble
into a remarkable organelle-like structure larger than the ribosome
(Straight et al., 2007). This suggested that the gene cluster is more
than an inactive evolutionary relic. Activity could also be assigned
to individual protein components of the B. subtilis PKS. Fouriertransform mass spectrometry (FTMS) demonstrated that PksC acts
as malonyl-loading acyltransferase (Calderone et al., 2006) and that
during biosynthesis a keto group is converted into a carbon branch
(␤-branch) by a reaction analogous to the conversion of acetoacetylCoA to isopentenylpyrophosphate in terpene biosynthesis. By gene
inactivation and mass spectrometric studies, we could ultimately
assign the bae cluster to a known natural product, the antibiotic
bacillaene (Chen et al., 2006). Using strain OKB105, a derivative
of B. subtilis 168 harboring an intact copy of the PPTase Sfp, we
also assigned bacillaene to the homologous pksX gene cluster in B.
Author's personal copy
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
33
Fig. 5. Presence of gene clusters involved in polyketide synthesis in FZB42. Progressive Mauve alignment was performed in comparison with the genomes of B. subtilis, B.
licheniformis and B. pumilus. Difficidin and macrolactin genes are without orthologues in other known Bacillus genomes. The same colour code as described in Fig. 3 was used.
subtilis 168 (Chen et al., 2006). Bacillaene has been described as
a highly unstable inhibitor of prokaryotic protein synthesis with
a partially characterized open-chain polyenic structure and the
empiric formula C35 H48 O7 (Patel et al., 1995). However, since the
NRPS modules indicated the presence of two nitrogen atoms, this
formula was suspected to be incorrect. Indeed, in vitro experiments
performed with components of the pksX NRPS modules strongly
suggested that bacillaene contains Gly and Ala (Dorrestein et al.,
2006). The full structure elucidation was complicated by the instability of the polyketide, however, the Clardy and Walsh groups
eventually succeeded in a preliminary structure by NMR analysis of partially purified extracts, establishing that the compound
is an open-chain enamine acid with an extended polyene system
(Butcher et al., 2007).
A comparison of the bacillaene structure to the architecture of
the PKS revealed numerous deviations from the colinearity rule,
according to which the modular architecture usually closely reflects
the order of functional groups in a polyketide. This is in contrast
to PKSs with integrated AT domains (cis-AT PKSs), but very common among trans-AT PKSs. To gain insights into the mechanism
of polyketide chain elongation, several truncated versions of the
Bae PKS were generated by fusing the 3 -terminal thioesterase
(TE) gene region to upstream modules (Moldenhauer et al., 2007).
Since the TE hydrolyzes assembled polyketides that are covalently
bound to the enzyme, shortened polyketides were expected as new
products. However, for all constructed strains, even those with
inactivated TE regions in an otherwise full-length bae PKS system, entire series of bacillaene intermediates in various stages of
chain elongation were detected, which were released from the PKS
by an as-yet unknown mechanism. This serendipitous discovery
allowed us to directly observe virtually the entire bacillaene pathway by analysis of its intermediates. In this way, non-canonical
PKS functions, such as the introduction of double bonds on two
successive modules, or the timing of ␤-branching could be examined. The data also confirmed earlier observations that a double
bond is introduced by a cytochrome P450 in a post-PKS process
by dehydrogenation of a single bond (Reddick et al., 2007). Further insights into the function of the bacillaene and other trans-AT
PKSs were obtained by an evolutionary analysis of keto synthases
(KS) domains, which revealed a tight correspondence between
substrate specificity and phylogeny of each KS (Nguyen et al.,
2008). This resulted in an extension of the classical colinearity
rules to trans-AT PKSs, which now allows to correlate PKS architecture with polyketide structure also for these seemingly aberrant
enzymes.
4.2. Difficidin and oxydifficidin
Similar to the assignment of the bacillaene bae gene cluster,
difficidin and oxydifficidin were identified as products of the dfn
(formerly dif) gene cluster in B. amyloliquefaciens FZB42 by MS
analysis of inactivation mutants (Chen et al., 2006). The cluster is
without counterpart in the genome of Bacillus subtilis 168, however the compounds were first detected in culture broths of two
other B. subtilis strains. The polyketides are highly unsaturated 22membered macrocylic polyene lactone phosphate esters (Wilson et
al., 1987) with broad spectrum antibacterial activity (Zimmerman
et al., 1987). Difficidin has been shown to inhibit protein biosynthesis (Zweerink and Edison, 1987), but the exact molecular target
remains unknown. The dfn gene sequences exhibit a reasonable
Author's personal copy
34
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
colinearity with the polyketide structure to deduce a biosynthetic
model proposed by Chen et al., 2006. The only putative post-PKS
modification is the introduction of the phosphate ester function,
which is likely to be catalyzed by the kinase homolog DfnB. In
contrast to bae, but like the third pks gene cluster (mln), the dfn
system encodes only one trans-AT that presumably acts iteratively
in acyl transfer. Difficidin represents the main antibacterial activity
produced by FZB42 (Chen et al., 2006). Especially promising is its
suppressive action against the enterobacterium Erwinia amylovara,
a devasting plant pathogen causing necrotrophic fire blight disease
of apple, pear, and other rosaceous plants (Chen et al., 2009).
4.3. Macrolactin
Apart from bacillaene and difficidin/oxydifficidin, a third
polyketide with macrolide-like structure, macrolactin, is produced
by B. amyloliquefaciens FZB42. Macrolactins, originally detected in
an unclassified deep-sea marine bacterium, have been previously
reported from several other Bacillus strains (Jaruchoktaweechai et
al., 2000). The macrolactin carbon skeleton contains three separate diene structure elements in a 24-membered lactone ring
(Gustafson et al., 1989). Until now, at least 17 macrolactins have
been described and one of them, 7-O-malonyl-macrolactin A, has
been recently reported as efficient against Gram-positive bacterial pathogens (Romero-Tabarez et al., 2006). Four members of the
macrolactin family were detected in B. amyloliquefaciens FZB 42,
macrolactin A, macrolactin D as well as 7-O-malonyl- and 7-Osuccinyl-macrolactin. As a consequence of the structure assignment
of macrolactin and inactivation of the gene cluster previously designated pks2, the cluster was assigned to macrolactin synthesis in
FZB42 (Schneider et al., 2007) and renamed to mln-gene cluster
(Fig. 5). The biosynthesis of macrolactin was further investigated
by feeding experiments with 13 C-acetate. 13 C-labelled macrolactin
A revealed an alternating labelling of its carbon skeleton with 13 C,
indicating that acetate/malonate was used as the sole precursor for
the carbon backbone. The macrolactin structure is compatible with
Fig. 6. Phylogenetic tree based on alignment of partial cheA nucleotide sequences. The cheA gene encodes the two-component histidine kinase CheA. The scale bar indicates
0.1 nucleotide substitutions per nucleotide position. Apart of selected representatives of B. amyloliquefaciens, other strains belonging to the B. subtilis group including the type
strains of B. amyloliquefaciens (DSM7), B. licheniformis (BL01250), B. subtilis ssp. subtilis (168), B. subtilis ssp. spizenii (DSM 397) and B. mojavensis (DSM 9205) were used in this
analysis. Strains with commercial importance are IN937 and INR7 (J. Kloepper, Auburn AL), FZB24 (Abitep GmbH, Germany and Taegro® Earth Biosciences Inc., USA), GB03
(Kodiak® , Gustafson Biologicals, Plano, TX, USA), QST713 (Serenade® , Agraquest Inc., Davis, CA, USA), RUS1 (Baksib F® , Siberia, Russia), U-5036 and U-5017 (UCM, Ukranian
Collection of Microorganisms, Kyiv, Ukraine), B9601-Y2 (Yunnan Agricultural University, China), and strains B2 and B3 (Nanjing Agricultural University, China). Strains tested
for secondary metabolites are indicated in bold letters. Polyketide production and presence of the respective genes are indicated in brackets along the strain names. Detection
of polyketides and of the respective genes were performed as described previously (Chen et al., 2006; Schneider et al., 2007). A phylogenetic tree based on partial gyrA
nucleotide sequences (encoding DNA gyrase subunit A) yielded a similar dendrogram (not shown).
Author's personal copy
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
the domain organization of the mln-operon. Similarly to bae and
dfn, mln is a modular polyketide synthase system of type I which
exhibits a trans-acyltransferase architecture using a discrete acyltransferase enzyme that iteratively assembles macrolactin. Various
post-assembly-line modifications of the macrolactin scaffold can
take place, such as O-methylation at the hydroxyl groups, attachment of a malonyl or succinyl residue, or glycosidation. Currently,
the tailoring reactions cannot be assigned to any enzyme, since corresponding methyl-, acyl- and glycosyl transferases are not found
in the proximity of the mln cluster encoding macrolactin biosynthesis.
5. Comparison with B. subtilis: differences in spectrum of
secondary metabolites
35
et al., 2007). A recent survey of the genomes of four other B.
subtilis strains (http://www.bacillusgenomics.org/bsubtilis/), corroborated our finding that strains belonging to B. subtilis sensu
stricto are unable to produce the polyketides difficidin and macrolactin and are often impaired in their ability to produce lipopeptides
other than surfactin. This suggests that FZB42 can be considered
a paradigm for an own group of plant-associated Gram-positive
bacteria with a huge potential for biocontrol and plant growth promotion. The complete genome sequence along with its amenability
to genetic manipulation, should facilitate exploitation of the hitherto unappreciated potential of strain FZB42 to produce secondary
metabolites for developing agrobiotechnological agents with predictable features.
Acknowledgments
In contrast to its large potential for non-ribosomal synthesis of
lipopeptides and polyketides, FZB42 does not produce most of the
ribosomally synthesized peptide antibiotics that B. subtilis 168 does.
FZB42 does not contain the gene clusters of lantibiotics subtilosin
and the SPß prophage-encoded sublancin. Moreover, the bacterium
does not produce the antibiotic-like killing factor Skf (sporulation
killing factor) or the toxic protein SdpC (sporulation delay protein,
Gonzalez-Pastor et al., 2003). Notably, SdpC is present only in B. subtilis strains and orthologues of it have not been identified in other
bacteria including all Bacillus species sequenced to date (Butcher
and Helmann, 2006).
Our genome analysis revealed presence of two partial gene clusters involved in resistance against mersacidin and subtilin, however
the genes involved in synthesis of these peptides are missing (Fig. 1).
Using a B. subtilis 168 derivative devoid of alternative factor sigma
W, we recently could demonstrate that FZB42 produces a compound inhibiting growth of the B. subtilis sigW mutant (Scholz,
unpublished results).
6. Outlook: FZB42, a paradigm for plant-associated Bacillus
amyloliquefaciens
In order to characterize FZB42 taxonomically, the gyrA and
cheA gene sequences were chosen. The highly conserved gyrA
gene encodes the DNA gyrase subunit A; the more variable cheA
gene encodes the two-component sensor histidine kinase CheA,
which is crucial for regulating bacterial chemotaxis. Both genes
have previously been shown to be effective for resolving closely
related taxa of the B. subtilis group (Reva et al., 2004). Alignments performed with partial gyrA and cheA nucleotide sequences
revealed identical results. Fig. 6 presents a phylogenetic tree,
deduced from the alignment of partial cheA sequences, which
demonstrates relationship among the members of the B. subtilis
group isolated from the plant rhizosphere. Two related groups
can be distinguished. Group 1 represents B. amyloliquefaciens
and group 2, consisting of B. subtilis ssp. subtilis, B. subtilis ssp.
spizenii, and B. mojavensis, contains the closest relatives of B. subtilis. In contrast, B. licheniformis, although belonging to the B.
subtilis group, too, is more distantly arranged. The tree reveals
that B. amyloliquefaciens FZB42 is a representative of a group
of plant-colonizing and/or endophytic Bacillus strains (group 1a).
It has been previously proposed to consider this specific group
of rhizobacteria as an ecomorph, which is distinct from the
B. amyloliquefaciens type strain F (Reva et al., 2004). In fact,
nearly all Bacillus strains that are currently commercialized for
their plant growth promoting and biocontrol activities belong to
this ecotype. In contrast to the other members of the B. subtilis group and to the B. amyloliquefaciens type strain, too, these
strains produce a wide spectrum of polyketides including bacillaene, difficidin and macrolactin like FZB42 (Fig. 6, Schneider
This project was supported by funds of the competence network
Genome Research on Bacteria (GenoMikPlus), financed by the German Ministry for Education and Research and by funding of the
DFG (SFB624 and PI430/5-2 of the priority programme SPP1152) to
J.P. Yueqiu He (Yunnan Agricultural University), Xuewen Gao (Nanjing Agricultural University). Joseph Kloepper (Auburn University)
and Brian McSpadden Gardener (Ohio State University) are thanked
for kindly providing Bacillus strains with plant growth promoting
activities.
References
Albertini, A.M., Caramori, T., Scoffone, F., Scotti, C., Galizzi, A.T., 1995. Sequence
around the 159◦ region of the Bacillus subtilis genome: the pksX locus spans
33.6 kb. Microbiology 141, 299–309.
Bais, H.P., Fall, R., Vivanco, J.M., 2004. Biocontrol of Bacillus subtilis against infection
of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation
and surfactin production. J. Plant Physiol. 134, 307–319.
Bongiorni, C., Ishikawa, S., Stephenson, S., Ogasawara, N., Perego, M., 2005. Synergistic regulation of competence development in Bacillus subtilis by two Rap-Phr
systems. J. Bacteriol. 187, 4353–4361.
Branda, S.S., Gonzalez-Pastor, J.E., Ben-Jehuda, S., Losick, R., Kolter, R., 2001. Fruiting
body formation by Bacillus subtilis. Proc. Natl. Acad. Sci. U.S.A. 98, 11621–11626.
Butcher, B.G., Helmann, J.D., 2006. Identification of Bacillus subtilis sigma-dependent
genes that provide intrinsic resistance to antimicrobial compounds produced by
Bacilli. Mol. Microbiol. 60, 765–782.
Butcher, R.A., Schroeder, F.C., Fischbach, M.A., Straight, P.D., Kolter, R., Walsh, D.,
Clardy, J., 2007. The identification of bacillaene, the product of the pksX megacomplex in Bacillus subtilis. Proc. Natl. Acad. Sci. U.S.A. 104, 1506–1509.
Calderone, C.T., Kowtoniuk, W.E., Kelleher, N.L., Walsh, C.T., Dorrestein, P.C., 2006.
Convergence of isoprene and polyketide biosynthetic machinery: isoprenyl-Scarrier proteins in the pksX pathway of Bacillus subtilis. Proc. Natl. Acad. Sci.
U.S.A. 103, 8977–8982.
Chen, C.L., Chang, L.K., Chang, Y.S., Liu, S.T., Tschen, J.S., 1995. Transposon mutagenesis and cloning of the genes encoding the enzymes of fengycin biosynthesis in
Bacillus subtilis. Mol. Gen. Genet. 248, 121–125.
Chen, X.H., Vater, J., Piel, J., Franke, P., Scholz, R., Schneider, K., Koumoutsi, A., Hitzeroth, G., Grammel, N., Strittmatter, A.W., Gottschalk, G., Süssmuth, R.D., Borriss,
R., 2006. Structural and functional characterization of three polyketides synthase
gene clusters in Bacillus amyloliquefaciens FZB42. J. Bacteriol. 188, 4024–4036.
Chen, X.H., Koumoutsi, A., Scholz, R., Eisenreich, A., Schneider, K., Heinemeyer, I.,
Morgenstern, B., Voss, B., Hess, W.R., Reva, O., Junge, H., Voigt, B., Jungblut, P.R.,
Vater, J., Süssmuth, R., Liesegang, H., Strittmatter, A., Gottschalk, G., Borriss,
R., 2007. Comparative analysis of the complete genome sequence of the plant
growth-promoting bacterium Bacillus amyloliquefaciens FZB42. Nat. Biotechnol.
25, 1007–1014.
Chen, X.H., Scholz, R., Borriss, M., Junge, H., Mögel, G., Kunz, S., Borriss, R., 2009.
Difficidin and bacilysin produced by plant-associated Bacillus amyloliquefaciens
are efficient in controlling fire blight disease. J. Biotechnol. 140, 38–44.
Cheng, Y.-Q., Tang, G.-L., Shen, B., 2003. Type I polyketide synthase requiring a discrete acyltransferase for polyketide biosynthesis. Proc. Natl. Acad. Sci. U.S.A. 100,
3149–3154.
Darling, A.C.E., Mau, B., Blattner, F.R., Perna, N.T., 2004. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome Res. 14,
1394–1404.
Dorrestein, P.C., Blackhall, J., Straight, P.D., Fischbach, M.A., Garneau-Tsodikova, S.,
Edwards, D.J., Mc Laughlin, S., Lin, M., Gerwick, W.H., Kolter, R., Walsh, C.T.,
Lelleher, N.L., 2006. Activity screening of carrier domains within nonribosomal
peptide synthetases using complex substrate mixtures and large molecule mass
spectrometry. Biochemistry 45, 1537–1546.
Author's personal copy
36
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
Duitman, E.H., Hamoen, L.W., Rembold, M., Venema, G., Seitz, H., Saenger, W., Bernhard, F., Reinhardt, R., Schmidt, M., Ullrich, C., Stein, T., Leenders, F., Vater, J.,
1999. The mycosubtilin synthetase of Bacillus subtilis ATCC 6633: a multifunctional hybrid between a peptide synthetase, an amino transferase, and a fatty
acid synthase. Proc. Natl. Acad. Sci. U.S.A. 96, 13294–13299.
Emmert, E.A.B., Klimowicz, A.K., Thomas, M.G., Handelsman, J., 2004. Genetics
of zwittermicin A production by Bacillus cereus. Appl. Environ. Microbiol. 70,
104–113.
Fritze, D., 2004. taxonomy of the genus Bacillus and related genera: the aerobic
endospore-forming bacteria. Phytopathology 94, 1245–1248.
Gonzalez-Pastor, J.E., Hobbs, E.C., Losick, R., 2003. Cannibalism by sporulating bacteria. Science 301, 510–513.
Grosch, R., Junge, H., Kofoet, A., 2001. Evaluation of isolates of Bacillus-group against
Fusarium oxysporum f. sp. radicis-lycopersici on tomato plants in soilless culture.
Gartenbauwissenschaft 66, 262–269 (in German).
Gustafson, K., Roman, M., Fenical, W., 1989. The macrolactins, a novel class of antiviral
and cytotoxic macrolides from a deep-sea marine bacterium. J. Am. Chem. Soc.
111, 7519–7524.
Hayashi, K., Ohsawa, T., Kobayashi, K., Ogasawara, N., Ogura, M., 2005. The H2 O2
stress-responsive regulator PerR positively regulates srfA expression in Bacillus
subtilis. J. Bacteriol. 187, 6659–6667.
Hofemeister, J., Conrad, B., Adler, B., Hofemeister, B., Feesche, J., Kucheryava, N., Steinborn, G., Franke, P., Grammel, N., Zwintscher, A., Leenders, F., Hitzeroth, G., Vater,
J., 2004. Genetic analysis of the biosynthesis of non-ribosomal peptide- and
polyketide-like antibiotics, iron uptake and biofilm formation by Bacillus subtilis
A1/3. Mol. Gen. Genomics 272, 363–378.
Idriss, E.E.S., Makarewicz, O., Farouk, A., Rosner, K., Greiner, R., Bochow, H., Richter,
T., Borriss, R., 2002. Extracellular phytase activity of Bacillus amyloliquefaciens
FZB45 contributes to its plant-growth promoting effect. Microbiology 148,
2097–2109.
Inaoka, T., Takahashi, K., Ohnishi-Kameyama, M., Yoshida, M., Ochi, K., 2003. Guanine nucleotides, guanosine 5 -diphosphate and GTP-co-operatively regulate
the production of an antibiotic bacilysin in Bacillus subtilis. J. Biol. Chem. 278,
2169–2176.
Jaruchoktaweechai, C.S., Suwanboriux, K., Tanasupawatt, S., Kittakoop, P., Menasveta,
P., 2000. New macrolactins from a marine Bacillus sp. Sc026. J. Nat. Prod. 63,
984–998.
Julkowska, D., Obuchowski, M., Holland, I.B., Seror, S.J., 2004. Branched swarming
patterns on a synthetic medium formed by wild-type Bacillus subtilis strain 3610:
detection of cellular morphologies and constellations of cells as the complex
architecture develops. Microbiology 150, 1839–1849.
Julkowska, D., Obuchowski, M., Holland, I.B., Seror, S.J., 2005. Comparative analysis of
the development of swarming communities of Bacillus subtilis 168 and a natural
wild type: critical effects of surfactin and the composition of the medium. J.
Bacteriol. 187, 65–76.
Kearns, D.B., Losick, R., 2003. Swarming motility in undomesticated Bacillus subtilis.
Mol. Microbiol. 49, 581–590.
Khosla, C., Tang, Y., Chen, A.Y., Schnarr, N.A., Cane, D.E., 2007. Structure and
mechanism of the 6-deoxyerythronolide B synthase. Ann. Rev. Biochem. 76,
195–221.
Koumoutsi, A., Chen, X.-H., Henne, A., Liesegang, H., Hitzeroth, G., Franke, P., Vater,
J., Borriss, R., 2004. Structural and functional characterization of gene clusters
directing nonribosomal synthesis of bioactive cyclic lipopeptides in Bacillus amyloliquefaciens strain FZB42. J. Bacteriol. 186, 1084–1096.
Koumoutsi, A., 2006. PhD thesis. Humboldt University Berlin.
Koumoutsi, A., Chen, X.-H., Vater, J., Borriss, R., 2007. DegU and YczE regulate the synthesis of bacillomycin D by Bacillus amyloliquefaciens strain FZB42. Appl. Environ.
Microbiol. 73, 6953–6964.
Lin, T.P., Chen, C.L., Chang, L.K., Tschen, J.S., Liu, S.T., 1999. Functional and transcriptional analyses of a fengycin synthetase gene, fenC, from Bacillus subtilis.
J. Bacteriol. 181, 5060–5067.
May, J.T., Wendrich, T.M., Marahiel, M.A., 2001. The dhb operon of Bacillus
subtilis encodes the biosynthetic template for the catecholic siderophore 2,3dihydroxybenzoate-glycine-threonin trimeric ester bacillibactin. J. Biol. Chem.
276, 7209–7217.
Moldenhauer, J., Chen, X.-H., Borriss, R., Piel, J., 2007. Biosynthesis of the antibiotic
bacillaene produced by a giant polyketides synthase complex of the trans-AT
family. Angew. Chem. Int. Ed. Engl. 46, 8195–8197.
Mootz, H.D., Finking, R., Marahiel, M.A., 2001. 4 -Phosphopantetheine transfer in
primary and secondary metabolism of Bacillus subtilis. J. Biol. Chem. 276,
37289–37298.
Moyne, A.-L., Cleveland, T.E., Tuzun, S., 2004. Molecular characterization and analysis of the operon encoding the antifungal lipopeptide bacillomycin D. FEMS
Microbiol. Lett. 234, 43–49.
Nakano, M.M., Marahiel, M.A., Zuber, P., 1991. transcription initiation region of the
srfA operon, which is controlled by the comP-comA signal transduction system
in Bacillus subtilis. J. Bacteriol. 170, 5662–5668.
Nguyen, T.A., Ishida, K., Jenke-Kodama, H., Dittmann, E., Gurgui, C., Hochmuth, T.,
Taudien, S., Platzer, M., Hertweck, C., Piel, J., 2008. Exploiting the mosaic structure
of trans-acyltransferase polyketide synthases for natural product discovery and
pathway dissection. Nat. Biotechnol. 26, 225–233.
Nishikori, T., Naganawa, H., Muraoka, Y., Aoyagi, T., Umezawa, H., 1986. Plipastatins:
new inhibitors of phospholipase A2, produced by Bacillus cereus BMG302-fF67.
III. Structural elucidation of plipastatins. J. Antibiot. 39, 755–761.
Omura, S., Ikeda, H., Ishikawa, J., Hanamoto, A., Takahashi, C., Shinose, M., Takahashi,
Y., Horikawa, H., Nakazawa, H., Osonoe, T., Kikuchi, H., Shiba, T., Sakaki, Y., Hat-
tori, M., 2001. Genome sequence of an industrial microorganism Streptomyces
avermitilis: deducing the ability of producing secondary metabolites. Proc. Natl.
Acad. Sci. U.S.A. 98, 12215–12220.
Patel, P.S., Huang, S., Fisher, S., Pirnik, D., Aklonis, C., Dean, L., Meyers, E.,
Fernandes, P., Mayerl, F., 1995. Bacillaene, a novel inhibitor of prokaryotic protein synthesis produced by Bacillus subtilis: production, taxonomy, isolation,
hysico-chemical characterization and biological activity. J. Antibiot. (Tokyo) 48,
997–1103.
Peypoux, F., Bromating, J.M., Wallach, J., 1999. Recent trends in the biochemistry of
surfactin. Appl. Microbiol. Biotechnol. 51, 553–563.
Piel, J., 2002. A polyketide synthase-peptide synthetase gene cluster from an uncultured bacterial symbiont of Paederus beetles. Proc. Natl. Acad. Sci. U.S.A. 99,
1402–1407.
Reddick, J.J., Antolak, S.A., Raner, G.M., 2007. PksS from Bacillus subtilis is a
cytochrome P450 involved in bacillaene metabolism. Biochem. Biophys. Res.
Commun. 358, 363–367.
Reva, O.N., Dixelius, C., Meijer, J., Priest, F.G., 2004. Taxonomic characterization and
plant colonizing abilities of some bacteria related to Bacillus amyloliquefaciens
and Bacillus subtilis. FEMS Microbiol. Ecol. 48, 249–259.
Reva, O.N., Tümmler, B., 2005. Differentiation of regions with atypical oligonucleotide compososition in bacterial genomes. BMC Bioinformatics 6,
251.
Roggiani, M., Dubnau, D., 1993. ComA, a phosphorylated response regulator protein of Bacillus subtilis, binds to the promoter region of srfA. J. Bacteriol. 175,
3182–3187.
Romero-Tabarez, M., Jansen, R., Sylla, M., Lünsdorf, H., Häußler, S., Santosa, D.A., Timmis, K.N., Molinari, G., 2006. 7-O-Malonyl macrolactin A, a
new macrolactin antibiotic from Bacillus subtilis active against methicillinresistant Staphylococcus aureus, vancomycin-resistant enterococci and a
small-colonyvariant of Burkholderia cepacia. Antimicrob. Agents Chemother. 50,
1701–1709.
Schneider, K., Chen, X.-H., Vater, J., Franke, P., Nicholson, G., Borriss, R., Süssmuth, R.D.,
2007. Macrolactin is the polyketide biosynthesis product of the pks2 cluster of
Bacillus amyloliquefaciens FZB42. J. Nat. Prod. 70, 1417–1423.
Serror, P., Sonenshein, A.L., 1996. Cod Y is required for nutritional repression of
Bacillus subtilis genetic competence. J. Bacteriol. 178, 5910–5915.
Shen, B., 2003. Polyketide biosynthesis beyond the type I, II and III polyketide synthase paradigma. Curr. Opin. Chem. Biol. 7, 285–295.
Stachelhaus, T., Mootz, H.D., Marahiel, M.A., 2002. Nonribosomal assembly of peptide
antibiotics on modular protein templates. In: Sonenshein, A.L., Hoch, J.A., Losick,
R. (Eds.), Bacillus subtilis and its Closest Relatives. ASM, Washington, D.C., pp.
415–435.
Stein, T., Vater, J., Kruft, V., Otto, A., Wittmann-Liebold, B., Franke, P., Panico, M.,
McDowell, R., Morris, H.R., 1996. The multiple carrier model of nonribosomal
peptide biosynthesis at modular multienzymatic templates. J. Biol. Chem. 271,
15428–15435.
Stein, T., 2005. Bacillus subtilis antibiotics: structure syntheses and specific functions.
Mol. Microbiol. 56, 845–857.
Steller, S., Vollenbroich, D., Leenders, F., Stein, T., Conrad, B., Hofemeister, J., Jacques, P.,
Thonart, P., Vater, J., 1999. Structural and functional organization of the fengycin
synthetase multienzyme system from Bacilluas subtilis b213 and A1/3. Chem.
Biol. 6, 31–41.
Steinborn, G., Hajirezaei, M.-R., Hofemeister, J., 2005. bac genes for recombinant
bacilysin and anticapsin production in Bacillus host strains. Arch. Microbiol. 183,
71–79.
Straight, P.D., Willey, J.M., Kolter, R., 2006. Interactions between Streptomyces coelicolor and Bacillus subtilis: role of surfactants in raising aerial structures. J.
Bacteriol. 188, 4918–4925.
Straight, P.D., Fischbach, M.A., Walsh, C.T., Rudner, D.Z., Kolter, R., 2007. A singular
enzymatic megacomplex from Bacillus subtilis. Proc. Natl. Acad. Sci. U.S.A. 104,
305–310.
Thimon, L., Peypoux, F., Wallach, J., Michel, G., 1995. Effect of the lipopeptide antibiotic, iturin A, on morphology and membrane structure of yeast cells. FEMS
Microbiol. Lett. 128, 101–106.
Tsuge, K., Ano, T., Hirai, M., Nakamura, Y., Shoda, M., 1999. The genes degQ, pps, and
lpa-8 (sfp) are responsible for conversion of Bacillus subtilis 168 to plipastatin
production. Antimicrob. Agents Chemother. 43, 2183–2192.
Tsuge, K., Akiyama, T., Shoda, M., 2001. Cloning, sequencing, and characterization of
the iturin A operon. J. Bacteriol. 183, 6265–6273.
Tsuge, K., Inoue, S., Ano, T., Itaya, M., Shoda, M., 2005. Horizontal transfer of iturin
A operon, itu, to Bacillus subtilis 168 and conversion into an iturinA producer.
Antimicrob. Agents Chemother. 45, 3566–3573.
Vollenbroich, D., Pauli, G., Özel, M., Vater, J., 1997. Antimycoplasma properties and
application in cell culture of surfactin, a lipopeptide antibiotic from Bacillus
subtilis. Appl. Environ. Microbiol. 63, 44–49.
Walsh, C.T., Gehring, A.M., Weinreb, P.H., Quadri, L.E.N., Flugel, R.S., 1997. Posttranslational modification of polyketide and nonribosomal peptide synthases.
Curr. Opin. Chem. Biol. 1, 309–315.
Walsh, C.T., 2004. Polyketides and nonribosomal peptide antibiotics: modularity and
versatility. Science 303, 1805–1810.
Wilson, K.E., Flor, J.F., Schwartz, R.E., Joshua, H., Smith, J.L., Pelak, B.A., Liesch, J.M.,
Hensens, O.D., 1987. Difficidin and oxydifficidin: novel broad spectrum antibacterial antibiotics produced by Bacillus subtilis. 2. Isolation and physico-chemical
characterization. J. Antibiot. 40, 1682–1691.
Yao, S., Gao, X., Fuchsbauer, N., Hillen, W., Vater, J., Wang, J., 2003. Cloning, sequencing, and characterization of the genetic region relevant to biosynthesis of
Author's personal copy
X.H. Chen et al. / Journal of Biotechnology 140 (2009) 27–37
the lipopeptides iturin A and surfactin in Bacillus subtilis. Curr. Microbiol. 47,
272–277.
Zimmerman, S.B., Schwartz, C.D., Monaghan, R.L., Pelak, B.A., Weissberger, B., Gilfilan, E.C., Mochales, S., Hernandez, S., Currie, S.A., Tejera, E., Stapley, E.O., 1987.
Difficidin and oxydifficidin: novel broad spectrum antibacterial antibiotics pro-
37
duced by Bacillus subtilis. 1. Production, taxonomy and antibacterial activity. J.
Antibiot. 40, 1677–1681.
Zweerink, M.M., Edison, A., 1987. Difficidin and oxydifficidin: novel broad spectrum antibacterial antibiotics produced by Bacillus subtilis. III. Mode of action
of difficidin. J. Antibiot. 40, 1691–1692.