Microbiology (2013), 159, 757–770
DOI 10.1099/mic.0.064691-0
An updated metabolic view of the Bacillus subtilis
168 genome
Eugeni Belda,1,2,3 Agnieszka Sekowska,4 François Le Fèvre,1,2,3
Anne Morgat,5 Damien Mornico,1,2,3 Christos Ouzounis,6,7
David Vallenet,1,2,3 Claudine Médigue1,2,3 and Antoine Danchin4,8
1
Correspondence
CEA, Institut de Génomique, Génoscope Laboratoire d’Analyse Bioinformatique en Génomique et
Métabolisme, 2 rue Gaston Crémieux, 91057 Evry, France
Eugeni Belda
[email protected]
2
Antoine Danchin
[email protected]
3
CNRS-UMR 8030, 2 rue Gaston Crémieux, 91057 Evry, France
UEVE, Université d’Evry, boulevard François Mitterrand, 91025 Evry, France
4
AMAbiotics SAS, Bldg G1, 2 rue Gaston Crémieux, 91000 Evry, France
5
Swiss Institute of Bioinformatics, CMU, 1 Michel-Servet, CH-1211 Genève 4, Switzerland
6
Institute of Applied Biosciences, Centre for Research and Technology Hellas (CERTH),
Thessaloniki, Greece
7
Donnelly Centre for Cellular and Biomolecular Research University of Toronto, Toronto, ON,
Canada
8
Department of Biochemistry, Li KaShing Faculty of Medicine, The University of Hong Kong, 21,
Sassoon Road, Hong Kong SAR, China
Received 7 November 2012
Revised 10 January 2013
Accepted 15 February 2013
Continuous updating of the genome sequence of Bacillus subtilis, the model of the Firmicutes, is
a basic requirement needed by the biology community. In this work new genomic objects have
been included (toxin/antitoxin genes and small RNA genes) and the metabolic network has been
entirely updated. The curated view of the validated metabolic pathways present in the organism as
of 2012 shows several significant differences from pathways present in the other bacterial
reference, Escherichia coli: variants in synthesis of cofactors (thiamine, biotin, bacillithiol), amino
acids (lysine, methionine), branched-chain fatty acids, tRNA modification and RNA degradation. In
this new version, gene products that are enzymes or transporters are explicitly linked to the
biochemical reactions of the RHEA reaction resource (http://www.ebi.ac.uk/rhea/), while novel
compound entries have been created in the database Chemical Entities of Biological Interest
(http://www.ebi.ac.uk/chebi/). The newly annotated sequence is deposited at the International
Nucleotide Sequence Data Collaboration with accession number AL009126.4.
INTRODUCTION
Thousands of genome sequences have now been published.
New ones appear daily. It is no longer possible to annotate
manually even the restricted set of the major features of the
genomic objects they encode. Yet, it is essential that the
Abbreviations: BBH, bidirectional best hits; CanOE, candidate genes for
orphan enzymes; CDS, predicted protein-coding sequence; ChEBI,
chemical entities of biological interest; EC, Enzyme Commission; GR,
gene–reaction; INSDC, International Nucleotide Sequence Database
Collaboration; PGDB, Pathway Genome DataBase; PkGDB, Prokaryotic
Genome DataBase.
The GenBank/EMBL/DDBJ accession number for the annotated
genome sequence of B. subtilis is AL009126.4.
Three supplementary tables are available with the online version of this
paper.
064691 G 2013 SGM
work pursued by a great many groups worldwide is
continuously imported into ongoing genome annotations
(Ouzounis & Karp, 2002), keeping them as accurate as
possible. During this process proper validation is essential
as percolation of annotations [in particular wrong annotations (Gilks et al., 2002)] is unavoidable, affecting the way
investigators build up their experiments in the future. This
need implies that annotation must be automatically
transferred from reliable sources. Model organisms play a
central role in this process. Bacillus subtilis is the only
model organism for the Firmicutes clade that has been
continuously annotated in depth. Its genome has been
resequenced and reannotated by Barbe et al. (2009), and it
is now time to deliver to the international community an
updated annotation that can be used both for specific work
on this organism and for making inferences not only for
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Printed in Great Britain
757
E. Belda and others
bacteria belonging to the Firmicutes, but also for many
other organisms.
At the time of the last update, the B. subtilis genome was
proposed to harbour 4244 predicted protein-coding
sequences (CDSs), 48 % with identified functions. Since
then, a considerable number of studies have allowed
investigators to complete, or at least improve, specific
annotations, including identification of new genes.
Sporulation was the most frequently recognized source of
biological interest for this organism. Recent work showed
that a specific biologically relevant feature of this organism
was that a significant portion of the genome sequence is
allocated to metabolic interactions with plants (Barbe et al.,
2009). It is now widely recognized that B. subtilis is an
epiphyte [it can even be found within plants (Huang et al.,
2011)]. More and more B. subtilis functions are found to be
typically relevant to plant-associated metabolism (this is
also in line with the traditional use of this organism in food
processing), including coping with surfaces by swarming
and making and remodelling biofilms (Chen et al., 2012;
Diethmaier et al., 2011; Ostrowski et al., 2011; Sekowska
et al., 2009). A major interest in the knowledge derived
from genome sequencing is the identification of active
metabolic pathways. This is particularly important for
modelling the physiology of the cell and for industrial
applications. Beside plant-related carbohydrate metabolism, a significant number of new enzyme activities were
discovered in B. subtilis, both in vivo and in vitro, and it is
now time to construct a comprehensive metabolic chart of
the organism. The present work collected the annotations
from publications that appeared after the 2009 sequence
release, as well as annotations of novel genomic objects
such as small untranslated RNAs, riboswitches and small
CDSs, with emphasis on metabolic discoveries that were
experimentally validated. The annotated sequence is
available to investigators as release AL009126.4 of the
corresponding entry at the International Nucleotide
Sequence Database Collaboration repositories (INSDC:
DDBJ/EMBL-EBI/GenBank) (Karsch-Mizrachi et al.,
2012). Additionally, the annotated genome sequence
together with an updated reconstruction of the metabolic
network of B. subtilis 168 are also available in the
MicroScope platform, which is based on the Prokaryotic
Genome DataBase (PkGDB) (https://www.genoscope.cns.
fr/agc/microscope/home/index.php) (Vallenet et al., 2009).
METHODS
General annotation procedures. As described by Barbe et al.
(2009) gene prediction was performed using the AMIGene software
(Bocs et al., 2003), using the gene models built with the previously
published version of B. subtilis annotations. The CDSs were assigned a
locus_tag similar to that of the previous annotations, i.e. ‘BSUxxxxx’.
In the case of additional gene predictions, we used a label
corresponding to the position of the gene in the genome sequence.
Sequence data for comparative analyses were obtained from the
National Center for Biotechnology Information database (RefSeq
section, http://www.ncbi.nlm.nih.gov/RefSeq). Putative orthologues
758
and synteny groups (i.e. displaying conservation of the chromosomal
co-localization between pairs of orthologous genes from different
genomes) were computed between the newly sequenced genome and
all the other complete genomes as described by Vallenet et al. (2006).
A further literature-screening procedure for previously characterized
CDSs involved searching for its name (and the name of synonyms) in
the PubMed and PubMed Central data libraries, as well as via Google
Web. Relevant PubMed identifiers were included in the genomic
object record when this information could be used to validate the
status of the annotation, with a particular emphasis on experimental
validations.
Newly annotated non protein-coding RNA genomic objects were
given a gene name (when possible) and labelled with the locus_tag
‘BSUmisc_RNA_xxx’, (xxx is a digital number indicating the start
position of the RNA). For example fsrA labels a regulatory RNA
controlling iron-dependent metabolism (Smaldone et al., 2012) and
its locus_tag is BSUmisc_RNA1483460D.
Automatic genome-scale metabolic network reconstruction.
The metabolic network of B. subtilis was generated by combining the
automatic and manually curated annotations extracted from PkGDB
using Pathway Tools (version 16.0), the BioCyc pathways reconstruction software (Latendresse et al., 2012). Starting with reference
metabolic pathways stored in the MetaCyc database (Caspi et al.,
2012), together with the set of genome annotations, the Pathway
Tools built up a specific Pathway Genome DataBase (PGDB) in a
two-step process: the Reactome projection and subsequently the
Pathway projection, as described by Karp et al. (2011). Reactome is an
open-source, open access, manually curated and peer-reviewed
pathway database. Pathway annotations are authored by expert
biologists in collaboration with Reactome editorial staff and crossreferenced to many bioinformatics databases. Metabolic pathways in
MetaCyc are defined as linked sets of reaction steps that participate in
a single biological process in a single organism, considered as active in
a sequence without lag or delay (Green & Karp, 2006). The Reactome
projection is performed by inferring associations between genes and
metabolic reactions extracted from MetaCyc. To this purpose, the
Enzyme Commission (EC) numbers linked to gene names are first
examined, followed by the gene product names and the gene ontology
terms found in the gene records of the genome. The corresponding
reactions are deemed ‘matched reactions’. However, this procedure
may lead to erroneous results because some records may be wrongly
formatted or may be linked to unspecific EC numbers or to
inaccurate or unclear textual annotations. This leads the Pathway
Tools either to overpredict reactions or to miss relevant enzymatic
reactions. We thus enhanced the matching procedure by directly
using the regular MetaCyc reactions’s frame identifier as the
functional annotation. PathoLogic is a component of the Pathway
Tools suite of MetaCyc that creates a new pathway database
containing the predicted metabolic pathways of an organism. Our
export procedure from PkGDB directly linked the genes to manually
validated MetaCyc reaction identifiers in the PathoLogic input
format. Alternatively, if no MetaCyc reaction has been validated
and associated to a particular gene, the procedure exports the partial
EC number and/or the gene product common name.
In the second step, the Pathway Tools software was used to infer
complete metabolic pathways through a pathway projection process
(Karp et al., 2011) that takes as the input the initial set of catalysed
reactions and the set of reference pathways available in MetaCyc. This
required the development of an in-house metabolic database having a
data structure identical to that of MetaCyc: this database was named
MicroRefCyc. It behaves as a repository for existing pathways that are
absent from MetaCyc. MicroRefCyc pathways are submitted on a
regular basis to both the external Microme repository (http://www.
microme.eu/) and the MetaCyc pathway resource (http://metacyc.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
org/). Pathways extracted from MetaCyc and MicroRefCyc were
merged together and used as the set of reference pathways to initiate
the PathoLogic pathway projection process. Finally, some reactions
lacking an associated gene (orphan reactions) were retrieved on the
basis of their presence in an inferred pathway. The reconstructed
metabolic genome scale network is included in the MicroCyc
repository available at http://www.genoscope.cns.fr/agc/microcyc.
Metabolic network curation process
The process of metabolic data curation is based on the specifications
of the MicroScope platform (Vallenet et al., 2009). It can be divided
into three main parts: first, the manual curation of individual gene–
reaction (GR) associations; second, the curation of automatically
generated metabolic pathway projections and the definition of new
metabolic pathways and pathway variants; third, the projection of GR
associations in two additional public resources for chemical
compounds [Chemical Entities of Biological Interest ontology
(ChEBI) (de Matos et al., 2012)] and reactions (RHEA) (Alcántara
et al., 2012). A schematic representation of the curation strategy is
provided in Fig. 1.
Curation of individual GR associations. As a starting point, the
annotation of the B. subtilis 168 genome stored in the MicroScope
platform at the end of December 2011 [it was continuously annotated
since the previous release of the B. subtilis genome sequence at the
INSDC (Barbe et al., 2009)] was used as the input data of the
PathoLogic algorithm, and a first set of MetaCyc reactions that can be
catalysed by the enzymes encoded in the genome was automatically
inferred (Karp et al., 2011). This step produced a set of automatically
retrieved GR associations that was subsequently divided into two
main subsets. This implied creating groups of similar genes, which we
identified as similar using BLASTP best hits of the corresponding
proteins. The best hit of a particular protein to all the proteins coded
by a target genome is the protein in that proteome that represents a
best BLASTP match. With complete genomes the procedure can be
symmetrical. The match is bidirectional when the two proteins are
best hits of each other. Accordingly, a bidirectional best hit (BBH)
indicates a very strong similarity between two proteins. It is used as in
silico evidence for orthology (Overbeek et al., 1999; Tatusov et al.,
2000). The first subset contained B. subtilis CDSs sharing BBH
relationships with E. coli K-12 counterparts, having the same gene
name or the same complete EC number annotation, which supports a
common functional role. For these CDSs, GR associations were
automatically transferred from the corresponding E. coli K-12 BBH
coming from curated EcoCyc data (Keseler et al., 2011). The second
subset comprised the remaining B. subtilis 168 CDSs having predicted
GR associations (with and without BBH relationships with E. coli K12); this set was manually curated to avoid overassignments of
biochemical reactions associated to CDSs with generic functional
annotations, and to reconcile B. subtilis 168 predictions with E. coli K12 and International Union of Biochemistry and Molecular Biologycompliant annotations of enzymatic activities.
Manual curation of GR associations was also extended to B. subtilis
168 CDSs annotated as enzymes and putative enzymes, and for which
no MetaCyc reaction had been predicted by PathoLogic. The
MicroScope platform gathers results from several computational
methods that can assist biologists in the curation process: this
includes BLASTP similarity profiles of each CDS against the UniProt
resource (Dimmer et al., 2012), PRIAM (an automatic profiling
sofware for automatic annotation of metabolism) profiles for EC
number predictions (Claudel-Renard et al., 2003), protein domain
profiling based on InterPro domains (Hunter et al., 2012) and protein
family assignments based on clusters of orthologues (Tatusov et al.,
2003), Fellowship for Interpretation of Genomes families (Meyer et al.,
2009) and high-quality automated and manual annotation of protein
family repositories (Lima et al., 2009). The outcome of all these
http://mic.sgmjournals.org
methods was further improved by extensive manual literature
searches to update B. subtilis 168 gene annotations and to broaden
our previous knowledge of biochemical reactions that can be
catalysed by enzymes encoded in the genome.
BsubCyc integration. Finally, GR associations coming from the
BsubCyc database (http://bsubcyc.org/) were also integrated into
MicroScope. BsubCyc is a Tier 2 BioCyc PGDB of B. subtilis 168 based
on the reannotation of 2009 (Barbe et al., 2009) that contains curated
data of B. subtilis 168 metabolism absent in MetaCyc repository (GR
associations and metabolic pathways) together with automatic
predictions based on the output of the PathoLogic algorithm
(Latendresse et al., 2012). Both B. subtilis 168 PGDBs (BsubCyc and
MicroScope) were compared at the level of individual GR associations, and BsubCyc-specific associations supported by experimental
evidence (not coming from automatic PathoLogic predictions) were
integrated in MicroScope.
Curation of automatically generated pathways projection. Manual
curation was also extended to the set of automatically projected
metabolic pathways generated by the PathoLogic algorithm
(Latendresse et al., 2012), based on the updated genome annotation
that includes automatic and manually validated GR associations (see
above). MetaCyc V16.0 was used as the reference repository for the
pathway projection procedure. This frequently produced incomplete
pathways, i.e. pathways with holes (enzymatic reactions not linked to
any gene). This situation may come from missing enzymatic activities
in annotations, presence of alternative pathway variants different from
those defined in MetaCyc, or non-functionality of the projected
MetaCyc pathway in the organism of interest. This problem prompted
us to curate manually the pathways that were incomplete, using
MicroScope. Several pathway-hole filling strategies have been
integrated in this platform, including CanOE (Smith et al., 2012) and
PhyloProfile (Engelen et al., 2012). CanOE is software that takes
advantage of the genomic and metabolic context, the similarity profiles
of the B. subtilis 168 proteome against UniProt entries validated by
experimental evidence of the missing enzymatic activity, to identify
missing activities in comparison with other genes of the pathway.
PhyloProfile analyses the co-evolution between phylogenetic profiles of
candidate genes. Finally, extensive manual literature searches were
performed to find experimental evidence of the missing enzymatic
activities in B. subtilis 168. When a candidate gene was found, its
functional annotation was updated accordingly (see Fig. 2 for an
example of pathway hole filling).
In addition, when the curation process allowed identification of a
novel metabolic pathway or a pathway variant not currently present
in the MetaCyc resource, the Pathway Tools were used to design novel
pathways and collect them in the MicroRefCyc database. In a similar
manner, BsubCyc-specific pathway entities absent in MetaCyc
repository and supported by experimental evidence were also
integrated into the MicroRefCyc database.
Curation of chemical compounds and reactions in ChEBI and
RHEA resources. The process of manual curation of GR associations
and pathway projections allowed us to identify several B. subtilis
enzyme activities that were not currently present in the MetaCyc
reaction repository. In order to deal with these situations, the
MicroScope platform allows the user to validate biochemical reactions
using two different reaction resources, mainly MetaCyc (Caspi et al.,
2012) and RHEA (Alcántara et al., 2012). RHEA is a European
resource of expert-curated biochemical reactions that are defined
based on chemical compounds from the ChEBI, and that are
stoichiometrically balanced for mass and charge at pH 7.3 (Alcántara
et al., 2012; de Matos et al., 2012). All small molecules of the MetaCyc
reactions associated to B. subtilis 168 validated genes were automatically mapped into RHEA, using the MetaCyc-RHEA cross-references
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
759
E. Belda and others
B. subtilis 168
PKGDB vN
BsubCyc PKGDB
Reactome inference (PathoLogic)
BsubCyc-specific
gene-Rxn
associations
Gene-Rxn associations
identical for BBH in
E. coli K12
Gene-Rxn associations
different for BBH in
E. coli K12
Genes no predicted
Rxn association;
enzyme-putative
enzyme
Gene-Rxn associations
with no BBH in
E. coli K12
Automatic
validation
BsubCyc-specific
metabolic
pathways
Manual
curation
B. subtilis 168
PKGDB vN+1
Pathway projection (PathoLogic)
Incomplete
MetaCyc
pathways
New metabolic
pathways/pathway
variants
Complete
MetaCyc
pathways
CanOE module
Similarity module
“Non-”/“functional”
Phylogenetic module
Pathway-hole
filling
Candidate
gene
annotation
update
Pathway
validation
“Variant-needed”
MicroRefCyc
metabolic
pathways
Literature searches
“False-positive”
HAMAP family searches
BRENDA annotations
B. subtilis 168
PKGDB vN+2
Reaction projection on RHEA
database
Reactions absent
in RHEA
Reactions present
in RHEA
Automatic mapping
RHEA–MetaCyc cross-reference
B. subtilis 168
PKGDB vN+3
New RHEA
reactions
New ChEBI
compounds
Fig. 1. Schematic representation of the curation workflow of B. subtilis 168 metabolic data. The curation process can be
divided into three main parts. First, the curation of individual GR associations. Second, the curation of MetaCyc pathway
projections and the creation of new metabolic pathways and pathway variants identified during the curation process and absent
in MetaCyc pathway repository. Third, projection of MetaCyc reactions and B. subtilis 168 specific reactions on ChEBI and
RHEA resources for compounds and reactions. Metabolic data for GR associations and metabolic pathways from BsubCyc
database (http://www.bsubcyc.org/) with supporting experimental evidence were also integrated. All annotation updates, GR
associations and metabolic pathway projections and reconstructions are stored in the specific B. subtilis entry of PKGDB in the
MicroScope database environment (Vallenet et al., 2009).
760
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
Fig. 2. Pathway hole-filling strategies for the 1,4-dihydroxy-2-naphthoate biosynthesis pathway. (a) Initial projections of
MetaCyc pathway (MetaCyc PWY-5837) in B. subtilis 168 based on MicroScope annotations. Two pathway holes are
observed at the level of EC 4.2.99.20 and EC 3.1.2.28 reactions. (b) Schematic representation of a metabolon predicted with
CANOE (Smith et al., 2012). Pathway genes are co-localized in the chromosome of B. subtilis 168, where a candidate gene,
BSU30810, is retrieved as possibly associated to the orphan enzymatic activity (EC 4.2.99.20). (c) Best co-evolved genes with
BSU30810 based on correlation scores of the corresponding phylogenetic profiles calculated over all microbial genomes
stored at the MicroScope platform [for details of the algorithm, see Engelen et al. (2012)]. The best co-evolved genes with the
candidate gene BSU30810 are two neighbour genes involved in other steps of the pathway (BSU30800 and BSU30820). (d)
BLASTP profiles of BSU30810 against protein entries of SwissProt. The best hit corresponds to the BSU30810 protein entry in
SwissProt, where it appears annotated as a putative esterase, with an incomplete EC number association. The next best hits
show significant similarities with the orphan activity EC 4.2.99.20 in initial pathway projections, confirming the initial predictions
of CANOE.
stored in the RHEA database. MetaCyc reactions involving small
molecules that do not have RHEA cross-references were also manually
curated: the corresponding MetaCyc cross-references were added if
the reactions were already present in RHEA, or the reactions were
created de-novo if they were absent. For those cases of missing
compounds in ChEBI with correct 2D structure at pH 7.3, the
corresponding compounds were created de novo in ChEBI. The
Marvin suite of tools from ChemAxon (http://www.chemaxon.com)
was used to finalize the 2D structure of these new compounds at
pH 7.3 so that they are now available for reaction creation into
RHEA. The same procedure was followed for specific biochemical
http://mic.sgmjournals.org
reactions characterized during the manual curation of GR associations that are absent in the MetaCyc reaction repository.
The resulting curated B. subtilis metabolic network is available at
the BacilluScope project website (https://www.genoscope.cns.fr/agc/
microscope/home/?index.php) and is included in the MicroCyc
repository (http://www.genoscope.cns.fr/agc/microcyc).
Experimental validation of annotations. In some cases, when
there was ambiguity in the function ascribed to a particular gene we
performed experiments designed to validate the proposed annotation.
The strain used was strain 168, resequenced in 2009 and deposited at
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
761
E. Belda and others
the DSMZ under accession number DSM 23778 (Anagnostopoulos &
Spizizen, 1961; Barbe et al., 2009). Experiments were performed by
growing relevant B. subtilis strain 168 mutants in liquid or on solid
media, as described previously (Barbe et al., 2009).
Cells were grown either in Luria–Bertani (LB) medium (Bertani,
1951) or in ED minimal medium: K2HPO4, 8 mM; KH2PO4, 4.4 mM;
glucose, 27 mM; Na3-citrate, 0.3 mM; L-glutamine, 15 mM; Ltryptophan, 0.244 mM; ferric citrate, 33.5 mM; MgSO4, 2 mM;
MgCl2, 0.61 mM; CaCl2, 49.5 mM; FeCl3, 49.9 mM; MnCl2,
5.05 mM; ZnCl2, 12.4 mM; CuCl2, 2.52 mM; CoCl2, 2.5 mM;
Na2MoO4, 2.48 mM, with the relevant carbon, nitrogen or sulfur
source when needed, omitting the counterpart (e.g. MgCl2 replaced
MgSO4 when cells were assayed for a sulfur source).
LB and ED plates were prepared by addition of 17 g Bacto agar or
Agar Noble (Difco) l21, respectively, to the medium. Bacteria were
grown at 37 uC. The optical density of bacterial cultures was
measured at 600 nm. For assaying growth on sulfur compounds on
plates, either the MgSO4-containing medium or the sulfur-free basal
medium was used (MgSO4 was replaced by MgCl2 as described
above). In the latter case, 10 ml of the sulfur source under
investigation was applied onto paper discs (generally 100 mM stock
solution) deposited on the plate, after the bacteria had been uniformly
spread on the surface of the plate, and growth was measured around
the disc. A similar procedure was used for assaying nitrogen sources,
omitting glutamine from the medium and placing the putative
nitrogen source on the disc.
RESULTS AND DISCUSSION
General features of reannotation
The B. subtilis sequence now comprises 4458 genomic
objects (4422 in the preceding release); 1217 genes have
been reannotated and updated references have been linked
to genes (3644 PubMed references in the current
annotation in comparison with 3376 references in the
preceding release). A significant effort has been made in the
curation of the associations between genes encoding
enzymes and the biochemical reactions catalysed by these
enzymes in order to provide a curated and comprehensive
reconstruction of the global metabolic map of B. subtilis. In
the current release, 1083 CDSs have been associated to
1097 chemical reactions (a total of 1751 GR associations).
In these associations, 384 functions were validated
automatically after association to the corresponding E. coli
K-12 reactions (EcoCyc data) by an automatic transfer:
BBHs between CDSs sharing the same gene name or
complete EC number annotations, supporting a common
functional role. Furthermore, 595 gene functions have been
manually validated in order to avoid missing overassignments of the PathoLogic algorithm, validating only
biochemical reactions for which there is experimental
supporting evidence (Fig. 3a).
In terms of chemical reactions, of the 1097 reactions
associated to B. subtilis CDSs, the majority (685) are smallmolecule reactions common to MetaCyc and RHEA
resources – 367 reactions are MetaCyc specific (absent in
RHEA), corresponding mainly to reactions involving
macromolecules (DNA, proteins) that are not managed
762
by RHEA [which is a resource specific for chemical
reactions involving small chemical entities (Alcántara
et al., 2012)]. Finally, we identified 45 reactions that are
absent in MetaCyc and that appear associated to the novel
metabolic pathways and pathway variants of B. subtilis that
were characterized during the curation process.
Manual curation has been also extended to the set of
automatically predicted metabolic pathways using the
PathoLogic algorithm, based on the MetaCyc V16.0
pathway repository. This PGDB contains a total number
of 283 MetaCyc pathways. Among those, 164 were
complete after automatic PathoLogic projections (all
pathway steps with associated CDSs), whereas 119
remained incomplete, having one or more pathway holes
at different steps of the pathway. These incomplete
pathways were manually curated in order to find candidate
genes for missing enzymatic activities and to assign a
functional status according to the experimental evidence
available. The global results of this curation process are
represented in Fig. 3(b). Twenty-four pathway holes were
resolved after manual curation, resulting in a total number
of 22 initially incomplete pathways with ‘validated’ status.
Among those, eight are now complete, whereas 14 remain
incomplete. Pathways that are functional according to
experimental evidence but where some reaction steps do
not yet have an identified gene in B. subtilis remain
incomplete in terms of gene identification but are
nevertheless deemed ‘validated’. This status could be used
to define new targets for experimental studies meant to
identify the corresponding catalyst (see Table S1, available
in the online version of this paper). The full list of the 283
MetaCyc pathways with the corresponding functional
status according to MicroScope pathway curation tool is
provided in Table S2.
Finally, manual curation of GR associations and automatic
pathway projections resulted in the characterization of 27
new metabolic pathways or pathway variants of B. subtilis
that are not represented in MetaCyc V16.0 pathway
repository (Table S3). These metabolic pathways have been
instantiated within the internal PGDB MicroRefCyc by
using the Pathway tools software package. The complete
metabolic network of B. subtilis 168 comprising manually
curated and automatically validated GR associations
together with full set of MetaCyc V16.0 and MicroRefCyc
metabolic pathways are freely available in a specific
MicroCyc PGDB in MicroScope environment at the
following URL http://microcyc.genoscope.cns.fr/BACSU7/
organism-summary.
Beside emphasis on metabolic pathways, discussed below,
several new features have now been included in the genome
annotation.
Specific genomic objects
Toxin/antitoxin genes. Association of gene products that
combine a deleterious activity together with a control
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
(a)
(b)
Gene remnant 0.21
Function proposed based on presence
of conserved amino acid motif, structural
feature or limited homology
Unknown
7.85
Function of strongly
homologous gene
Deleted
Function of homologous gene
experimentally demonstrated
in another organism
39
Validated;
incomplete
25.13
Function experimentally demonstrated
0.21
in the studied species
14
Validated;
complete
Function experimentally
demonstrated in the studied strain
63.14
10
47
Variant
needed
3.46
0
11
20
30
40
50
60
70
8
Predicted;
complete
164
0
20 40 60 80 100 120 140 160 180
Genes with validated GR associations (%)
Fig. 3. (a) Distribution of GR associations, in agreement with supporting evidence. The x-axis represents the fraction of B.
subtilis genes with validated GR associations (955 CDSs). The y-axis corresponds to the different categories of the ‘class’
qualifier of MicroScope gene annotations. (b) Global results of the pathway curation process. Manual curation has been
restricted to incomplete pathways in initial PathoLogic projections based on the MetaCyc V16.0 repository. Validated functional
pathways after manual curation process; variant needed, pathways where MetaCyc representation does not correspond to the
real metabolic profile present in B. subtilis; unknown, pathways where there is not enough evidence to declare them as
functional; deleted, false-positive predictions of PathoLogic algorithm; predicted; automatic predictions by PathoLogic
algorithm, restricted to pathways where all the reaction steps have a corresponding protein-coding gene associated (complete
pathways).
inhibitor are widespread in all domains of life. A case in
point is the association between restriction endonucleases
and associated anti-restriction methylases. This type of
association has been extended and is nowadays generally
identified under the ‘toxin–antitoxin’ generic name (Hayes
& Van Melderen, 2011). Many physico-chemical processes
lie under this banner: the toxin can be a protein destabilizing membranes, a protease, a restriction endonuclease
or, most often, an RNase that may have a variety of targets,
sometimes with broad specificity, sometimes with
extremely narrow sites of action. Most of these systems,
restriction endonucleases aside, belong to two classes.
Type I toxins are often short proteins, disrupting
membranes. The antitoxin of type I systems is a small
RNA which base pairs with the toxin mRNA to prevent
protein synthesis (Fozo et al., 2010). The TxpA/RatA with
its RNA was already identified in the previous release. In
phage SPbeta the BsrG/RtaD (SR4) is expressed in a
thermosensitive fashion and controls cell lysis (Jahn et al.,
2012). We now included a collection of several related
peptides that are probably type I toxins: YoyG, YosA, YjcZ,
YczM, YksV, YuzJ and SscA, for which the putative
downstream RNA has not yet been identified. They all
contain variants of the peptide motif FALLVVLFILLIIVG.
The latter, SscA, displayed a poor germination phenotype
when disrupted (Kodama et al., 2011).
The antitoxin of the type II systems is a protein that binds
to and inhibits the toxin protein. Several such systems have
been documented in B. subtilis 168: NdoA/NdoAI modulates the effect of lethal stresses, via the action of a
http://mic.sgmjournals.org
sequence-specific RNase (Wu et al., 2011); SpoIISA/
SpoIISB, inactivation of SpoIISA has no effect on
sporulation but it fully restores sporulation of a spoIISB
null mutant, indicating that SpoIISB is required only to
counteract the negative effect of SpoIISA on sporulation
(Florek et al., 2011); YxiD (now RtbD) was annotated as a
putative transposase, experiments have recently demonstrated that it is in fact the antitoxin of toxic RNase of the
RtbD(YxiD)/RtbE(YxiE) system. The same is true for the
couples RtbG/RtbF and RtbI(YocI)/RtbJ(YocJ), with SsrSA
small RNA in between (Beckmann et al., 2012) RtbL/RtbM
and RtbN(YwqJ)/RtbO(YwqK) (Holberger et al., 2012).
Small untranslated RNAs and riboswitches. Several small
untranslated RNAs have been identified (Irnov et al., 2010),
many without known function, while some are involved in
the control of iron homeostasis (Gaballa et al., 2008). SrrA
(SR1) controls expression of a transcriptional activator of
arginine catabolism, AhrC (Heidrich et al., 2006), while it
also encodes RgpA, which regulates expression of gapA
(Gimpel et al., 2010), and FsrA, which controls ironresponsive genes (sdhCAB, citB, yvfW, leuCD) (Gaballa et al.,
2008). Apart from the previously identified S-boxes and Tboxes, several novel riboswitches have been identified
(Breaker, 2011).
Information transfer. Besides the previously identified
features of macromolecule modifications, the biosynthesis of
themodifiednucleotide2-methylthio-N6-threocarbamoyladenosine,
a modified residue at position 37, adjacent to the 39-end
of the anticodon of several tRNAs has been characterized
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
763
E. Belda and others
(MicrorefCyc, PWYIPS-51). This complex modification,
which is found in Firmicutes and extends that found in
gamma-proteobacteria, combines synthesis and transfer of the
threonylcarbamoyl- group to the N6 atom of the adenine
residue at position 37 of tRNA catalysed by a complex of four
proteins, TsaC (YwlC, RimN), TsaE (YdiB), TsaB (YdiC) and
TsaD (YdiE) tRNA threonylcarbamoyltransferase, and then the
transfer of methylthiol group to threonylcarbamoyladenosine
37 by the class 3 methylthiotransferase MtaB (YqeV, RimO)
(Anton et al., 2008, 2010; Arragain et al., 2010; Lauhon, 2012).
characterization prevented us from changing its status of ‘Y’
(unknown) protein.
Metabolic pathways
The present update of the B. subtilis 168 genome
annotation provides us with a fairly comprehensive view
of its metabolism. We highlight below some of the novel
features that have been introduced in the present updated
annotation of the genome sequence, with emphasis on
metabolism.
The Firmicute degradosome and nanoRNases. The
existence of a variety of small untranslated RNAs as well
as regulatory regions in translated RNAs has placed RNA
degradation in the limelight. In E. coli a structure named
the degradosome is organized around RNase E (Carpousis,
2007). There is no structural counterpart of this RNase in
B. subtilis, and a remarkable feature of this organism is that
the way it degrades RNA appears to be completely different
to that found in E. coli. In particular, the counterpart of the
degradosome is almost entirely different, being built up
around RNase J (A and B) (Deikus & Bechhofer, 2009;
Mäder et al., 2008; Mathy et al., 2010) and RNase Y
(Commichau et al., 2009; Lehnik-Habrink et al., 2011).
There is, however, convergent evolution, in that this
degradosome seems to be strongly connected to nucleotide
metabolism [via polynucleotide phosphorylase, cytidylate
kinase and a short version of the mRNA-presenting protein
RpfA (YpfD), used as ribosomal protein S1 in its long
version in E. coli] and to the core of glycolysis (Danchin,
2009). Another essential function of the degradosome is
that of nanoRNases, degrading oligonucleotides shorter
than 5 nt, Orn in E. coli and at least two counterparts
of different descent in B. subtilis, NrnA and NrnB (Fang
et al., 2009). Remarkably, NrnA displays 39,59-adenosine
bisphosphate (pAp) 39 phosphatase activity, coupling RNA
degradation to sulfur and lipid metabolism (pAp is the
by-product of sulfate reduction, as well as of 4-phosphopantetheine synthesis). This type of coupling between
small molecules and macromolecule degradation will have
to be taken into account in further developments of
metabolic pathway tools and repositories. A further
substantiation of this widespread category of pathways is
seen with a similar double function, found for NrnC in
alpha-proteobacteria (Liu et al., 2012), suggesting that this
is an important metabolic feature of RNA catabolism.
Another exonuclease of unknown specificity, YhaM, may
also participate in the activity of the degradosome. It belongs
to a cluster of co-evolving proteins comprising YmfF and
YmfG, two peptidyl hydrolases of unknown substrates,
perhaps associated to transport of peptidyl siderophores and
connected to YheA [related to metabolism of aromatics
(YwbD, AroC, TyrA, AroF)], proteins such as YqjA (a
putative membrane-bound protein) and YacP (possibly
involved in rRNA maturation), to a network with tRNA
synthetases CysS and GltX and with sigma factor SigH
(Engelen et al., 2012). Nevertheless, the lack of further
764
Carbon metabolism. The pathway of myo-inositol
degradation to the glycolytic intermediate dihydroxyacetone phosphate has been updated, including the activities
of IolD and IolB, catalysing the hydrolysis of the pathway
intermediate 3D-(3,5/4)-trihydroxycyclohexane-1,2-dione
to yield 5-deoxy-D-glucuronic acid and the subsequent
isomerization to 2-deoxy-5-keto-D-gluconic acid (Yoshida
et al., 2008). The activity of IolI, the inosose isomerase that
catalyses the isomerization of 2-keto-myo-inositol intermediate to 1-keto-D-chiro-inositol, which is subsequently
transformed to D-chiro-D-inositol by the inositol
dehydrogenase IolG, has also been included (Yoshida
et al., 2006). These activities have been experimentally
validated by the inactivation of IolE, leading to
overproduction of D-chiro-inositol, a possibly interesting
metabolite as a drug candidate as it improves the efficiency
with which the body uses insulin and also promotes
ovulation, being a possible candidate for the treatment of
type 2 diabetes and polycystic ovary syndrome (Yoshida
et al., 2006).
The fermentative metabolism of B. subtilis 168 has been
further documented based on the experimental evidence
reviewed by Nakano & Zuber (1998) and Nakano et al.
(1997). In the absence of external electron acceptors, B.
subtilis 168 still grows anaerobically by fermentation,
provided glucose and pyruvate are supplied, with ethanol,
lactate, acetate and 2,3-butanediol as major fermentative
products detected by nuclear magnetic resonance and
HPLC experiments on anaerobic cell cultures with glucose
and pyruvate as carbon sources and ammonia as a nitrogen
source (Nakano & Zuber, 1998; Nakano et al., 1997).
Production of lactate and 2,3-butanediol continues even in
the presence of the anaerobic electron acceptors nitrate and
nitrite, pointing out the importance of these fermentative
processes for NADH reoxidation even for anaerobic
respiratory growth (Nakano & Zuber, 1998).
The annotation of the gene cluster yngJIHGFE has also
been updated to reflect its functional role, encoding
enzymes responsible for the degradation of L-leucine to
acetyl-CoA. Experimental inactivation of each gene and
analysis of the corresponding cell cultures in medium with
13
C-labelled L-leucine confirm that the pathway is functional during sporulation in B. subtilis 168 and that
individual knockouts of this gene cluster block L-leucine
degradation (Hsiao et al., 2010). The pathway might also
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
Table 1. cont.
Table 1. Growth properties of Bacillus subtilis 168
Growth is measured on minimal medium plates spread with a liquid
culture of B. subtilis (see Methods) onto which a 6 mm paper disk is
deposited, containing 10 ml of a 100 mM solution of the assayed
compound. In the case of serine a complete inhibition zone around
the disk is then followed by a zone of growth. Growth was also
monitored in liquid medium supplemented with 10 mM (final
concentration) of the compound of interest (except for arginine,
where 5 mM was used as arginine acts as a chaotropic agent, leading
to B. subtilis lysis). +, Growth; +/–, weak growth; –, no growth; NA,
not applicable.
Compound
Carbon
Nitrogen
Alanine
Arginine*
Asparagine
Aspartic acid
Cysteine
Glutamic acid
Glutamine
Glycine
Histidine
Isoleucine
Leucine
Lysine
Methionine
Phenylalanine
Proline
Serine
Threonine
Tryptophan
Tyrosine
Valine
Beta-alanine
1,5-Anhydro-D-sorbitol
Glutarate
Phenylgalactoside
D-Rhamnose
S-Methyl-cysteine
Compound
Acetylmethionine
Benzenesulfonic acid
Cholesterol sulfate
Coenzyme M
Dithiodibutyric acid
Dithiodipropionic acid
Formyl-methionine
Glutathione
Isethionate
Methanesulfonate
Methanethiol
Orthanilic acid
Propane sulfonate
Sinigrine
S-Methylmethionine
S-Methylcysteine
S-Methyl-galactoside (TMG)
Sulfanilamide
+
+
+
+/–
–
+/–
+
–
–
+
+/–
–
+/–
+/–
+
–
–
–
–
+
–d
–d
–d
–d
–d
–d
Sulfur
+
–
–
+
–
–
+
+/–
+
+
–
–
+
–
–
+
–
–
+
+
+
+
–
+
+
+/–
+
+
+
–
–
–
+
+D
+
–
–
+
–d
http://mic.sgmjournals.org
NA
NA
NA
NA
+d
Compound
Sulfobenzoic acid
Sulfur (S0)
Taurine
Thialysine
Thiaproline
Thio-dipropionic acid
Thiophene
Toluenesulfonic acid
Sulfur
–
+
+/–
+/–
+
–
–
–
*5 mM in liquid conditions.
DInhibition of 1 cm followed by growth of 0.5 cm.
dNot assayed in liquid medium.
be used in isoleucine and valine degradation during
sporulation.
The substrate specificity of carbohydrate-degrading
enzymes has to be carefully established. Beta-galactosidase
GanA from B. subtilis is very different from LacZ of E. coli.
It is an arabinogalactan hydrolase (Shipkowski &
Brenchley, 2006), which is restricted in its substrates. For
example, in contrast with LacZ its presence does not allow
growth on phenyl-galactoside, showing also that its
paralogue RhgZ (a possible galacturonisase) does not
confer this ability either (Table 1). D-Rhamnose is a fairly
rare component of the surface of many pathogens. Because
B. subtilis has been shown to counteract many actions of
pathogens on plants (Choudhary & Johri, 2009), it was
interesting to see whether it could use this carbon source.
Contrary to the situation with L-rhamnose (involving
rhaA, rhaB, rhaU), B. subtilis does not grow on D-rhamnose
(Table 1).
Energy metabolism. Three different metabolic pathways
(MicrorefCyc: PWYIPS-66, 67 and 95) have been created,
reflecting the two main branches of aerobic respiration
existing in B. subtilis 168, the quinol oxidase branch and
cytochrome oxidase branch, according to the presence of
three different types of terminal oxidases (Winstedt & von
Wachenfeldt, 2000). In the quinol oxidase branch,
menaquinol is directly reoxidized to menaquinone using
molecular oxygen as an electron acceptor by two different
types of quinol oxidases: the cytochrome aa3 oxidase
complex encoded by the qoxABCD gene cluster
(MicrorefCyc PWYIPS-66), which belongs to the haem–
copper superfamily of respiratory oxidases, coupling
menaquinol reoxidation with proton translocation across
the membrane (Santana et al., 1992); and two different
cytochrome bd oxidases encoded by the cydABCD or ythAB
gene clusters (MicrorefCyc PWYIPS-67), which do not
pump protons (Winstedt et al., 1998). CydABCD has been
shown to be a cytochrome bb9 oxidase rather than a
cytochrome bd oxidase with haem D being substituted by
high spin haem B at the oxygen reactive site, i.e.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
765
E. Belda and others
cytochrome b(558)b(595)b9 (Azarkina et al., 1999). The
exact role of YthAB is unknown as neither deletion nor
overproduction alter the expression of the bb9 oxidase
(Azarkina et al., 1999). In the cytochrome oxidase branch
(MicrorefCyc PWYIPS-71), a cytochrome bc1 complex
encoded by genes qcrCAB catalyses the reoxidation of
menaquinol intermediates by electron transfer to a C-type
cytochrome, which is subsequently reoxidized by a
cytochrome caa3 complex encoded by the genes
ctaCDEF, reducing molecular oxygen to water (Yu et al.,
1995).
Anabolism of small molecules. A new biotin biosynthesis
pathway variant has been defined, including the specific
activity of BioA and BioI enzymes in B. subtilis 168. BioA is
lysine-8-amino-7-oxononanoate aminotransferase that
uses specifically L-lysine as an amino group donor instead
of S-adenosylmethionine, which is commonly used by
BioA homologues in other bacterial species (Van Arsdell
et al., 2005), whereas BioI encodes a cytochrome p450
protein that catalyses the two-step oxidative cleavage of
long-chain acyl-ACP and long-chain fatty acids to yield
pimeloyl-ACP and pimelate, respectively, which are the
starting substrates of the pathway (Green et al., 2001; Stok
& De Voss, 2000). This activity allows the connection of
fatty acid metabolism with biotin biosynthesis, accounting
for the unexplained auxotrophy for biotin observed in
currently available genome-scale metabolic models of B.
subtilis 168 (Henry et al., 2009).
The lipoate biosynthesis pathway of B. subtilis 168 also
proceeds in a slightly different way when compared with
that of E. coli K-12, with the presence of octanoyl[GcvH] : protein N-octanoyltransferase LipL, which catalyses the two-step transfer of octanoyl moiety from
octanoyl-GcvH complex to the E2 subunit of pyruvate
dehydrogenase (Christensen et al., 2011).
The pathway of branched-chain fatty acid biosynthesis has
also been updated. This includes the initial two-step
degradation of branched-chain amino acids (L-valine, Lisoleucine and L-leucine) to the corresponding 2-ketoacylCoA intermediates (isobutyryl-CoA, 2-methylbutanoyl-CoA
and isovaleryl-CoA, respectively) by the combined action of
the branched-chain amino acid dehydrogenase Bcd and the
branched-chain keto-acid dehydrogenase enzyme complex
(BkdAA+BkdAB+BkdBB). This degradation is coupled
with the successive elongation cycles by using malonyl-ACP
as an extender substrate carried out the universal dissociated
type II fatty acid synthase system encoded by fab genes (Choi
et al., 2000; Debarbouille et al., 1999; Kaneda, 1991; Oku &
Kaneda, 1988). These elongation cycles are equivalent to
those involved in the classical pathway of straight-chain fatty
acid biosynthesis present in E. coli K-12 (with acetyl-CoA as
a primer substrate).
In the same way, the metabolism of methionine has been
updated with a novel pathway representation (MicrorefCyc
766
RXNIPS-71). This covers the reverse transulfuration
pathway from methionine to cysteine existing in B. subtilis
168 that allows the bacterium to grow with methionine as a
sulphur source, and that includes the specific cystathionine
b-synthase MccA that uses O-acetyl-serine instead of serine
as the substrate to produce cystathionine from homocysteine, and MccB, which encodes an enzyme with
cystathionine c-lyase, homocysteine c-lyase and cysteine
desulfhydrase activities (Hullo et al., 2007).
It is known that B. subtilis does not synthesize glutathione,
and it took a long time to identify the molecule that plays
the role of this reduced-sulfur-buffering molecule. It has
now been established that B. subtilis synthesizes bacillithiol, the alpha-anomeric glycoside of L-cysteinyl-Dglucosamine with L-malic acid. The whole pathway
(encoded by the genes bshBAbshA, bshBB and bshC) is
now present and documented in the genome sequence
annotations (Gaballa et al., 2010; Newton et al., 2009,
2011; Upton et al., 2012).
It has been recently reported that B. subtilis 168 is able to
synthesize in vivo an archaeal-type ether lipid by the action
of glycerol-1-phosphate (G1P) dehydrogenase AraM,
which catalyses the NADH-dependent reduction of dihydroxyacetone phosphate to G1P, followed by the condensation with the heptaprenyl moiety of heptaprenyl
diphosphate catalysed by heptaprenylglyceryl-phosphate
synthase PcrB (Guldan et al., 2008, 2011). This represents
the first experimental evidence of the presence of an
archaeal-type ether lipid derived from G1P in bacteria, a
remarkable feature from an evolutionary point of view,
since one of the main differences between Bacteria and
Archaea lies in the chemical composition of their
membrane phospholipids as well as the chirality of the
glycerol–phosphate head. Bacteria and Eukarya have esterlinked fatty acid phospholipids based on glycerol-3phosphate (G3P), whereas Archaea have ether-linked
isoprenoid phospholipids based on G1P (Lombard et al.,
2012). Phylogenomic approaches have revealed the presence of G1P and G3P dehydrogenases in the three domains
of life, strongly supporting the view of the ubiquitous
presence of phospholipid membranes in the common
ancestor population of Bacteria and Archaea despite their
different composition (Peretó et al., 2004). Various
analyses of metabolic pathways have long suggested a
significant proximity between Archaea and Firmicutes
(Bright et al., 1993; Jensen et al., 1988; Ruepp et al.,
1995; Sekowska et al., 2000) and, although horizontal gene
transfer events cannot be excluded, the presence of this
archaeal-type ether lipid biosynthesis pathway in B. subtilis
168 substantiates the idea of a very-early presence of
phospholipids of various types in the ancestral populations
of primitive cells.
Secondary metabolism. A novel aromatic polyketide
biosynthesis pathway has been entered (MicrorefCyc
PWYIPS-48) that includes the polyketide synthase BcsA,
which catalyses the biosynthesis of tri- and tetraketide
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
pyrones by the condensation of long-chain fatty acyl-CoA
thioesters (primer substrates) with malonyl-CoA (extender
substrates), and methyltransferase BcsB, which methylates
akylpyrones produced by BcsA-yielding akylpyrone methyl
ethers. YpbQ represents the first enzyme found to
methylate akylpyrones (Nakano et al., 2009).
Finally, annotation of the bacABHF genes has been updated
in order to reflect their roles in tetrahydrotyrosine biosynthesis, a non-proteogenic amino acid that is synthesized
by the products of the genes bacABHF, from which
anticapsin, the key residue essential for the antibiotic
activity of bacylisin, could be derived (Shomura et al.,
2012).
Other idiosyncratic features. B. subtilis has a complete
methionine salvage pathway. While the pathway accounts
for salvage of the thiomethyl group under oxidative
conditions (with a dioxygenase, MtnD, used at the
penultimate step of the pathway), the last transamination
step unexpectedly uses glutamine as the alpha-amino
donor group (Sekowska et al., 2004). This reaction is
surprising because it generates alpha-ketoglutaramate, a
reactive molecule that cyclizes and has toxic properties.
This entails the existence of a further metabolic step that
deamidates this toxic compound. MtnU, previously
annotated as a nitrilase, is in fact highly similar to known
omega amidases. As it has all the characteristic features of
these enzymes, we can be confident that this is indeed its
activity in B. subtilis. This step introduces an irreversible
reaction that prevents back-transamination of methionine.
This is supported by the observation that, while being a
sulfur source, methionine cannot be used as a carbon
source (Table 1).
The pathway of purine nucleotide degradation has been
updated by including the specific activity of PucG, which
encodes a specific pyridoxal-phosphate-dependent (S)ureidoglycine : glyoxylate aminotransferase that catalyses
the amino group transfer from the purine degradation
intermediate (S)-ureidoglycine to glyoxylate, yielding the
amino acid glycine and oxalurate (Ramazzina et al., 2010).
In this context, (S)-ureidoglycine can also deaminate
possibly spontaneously to glyoxylate by releasing urea
and subsequently ammonia, representing a novel mechanism by which the same metabolite [(S)-ureidoglycine] is
able to act as an amino group donor and acceptor under
conditions where the amino group acceptor (glyoxylate) is
absent (Ramazzina et al., 2010).
The presence of a specific lysine aminomutase (KamA)
suggested that this enzyme might be an intermediate in
lysine degradation. However, we found that lysine could
not be used as a carbon source under vegetative growth
(Table 1). In B. subtilis, yodP and kamA belong to a
transcriptional unit whose expression is upregulated
during sporulation, and, in fact, N(e)-acetyl-b-lysine has
not been detected in vegetatively growing or osmotically
stressed B. subtilis cells (Müller et al., 2011), pointing out a
http://mic.sgmjournals.org
possible role of N(e)-acetyl-b-lysine under sporulation.
Glutarate, an intermediate in several lysine degradation
pathways (MetaCyc pathway III, IV, V and IX), cannot be
used either (Table 1). The annotation of the kamA and
kamB (yodP) genes has been updated, reflecting their roles
encoding an AblA-related lysine-2,3-aminomutase and an
AblB-related acetyltransferase, respectively, involved in the
biosynthesis of the archaeal-type osmolite N(e)-acetyl-blysine (Müller et al., 2011). N(e)-acetyl-b-lysine is characteristic of methanogenic Archaea, allowing cellular
adaptation to high-osmotic conditions, and also having
other stabilizing properties for protein, membranes and
even entire cells (Pflüger et al., 2003). It is a compatible
solute that may protect bacteria against some stressful
conditions such as low temperature or extreme pH.
Alternatively, a putative lysine degradation pathway might
be triggered only under specific developmental conditions.
We note that the yodT upstream promoter of the region
encompassing kamBA is indeed under the control of sigma
E, a mother cell-specific promoter during sporulation. It
may be that lysine degradation is triggered only in the
mother cell when it undergoes lysis.
Conclusion
With the data explosion that parallels the development of
next-generation sequencing techniques it is no longer
possible to annotate most genomes in-depth. This makes
the role of model bacteria as essential partners in defining
the reference knowledge that is subsequently automatically
propagated in genome annotations all the more important.
B. subtilis 168 is the reference organism for the Firmicutes
and it is particularly useful that its genome sequence
annotation is updated in a regular fashion. More than 5000
new references connected to this organism have been
included in PubMed since the last annotation of the
genome, many of them corresponding to the identification
of gene functions. The present work upgraded the overall
identification of gene functions, while focusing on creating
a consistent picture of the intermediary metabolism of the
organism, a feature that will be essential for work on
systems and synthetic biology. It will be pursued over the
next few years with the aim of having a fairly complete
picture of the gene set and organization of this model
organism.
ACKNOWLEDGEMENTS
This work was supported by European Union’s 7th Framework
Programme Microme FP7-KBBE-2007-3-2-08-222886.
REFERENCES
Alcántara, R., Axelsen, K. B., Morgat, A., Belda, E., Coudert, E.,
Bridge, A., Cao, H., de Matos, P., Ennis, M. & other authors (2012).
Rhea – a manually curated resource of biochemical reactions. Nucleic
Acids Res 40 (Database issue), D754–D760.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
767
E. Belda and others
Anagnostopoulos, C. & Spizizen, J. (1961). Requirements for
Claudel-Renard, C., Chevalet, C., Faraut, T. & Kahn, D. (2003).
transformation in Bacillus subtilis. J Bacteriol 81, 741–746.
Enzyme-specific profiles for genome annotation: PRIAM. Nucleic
Acids Res 31, 6633–6639.
Anton, B. P., Saleh, L., Benner, J. S., Raleigh, E. A., Kasif, S. &
Roberts, R. J. (2008). RimO, a MiaB-like enzyme, methylthiolates the
universally conserved Asp88 residue of ribosomal protein S12 in
Escherichia coli. Proc Natl Acad Sci U S A 105, 1826–1831.
Anton, B. P., Russell, S. P., Vertrees, J., Kasif, S., Raleigh, E. A.,
Limbach, P. A. & Roberts, R. J. (2010). Functional characterization of
Commichau, F. M., Rothe, F. M., Herzberg, C., Wagner, E., Hellwig,
D., Lehnik-Habrink, M., Hammer, E., Völker, U. & Stülke, J. (2009).
Novel activities of glycolytic enzymes in Bacillus subtilis: interactions
with essential proteins involved in mRNA processing. Mol Cell
Proteomics 8, 1350–1360.
the YmcB and YqeV tRNA methylthiotransferases of Bacillus subtilis.
Nucleic Acids Res 38, 6195–6205.
Danchin, A. (2009). A phylogenetic view of bacterial ribonucleases.
Arragain, S., Handelman, S. K., Forouhar, F., Wei, F. Y., Tomizawa,
K., Hunt, J. F., Douki, T., Fontecave, M., Mulliez, E. & Atta, M. (2010).
de Matos, P., Adams, N., Hastings, J., Moreno, P. & Steinbeck, C.
(2012). A database for chemical proteomics: ChEBI. Methods Mol Biol
Identification of eukaryotic and prokaryotic methylthiotransferase for
biosynthesis of 2-methylthio-N6-threonylcarbamoyladenosine in
tRNA. J Biol Chem 285, 28425–28433.
Debarbouille, M., Gardan, R., Arnaud, M. & Rapoport, G. (1999). Role
Azarkina, N., Siletsky, S., Borisov, V., von Wachenfeldt, C.,
Hederstedt, L. & Konstantinov, A. A. (1999). A cytochrome bb9-type
quinol oxidase in Bacillus subtilis strain 168. J Biol Chem 274, 32810–
32817.
Barbe, V., Cruveiller, S., Kunst, F., Lenoble, P., Meurice, G.,
Sekowska, A., Vallenet, D., Wang, T., Moszer, I. & other authors
(2009). From a consortium sequence to a unified sequence: the
Bacillus subtilis 168 reference genome a decade later. Microbiology 155,
1758–1775.
Beckmann, B. M., Hoch, P. G., Marz, M., Willkomm, D. K., Salas, M. &
Hartmann, R. K. (2012). A pRNA-induced structural rearrangement
triggers 6S-1 RNA release from RNA polymerase in Bacillus subtilis.
EMBO J 31, 1727–1738.
Bertani, G. (1951). Studies on lysogenesis. I. The mode of phage
liberation by lysogenic Escherichia coli. J Bacteriol 62, 293–300.
Bocs, S., Cruveiller, S., Vallenet, D., Nuel, G. & Médigue, C. (2003).
AMIGene: annotation of microbial genes. Nucleic Acids Res 31, 3723–
3726.
Breaker, R. R. (2011). Prospects for riboswitch discovery and analysis.
Mol Cell 43, 867–879.
Bright, J. R., Byrom, D., Danson, M. J., Hough, D. W. & Towner, P.
(1993). Cloning, sequencing and expression of the gene encoding
glucose dehydrogenase from the thermophilic archaeon Thermoplasma acidophilum. Eur J Biochem 211, 549–554.
Prog Mol Biol Transl Sci 85, 1–41.
803, 273–296.
of bkdR, a transcriptional activator of the sigL-dependent isoleucine
and valine degradation pathway in Bacillus subtilis. J Bacteriol 181,
2059–2066.
Deikus, G. & Bechhofer, D. H. (2009). Bacillus subtilis trp leader RNA:
RNase J1 endonuclease cleavage specificity and PNPase processing.
J Biol Chem 284, 26394–26401.
Diethmaier, C., Pietack, N., Gunka, K., Wrede, C., Lehnik-Habrink,
M., Herzberg, C., Hübner, S. & Stülke, J. (2011). A novel factor
controlling bistability in Bacillus subtilis: the YmdB protein affects
flagellin expression and biofilm formation. J Bacteriol 193, 5997–
6007.
Dimmer, E. C., Huntley, R. P., Alam-Faruque, Y., Sawford, T.,
O’Donovan, C., Martin, M. J., Bely, B., Browne, P., Mun Chan, W. &
other authors (2012). The UniProt-GO Annotation database in 2011.
Nucleic Acids Res 40 (Database issue), D565–D570.
Engelen, S., Vallenet, D., Médigue, C. & Danchin, A. (2012). Distinct
co-evolution patterns of genes associated to DNA polymerase III
DnaE and PolC. BMC Genomics 13, 69.
Fang, M., Zeisberg, W. M., Condon, C., Ogryzko, V., Danchin, A. &
Mechold, U. (2009). Degradation of nanoRNA is performed by
multiple redundant RNases in Bacillus subtilis. Nucleic Acids Res 37,
5114–5125.
Florek, P., Levdikov, V. M., Blagova, E., Lebedev, A. A., Škrabana, R.,
Resetárová, S., Pavelcı́ková, P., Barak, I. & Wilkinson, A. J. (2011).
Carpousis, A. J. (2007). The RNA degradosome of Escherichia coli: an
The structure and interactions of SpoIISA and SpoIISB, a toxin–
antitoxin system in Bacillus subtilis. J Biol Chem 286, 6808–6819.
mRNA-degrading machine assembled on RNase E. Annu Rev
Microbiol 61, 71–87.
Fozo, E. M., Makarova, K. S., Shabalina, S. A., Yutin, N., Koonin, E. V.
& Storz, G. (2010). Abundance of type I toxin-antitoxin systems in
Caspi, R., Altman, T., Dreher, K., Fulcher, C. A., Subhraveti, P.,
Keseler, I. M., Kothari, A., Krummenacker, M., Latendresse, M. &
other authors (2012). The MetaCyc database of metabolic pathways
bacteria: searches for new candidates and discovery of novel families.
Nucleic Acids Res 38, 3743–3759.
and enzymes and the BioCyc collection of pathway/genome databases.
Nucleic Acids Res 40 (Database issue), D742–D753.
Chen, Y., Cao, S., Chai, Y., Clardy, J., Kolter, R., Guo, J. H. & Losick, R.
(2012). A Bacillus subtilis sensor kinase involved in triggering biofilm
formation on the roots of tomato plants. Mol Microbiol 85, 418–
430.
Gaballa, A., Antelmann, H., Aguilar, C., Khakh, S. K., Song, K. B.,
Smaldone, G. T. & Helmann, J. D. (2008). The Bacillus subtilis iron-
sparing response is mediated by a Fur-regulated small RNA and three
small, basic proteins. Proc Natl Acad Sci U S A 105, 11927–11932.
Gaballa, A., Newton, G. L., Antelmann, H., Parsonage, D., Upton, H.,
Rawat, M., Claiborne, A., Fahey, R. C. & Helmann, J. D. (2010).
Choi, K. H., Heath, R. J. & Rock, C. O. (2000). b-Ketoacyl-acyl carrier
Biosynthesis and functions of bacillithiol, a major low-molecularweight thiol in bacilli. Proc Natl Acad Sci U S A 107, 6482–6486.
protein synthase III (FabH) is a determining factor in branched-chain
fatty acid biosynthesis. J Bacteriol 182, 365–370.
Gilks, W. R., Audit, B., De Angelis, D., Tsoka, S. & Ouzounis, C. A.
(2002). Modeling the percolation of annotation errors in a database of
Choudhary, D. K. & Johri, B. N. (2009). Interactions of Bacillus spp.
protein sequences. Bioinformatics 18, 1641–1649.
and plants – with special reference to induced systemic resistance
(ISR). Microbiol Res 164, 493–513.
Gimpel, M., Heidrich, N., Mäder, U., Krügel, H. & Brantl, S. (2010). A
Christensen, Q. H., Martin, N., Mansilla, M. C., de Mendoza, D. &
Cronan, J. E. (2011). A novel amidotransferase required for lipoic
acid cofactor assembly in Bacillus subtilis. Mol Microbiol 80, 350–
363.
768
dual-function sRNA from B. subtilis: SR1 acts as a peptide encoding
mRNA on the gapA operon. Mol Microbiol 76, 990–1009.
Green, M. L. & Karp, P. D. (2006). The outcomes of pathway database
computations depend on pathway ontology. Nucleic Acids Res 34,
3687–3697.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159
Bacillus subtilis metabolic pathways 2012
Green, A. J., Rivers, S. L., Cheesman, M., Reid, G. A., Quaroni, L. G.,
Macdonald, I. D., Chapman, S. K. & Munro, A. W. (2001). Expression,
comprehensive database of Escherichia coli biology. Nucleic Acids Res
39 (Database issue), D583–D590.
purification and characterization of cytochrome P450 Biol: a novel
P450 involved in biotin synthesis in Bacillus subtilis. J Biol Inorg Chem
6, 523–533.
Kodama, T., Matsubayashi, T., Yanagihara, T., Komoto, H., Ara, K.,
Ozaki, K., Kuwana, R., Imamura, D., Takamatsu, H. & other authors
(2011). A novel small protein of Bacillus subtilis involved in spore
Guldan, H., Sterner, R. & Babinger, P. (2008). Identification and
germination and spore coat assembly. Biosci Biotechnol Biochem 75,
1119–1128.
characterization of a bacterial glycerol-1-phosphate dehydrogenase:
Ni2+-dependent AraM from Bacillus subtilis. Biochemistry 47, 7376–
7384.
Guldan, H., Matysik, F. M., Bocola, M., Sterner, R. & Babinger, P.
(2011). Functional assignment of an enzyme that catalyzes the
synthesis of an archaea-type ether lipid in bacteria. Angew Chem Int
Ed Engl 50, 8188–8191.
Hayes, F. & Van Melderen, L. (2011). Toxins–antitoxins: diversity,
evolution and function. Crit Rev Biochem Mol Biol 46, 386–408.
Latendresse, M., Paley, S. & Karp, P. D. (2012). Browsing metabolic
and regulatory networks with BioCyc. Methods Mol Biol 804, 197–216.
Lauhon, C. T. (2012). Mechanism of N6-threonylcarbamoyladenon-
sine (t(6)A) biosynthesis: isolation and characterization of the
intermediate threonylcarbamoyl-AMP. Biochemistry 51, 8950–8963.
(Epub ahead of print).
Heidrich, N., Chinali, A., Gerth, U. & Brantl, S. (2006). The small
Lehnik-Habrink, M., Newman, J., Rothe, F. M., Solovyova, A. S.,
Rodrigues, C., Herzberg, C., Commichau, F. M., Lewis, R. J. & Stülke,
J. (2011). RNase Y in Bacillus subtilis: a Natively disordered protein
untranslated RNA SR1 from the Bacillus subtilis genome is involved in
the regulation of arginine catabolism. Mol Microbiol 62, 520–536.
that is the functional equivalent of RNase E from Escherichia coli.
J Bacteriol 193, 5431–5441.
Henry, C. S., Zinner, J. F., Cohoon, M. P. & Stevens, R. L. (2009).
Lima, T., Auchincloss, A. H., Coudert, E., Keller, G., Michoud, K.,
Rivoire, C., Bulliard, V., de Castro, E., Lachaize, C. & other authors
(2009). HAMAP: a database of completely sequenced microbial
iBsu1103: a new genome-scale metabolic model of Bacillus subtilis
based on SEED annotations. Genome Biol 10, R69.
Holberger, L. E., Garza-Sánchez, F., Lamoureux, J., Low, D. A. &
Hayes, C. S. (2012). A novel family of toxin/antitoxin proteins in
Bacillus species. FEBS Lett 586, 132–136.
Hsiao, T. L., Revelles, O., Chen, L., Sauer, U. & Vitkup, D. (2010).
Automatic policing of biochemical annotations using genomic
correlations. Nat Chem Biol 6, 34–40.
Huang, B., Lv, C., Zhuang, P., Zhang, H. & Fan, L. (2011). Endophytic
colonisation of Bacillus subtilis in the roots of Robinia pseudoacacia L.
Plant Biol (Stuttg) 13, 925–931.
Hullo, M. F., Auger, S., Soutourina, O., Barzu, O., Yvon, M., Danchin,
A. & Martin-Verstraete, I. (2007). Conversion of methionine to
cysteine in Bacillus subtilis and its regulation. J Bacteriol 189, 187–
197.
Hunter, S., Jones, P., Mitchell, A., Apweiler, R., Attwood, T. K.,
Bateman, A., Bernard, T., Binns, D., Bork, P. & other authors (2012).
InterPro in 2011: new developments in the family and domain
prediction database. Nucleic Acids Res 40 (Database issue), D306–
D312.
Irnov, I., Sharma, C. M., Vogel, J. & Winkler, W. C. (2010).
Identification of regulatory RNAs in Bacillus subtilis. Nucleic Acids
Res 38, 6637–6651.
Jahn, N., Preis, H., Wiedemann, C. & Brantl, S. (2012). BsrG/SR4 from
Bacillus subtilis – the first temperature-dependent type I toxinantitoxin system. Mol Microbiol 83, 579–598.
Jensen, R. A., d’Amato, T. A. & Hochstein, L. I. (1988). An extreme-
halophile archaebacterium possesses the interlock type of prephenate
dehydratase characteristic of the Gram-positive eubacteria. Arch
Microbiol 149, 365–371.
Kaneda, T. (1991). Iso- and anteiso-fatty acids in bacteria:
biosynthesis, function, and taxonomic significance. Microbiol Rev
55, 288–302.
Karp, P. D., Latendresse, M. & Caspi, R. (2011). The pathway tools
pathway prediction algorithm. Stand Genomic Sci 5, 424–429.
Karsch-Mizrachi, I., Nakamura, Y., Cochrane, G. & International
Nucleotide Sequence Database Collaboration (2012). The
International Nucleotide Sequence Database Collaboration. Nucleic
Acids Res 40 (Database issue), D33–D37.
Keseler, I. M., Collado-Vides, J., Santos-Zavaleta, A., Peralta-Gil, M.,
Gama-Castro, S., Muñiz-Rascado, L., Bonavides-Martinez, C.,
Paley, S., Krummenacker, M. & other authors (2011). EcoCyc: a
http://mic.sgmjournals.org
proteome sets and manually curated microbial protein families in
UniProtKB/Swiss-Prot. Nucleic Acids Res 37 (Database issue), D471–
D478.
Liu, M. F., Cescau, S., Mechold, U., Wang, J., Cohen, D., Danchin, A.,
Boulouis, H. J. & Biville, F. (2012). Identification of a novel
nanoRNase in Bartonella. Microbiology 158, 886–895.
Lombard, J., López-Garcı́a, P. & Moreira, D. (2012). The early
evolution of lipid membranes and the three domains of life. Nat Rev
Microbiol 10, 507–515.
Mäder, U., Zig, L., Kretschmer, J., Homuth, G. & Putzer, H. (2008).
mRNA processing by RNases J1 and J2 affects Bacillus subtilis gene
expression on a global scale. Mol Microbiol 70, 183–196.
Mathy, N., Hébert, A., Mervelet, P., Bénard, L., Dorléans, A., Li de la
Sierra-Gallay, I., Noirot, P., Putzer, H. & Condon, C. (2010). Bacillus
subtilis ribonucleases J1 and J2 form a complex with altered enzyme
behaviour. Mol Microbiol 75, 489–498.
Meyer, F., Overbeek, R. & Rodriguez, A. (2009). FIGfams: yet another
set of protein families. Nucleic Acids Res 37, 6643–6654.
Müller, S., Hoffmann, T., Santos, H., Saum, S. H., Bremer, E. &
Müller, V. (2011). Bacterial abl-like genes: production of the archaeal
osmolyte N(e)-acetyl-b-lysine by homologous overexpression of the
yodP-kamA genes in Bacillus subtilis. Appl Microbiol Biotechnol 91,
689–697.
Nakano, M. M. & Zuber, P. (1998). Anaerobic growth of a ‘‘strict
aerobe’’ (Bacillus subtilis). Annu Rev Microbiol 52, 165–190.
Nakano, M. M., Dailly, Y. P., Zuber, P. & Clark, D. P. (1997).
Characterization of anaerobic fermentative growth of Bacillus subtilis:
identification of fermentation end products and genes required for
growth. J Bacteriol 179, 6749–6755.
Nakano, C., Ozawa, H., Akanuma, G., Funa, N. & Horinouchi, S.
(2009). Biosynthesis of aliphatic polyketides by type III polyketide
synthase and methyltransferase in Bacillus subtilis. J Bacteriol 191,
4916–4923.
Newton, G. L., Rawat, M., La Clair, J. J., Jothivasan, V. K., Budiarto, T.,
Hamilton, C. J., Claiborne, A., Helmann, J. D. & Fahey, R. C. (2009).
Bacillithiol is an antioxidant thiol produced in bacilli. Nat Chem Biol
5, 625–627.
Newton, G. L., Leung, S. S., Wakabayashi, J. I., Rawat, M. & Fahey,
R. C. (2011). The DinB superfamily includes novel mycothiol,
bacillithiol, and glutathione S-transferases. Biochemistry 50, 10751–
10760.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
769
E. Belda and others
Oku, H. & Kaneda, T. (1988). Biosynthesis of branched-chain fatty
Smith, A. A., Belda, E., Viari, A., Medigue, C. & Vallenet, D. (2012).
acids in Bacillus subtilis. A decarboxylase is essential for branchedchain fatty acid synthetase. J Biol Chem 263, 18386–18396.
The CanOE strategy: integrating genomic and metabolic contexts
across multiple prokaryote genomes to find candidate genes for
orphan enzymes. PLOS Comput Biol 8, e1002540.
Ostrowski, A., Mehert, A., Prescott, A., Kiley, T. B. & Stanley-Wall,
N. R. (2011). YuaB functions synergistically with the exopolysacchar-
Stok, J. E. & De Voss, J. (2000). Expression, purification, and
Ouzounis, C. A. & Karp, P. D. (2002). The past, present and future of
characterization of BioI: a carbon-carbon bond cleaving cytochrome
P450 involved in biotin biosynthesis in Bacillus subtilis. Arch Biochem
Biophys 384, 351–360.
genome-wide re-annotation. Genome Biol 3, comment2001.1.
Tatusov, R. L., Galperin, M. Y., Natale, D. A. & Koonin, E. V. (2000).
ide and TasA amyloid fibers to allow biofilm formation by Bacillus
subtilis. J Bacteriol 193, 4821–4831.
Overbeek, R., Fonstein, M., D’Souza, M., Pusch, G. D. & Maltsev, N.
(1999). The use of gene clusters to infer functional coupling. Proc Natl
Acad Sci U S A 96, 2896–2901.
Peretó, J., López-Garcı́a, P. & Moreira, D. (2004). Ancestral lipid
biosynthesis and early membrane evolution. Trends Biochem Sci 29,
469–477.
Pflüger, K., Baumann, S., Gottschalk, G., Lin, W., Santos, H. & Müller,
V. (2003). Lysine-2,3-aminomutase and beta-lysine acetyltransferase
genes of methanogenic archaea are salt induced and are essential for
the biosynthesis of Ne-acetyl-b-lysine and growth at high salinity. Appl
Environ Microbiol 69, 6047–6055.
Ramazzina, I., Costa, R., Cendron, L., Berni, R., Peracchi, A., Zanotti,
G. & Percudani, R. (2010). An aminotransferase branch point
connects purine catabolism to amino acid recycling. Nat Chem Biol
6, 801–806.
Ruepp, A., Müller, H. N., Lottspeich, F. & Soppa, J. (1995). Catabolic
ornithine transcarbamylase of Halobacterium halobium (salinarium):
purification, characterization, sequence determination, and evolution.
J Bacteriol 177, 1129–1136.
Santana, M., Kunst, F., Hullo, M. F., Rapoport, G., Danchin, A. &
Glaser, P. (1992). Molecular cloning, sequencing, and physiological
characterization of the qox operon from Bacillus subtilis encoding the
aa3–600 quinol oxidase. J Biol Chem 267, 10225–10231.
Sekowska, A., Coppée, J. Y., Le Caer, J. P., Martin-Verstraete, I. &
Danchin, A. (2000). S-adenosylmethionine decarboxylase of Bacillus
subtilis is closely related to archaebacterial counterparts. Mol
Microbiol 36, 1135–1147.
Sekowska, A., Dénervaud, V., Ashida, H., Michoud, K., Haas, D.,
Yokota, A. & Danchin, A. (2004). Bacterial variations on the
methionine salvage pathway. BMC Microbiol 4, 9.
Sekowska, A., Masson, J. B., Celani, A., Danchin, A. & Vergassola, M.
(2009). Repulsion and metabolic switches in the collective behavior of
bacterial colonies. Biophys J 97, 688–698.
Shipkowski, S. & Brenchley, J. E. (2006). Bioinformatic, genetic, and
The COG database: a tool for genome-scale analysis of protein
functions and evolution. Nucleic Acids Res 28, 33–36.
Tatusov, R. L., Fedorova, N. D., Jackson, J. D., Jacobs, A. R., Kiryutin,
B., Koonin, E. V., Krylov, D. M., Mazumder, R., Mekhedov, S. L. &
other authors (2003). The COG database: an updated version
includes eukaryotes. BMC Bioinformatics 4, 41.
Upton, H., Newton, G. L., Gushiken, M., Lo, K., Holden, D., Fahey,
R. C. & Rawat, M. (2012). Characterization of BshA, bacillithiol
glycosyltransferase from Staphylococcus aureus and Bacillus subtilis.
FEBS Lett 586, 1004–1008.
Vallenet, D., Labarre, L., Rouy, Z., Barbe, V., Bocs, S., Cruveiller, S.,
Lajus, A., Pascal, G., Scarpelli, C. & Médigue, C. (2006). MaGe: a
microbial genome annotation system supported by synteny results.
Nucleic Acids Res 34, 53–65.
Vallenet, D., Engelen, S., Mornico, D., Cruveiller, S., Fleury, L., Lajus,
A., Rouy, Z., Roche, D., Salvignol, G. & other authors (2009).
MicroScope: a platform for microbial genome annotation and
comparative genomics. Database (Oxford) 2009, bap021.
Van Arsdell, S. W., Perkins, J. B., Yocum, R. R., Luan, L., Howitt, C. L.,
Chatterjee, N. P. & Pero, J. G. (2005). Removing a bottleneck in the
Bacillus subtilis biotin pathway: bioA utilizes lysine rather than Sadenosylmethionine as the amino donor in the KAPA-to-DAPA
reaction. Biotechnol Bioeng 91, 75–83.
Winstedt, L. & von Wachenfeldt, C. (2000). Terminal oxidases of
Bacillus subtilis strain 168: one quinol oxidase, cytochrome aa3 or
cytochrome bd, is required for aerobic growth. J Bacteriol 182, 6557–
6564.
Winstedt, L., Yoshida, K., Fujita, Y. & von Wachenfeldt, C. (1998).
Cytochrome bd biosynthesis in Bacillus subtilis: characterization of
the cydABCD operon. J Bacteriol 180, 6571–6580.
Wu, X., Wang, X., Drlica, K. & Zhao, X. (2011). A toxin–antitoxin
module in Bacillus subtilis can both mitigate and amplify effects of
lethal stress. PLoS ONE 6, e23909.
Yoshida, K., Yamaguchi, M., Morinaga, T., Ikeuchi, M., Kinehara, M.
& Ashida, H. (2006). Genetic modification of Bacillus subtilis for
biochemical evidence that some glycoside hydrolase family 42 betagalactosidases are arabinogalactan type I oligomer hydrolases. Appl
Environ Microbiol 72, 7730–7738.
production of D-chiro-inositol, an investigational drug candidate for
treatment of type 2 diabetes and polycystic ovary syndrome. Appl
Environ Microbiol 72, 1310–1315.
Shomura, Y., Hinokuchi, E., Ikeda, H., Senoo, A., Takahashi, Y.,
Saito, J., Komori, H., Shibata, N., Yonetani, Y. & Higuchi, Y. (2012).
Yoshida, K., Yamaguchi, M., Morinaga, T., Kinehara, M., Ikeuchi, M.,
Ashida, H. & Fujita, Y. (2008). myo-Inositol catabolism in Bacillus
Structural and enzymatic characterization of BacD, an L-amino acid
dipeptide ligase from Bacillus subtilis. Protein Sci 21, 707–716.
subtilis. J Biol Chem 283, 10415–10424.
Smaldone, G. T., Antelmann, H., Gaballa, A. & Helmann, J. D. (2012).
complex (menaquinone : cytochrome c reductase) in Bacillus subtilis
has a nontraditional subunit organization. J Bacteriol 177, 6751–6760.
The FsrA sRNA and FbpB protein mediate the iron-dependent
induction of the Bacillus subtilis lutABC iron-sulfur-containing
oxidases. J Bacteriol 194, 2586–2593.
770
Yu, J., Hederstedt, L. & Piggot, P. J. (1995). The cytochrome bc
Edited by: W. J. Quax
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 00:13:16
Microbiology 159