Download A comparative genome analysis of the RpoS sigmulon shows a high

Microbiology (2011), 157, 1393–1401
DOI 10.1099/mic.0.042937-0
A comparative genome analysis of the RpoS
sigmulon shows a high diversity of responses
and origins
Alberto Santos-Zavaleta,1 Socorro Gama-Castro1
and Ernesto Pérez-Rueda2
1
Correspondence
Programa de Genómica Computacional, Centro de Ciencias Genómicas, Universidad Nacional
Autónoma de México, Cuernavaca, Morelos, Mexico
Ernesto Pérez-Rueda
[email protected]
2
Departamento de Ingenierı́a Celular y Biocatálisis, Instituto de Biotecnologı́a, Universidad Nacional
Autónoma de México, Cuernavaca, Morelos, Mexico
Received 24 June 2010
Revised
19 January 2011
Accepted 3 February 2011
The stationary-phase response mediated by the RpoS sigma factor (sS, s38) has been widely
studied as a general mechanism of activation of highly diverse genes that maintain cell viability. In
bacteria, genes for diverse functions have been associated with this response, showing that
bacteria use a large number of functions to contend with adverse conditions in their environment.
However, little is known about how the genes have been functionally recruited in diverse
organisms. In this work, we address the analysis of genes regulated by sS, based on a
comparative genomic-scale analysis considering four versatile bacterial species that represent
different lifestyles and taxonomic groups, Escherichia coli K-12, Geobacter sulfurreducens,
Borrelia burgdorferi and Bacillus subtilis, as well as the extent of conservation in bacterial
genomes, as a means of assessing the evolution of this sigmulon across all organisms completely
sequenced. The analysis presented here shows that genes associated with the sS response have
been recruited from diverse regulons to achieve a global response. In addition, and based on the
distribution of orthologues, we show a group of genes that is highly conserved among all
organisms, mainly associated with glycerol metabolism, as well as diverse functional genes
recruited in a lineage-specific manner.
INTRODUCTION
Adaptive responses associated with environmental changes
include the modification of the genetic program and, as a
consequence, changes in metabolism. In bacteria, gene
expression is modulated through the replacement of sigma
(s) factors on the core RNA polymerase (Martı́nez-Garcı́a
et al., 2001), which provides bacteria with the ability to
express different genes under different metabolic stimuli or
growth conditions (Hengge-Aronis, 1999, 2000). In the
bacterium Escherichia coli K-12, seven s factors have been
experimentally described, with RpoD (s70) responsible for
recognizing and activating almost 30 % of its 4605 genes,
whereas alternative s factors, such as RpoS (sS), the master
regulator of the stationary-phase response (Hengge-Aronis,
1996, 2002; Loewen et al., 1998), are associated with
regulation of a wide variety of physiological processes.
Abbreviations: TCA cycle, tricarboxylic acid cycle; TF, transcription
factor.
A supplementary figure, showing a single linkage-clustering algorithm
with none leaf order optimization applied with Manhattan distance as
similarity measure, is available with the online version of this paper.
042937 G 2011 SGM
The master regulators control a large diversity of genes,
hierarchically organized in regulons. Regulons are the
major regulatory systems that provide a controlled gene
expression response to environmental conditions (Dufour
et al., 2010). Diverse comparative studies have suggested
that regulons exhibit plasticity across the evolution of
bacterial species (Lozada-Chávez et al., 2006), e.g. the
PhoPQ regulon, which contains over 200 regulated genes
in E. coli and Salmonella typhimurium, of which only 13
operons are common to the two regulons (Monsieurs
et al., 2005). Additionally, Liu & Ochman (2007) have
suggested that the plasticity in bacterial regulons is
evidence of lineage-specific modifications. Therefore, it
seems that there is a selective pressure to maintain essential
core genes within regulons, as well as to allow for plasticity
and for lineage-specific changes in regulon composition
between species. In this regard, diverse s factors have been
recruited to regulate genes for non-essential functions,
although they may be important for adaptation or survival
of the cell in nature. For instance, King et al. (2006) and
Zambrano et al. (1993) observed a rapid loss of the RpoS
regulon when bacteria were subjected to environmental
pressures.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Printed in Great Britain
1393
A. Santos-Zavaleta, S. Gama-Castro and E. Pérez-Rueda
The RpoS sigma factor has been classified as a member of
the s70 protein family (Paget & Helmann, 2003). The
functional role of sS as master regulator during stationary
phase has been clearly defined in E. coli, although diverse
studies have been performed in other bacteria, such as
Pseudomonas aeruginosa, Salmonella sp. (Loewen et al.,
1998) and Legionella pneumophila (Hales & Shuman,
1999), and in plant pathogens such as Erwinia carotovora,
Erwinia herbicola and Ralstonia solanacearum (Loewen
et al., 1998; Martı́nez-Garcı́a et al., 2001). In the Firmicute
Bacillus subtilis, the same response has been described and
is associated with the sB factor. Paget & Helmann (2003)
suggested, based on a phylogenetic analysis, that sS of E.
coli, Geobacter sulfurreducens and Borrelia burgdorferi, and
sB of B. subtilis, belong to the same group of proteins and
are closely related to the primary s70 factors, although they
are dispensable for bacterial cell growth.
Accordingly, the availability of well-known genes under the
control of sS in E. coli, G. sulfurreducens, Borrelia
burgdorferi and B. subtilis, and the high number of
completely sequenced genomes, provide an excellent
opportunity to perform exhaustive analysis of sS-dependent
genes in a large diversity of organisms. In this regard,
organisms in addition to E. coli have been analysed to
identify genes under the control of this s factor, including G. sulfurreducens, a delta-proteobacterium in which
sS regulates genes with diverse functions such as those
involved in energy metabolism and the tricarboxylic acid
(TCA) cycle, among others, and plays a fundamental role
in regulating the normal physiology of the cell (Núñez
et al., 2006). In the spirochaete Borrelia burgdorferi, the
causative agent of Lyme disease, sS controls the expression
of 130 genes, many of which are differentially expressed
during tick feeding, and some of which are required for or
associated with mammalian host infection (Caimano et al.,
2004). Finally, B. subtilis sB, the functional orthologue of E.
coli sS, is responsible for the transcription of genes that can
confer stress resistance upon the vegetative cell, as well as
antibiotic resistance and pathogenicity, and which can affect
cellular differentiation, among other processes (Hecker
et al., 2007; Völker et al., 1999).
In this work we analysed the taxonomic distribution of sSregulated genes in order to deduce the common principles
by which this sigmulon has been assembled through
evolution. Based on the RpoS sigmulon of each organism,
orthologues were identified in all the genomes in order to
determine the genes that are universally conserved versus
those genes with probable lineage-specific distributions.
Here, we show that although these bacterial species contain
similar proportions of genes under sS control, the majority
of them have been independently recruited for all four
bacterial species. Thus, these results suggest not only a low
selective pressure to maintain regulon composition but
also that there are a large number of recruitment events in
bacteria. We suggest that the basic metabolic processes
associated with the regulatory networks for these organisms have been preferentially recruited. These results open
1394
diverse opportunities to understand and decipher the
regulatory systems in all the organisms, especially in
association with stress responses.
METHODS
Datasets and genomic sequences. A total of 900 genome
sequences were retrieved from GenBank (Benson et al., 2008), and
gene functional annotations were collected from the Clusters of
Orthologous Groups (COG) database (http://www.ncbi.nlm.nih.gov/
COG/) (Woodsmall & Benson, 1993). A total of 679 non-redundant
genomes were considered for the discussion of the data. A complete
list of non-redundant genomes can be obtained from http://
popolvuh.wlu.ca/Phyl_Profiles/NR_genomes/REDUNDANCY.html.
From this non-redundant dataset, 50 sequences corresponded to
Archaea genomes (15 Crenarchaeota, 33 Euryarchaeota, one
Korarchaeota and one Nanoarchaeota species), 611 corresponded
to Bacteria genomes (two Acidobacteria, 47 Actinobacteria, 12
Bacteroides, 13 Chlamydia, 32 Cyanobacteria, four Deinococcus–
Thermus, 46 Firmicutes Bacillales, 32 Clostridium, 47 Lactobacillus,
19 Mollicutes, one Fusobacterium, seven green nonsulfur bacteria,
five green sulfur bacteria, seven hyperthermophilic bacteria, one
Magnetococcus, one Planctomyces, 82 alpha-proteobacteria, 47 betaproteobacteria, 21 delta-proteobacteria, 19 epsilon-proteobacteria,
156 gamma-proteobacteria, nine Spirochaetes and one Synergistetes
species), and 18 corresponded to Eukarya genomes.
In this work, non-redundant genomes refer to representative bacterial
species, as defined by Janga & Moreno-Hagelsieb (2004). In brief, if
there are diverse strains of the same species, a representative genome
is considered; however, the order of elimination follows the
importance of certain species as model organisms (E. coli K-12, B.
subtilis) and then the order of importance follows the highest number
of genes having orthologues across phyla. For instance, Mycobacterium avium strain 104 is representative of diverse Mycobacterium
strains (Mycobacterium avium subsp. paratuberculosis, Mycobacterium
bovis, M. bovis BCG Pasteur 1173P2, Mycobacterium leprae and
Mycobacterium smegmatis mc2 155).
Data for regulatory interactions. A collection of 195 sS-dependent
genes (organized in 130 transcriptional units) experimentally
identified in E. coli K-12 was obtained from RegulonDB release 6.7
(Gama-Castro et al., 2008), a specialized database relating to gene
regulation in this bacterium and which contains experimental
information extracted from the literature. Alternatively, 67 sBdependent genes from B. subtilis were retrieved from DBTBS
(Sierro et al., 2008), a database devoted to the regulatory elements
in this firmicute. A total of 130 sS-dependent genes of Borrelia
burgdorferi and 342 sS-dependent genes of G. sulfurreducens were also
identified from the published literature (Caimano et al., 2004; Núñez
et al., 2006).
Identification of orthologous genes. Orthologues are defined as
proteins in different species that evolved from a common ancestor by
speciation (Fitch, 1970), and they usually have the same function.
Operationally, orthologous sequences are usually defined as the bestmatching homologues or bidirectional best hits (BDBHs) in another
organism (Tatusov et al., 1997). In this work, orthologues were
identified by using the BDBHs definition through depurated genomes
at 95 % identity, with an E-value of ¡1e-6, as described elsewhere
(Janga & Moreno-Hagelsieb, 2004). To corroborate the probable
function associated with all orthologues detected, we identified their
protein domain arrangements. Domain assignments were retrieved
from Pfam database release 24.0 (Finn et al., 2010), into which a large
library of domains has been deposited.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Microbiology 157
Genomic analysis of the RpoS sigmulon
RESULTS AND DISCUSSION
Characterization of the RpoS sigmulon in E. coli,
G. sulfurreducens, Borrelia burgdorferi and
B. subtilis
Borrelia burgdorferi and B. subtilis. Thus, we first describe
the regulatory levels and regulons associated with the sS
response in E. coli K-12 and their probable functional
implications. After that, we discuss the sigmulon organization in the other bacterial models.
The main goal of this study was to refine and expand the
knowledge of transcriptional regulation of genes under the
control of sS from the perspective of four different
bacterial species representing a wide diversity of taxonomic
characteristics and lifestyles: E. coli, G. sulfurreducens,
The E. coli K-12 RpoS sigmulon (Fig. 1) represents around
4.3 % of the genes in this bacterium, i.e. 195 of all its genes
are under the control of sS. When these genes were
dissected in terms of their regulatory mechanism, 19 %
were found to be exclusively dependent on RpoS, i.e. so far,
Fig. 1. The RpoS sigmulon in E. coli K-12. s Factors are shown in yellow in the upper and middle portions of the network. The
48 orange ellipses represent the DNA-binding TFs, and the 195 light-blue ellipses represent the target genes. The green and
red arrows represent transcriptional activation and repression, respectively.
http://mic.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
1395
A. Santos-Zavaleta, S. Gama-Castro and E. Pérez-Rueda
only promoters associated with RpoS had been identified.
In contrast, 62 % of the genes contained an additional
promoter associated with s70 or alternative s factors.
Finally, 58 % of the genes associated with this regulon were
finely regulated by 49 different transcription factors (TFs).
From these data, it is interesting that the expression of a
large fraction (38 %) of the genes in the sigmulon is
regulated by global regulators, such as Crp, Fnr and Lrp,
suggesting that sS-regulated genes may also play important
roles in other responses, as proposed for the hyperosmolarity and low-pH responses (Hengge-Aronis, 1996, 2002;
Loewen et al., 1998). Additionally, diverse regulons were
almost exclusively controlled by sS, such as the GadX
regulon, which is involved in the response to acid
environments (Masuda & Church, 2003; Tramonti et al.,
2003), in which 70 % of the genes included in the regulon
are controlled by sS. In summary, these findings suggest
not only that sS-regulated genes can participate specifically
in the growth phase response but also that diverse regulons
are associated with independent responses beyond sS.
Thus, we suggest that diverse genes associated with this
response have been recruited to play an important role in
the stationary-phase growth response, with no apparent
functional relationship among them.
In the following section, we discuss some of the evidence
concerning gene recruitment associated with the sS
response. In this regard, the sS network exhibits extensive
regulatory overlaps with other global and independent
regulons, such as with the Lrp protein regulon. Indeed,
certain modules of the sS-dependent general stress
response can be temporarily recruited by stress-specific
regulons, which are controlled by other stress-responsive
regulators that act together with the s70 RNA polymerase.
Thus, not only the expression of genes within a regulatory
network but also the architecture of the network itself can
be subject to regulation (Weber et al., 2005). As an
alternative, we explored other bacterial models to obtain a
more integrated scenario to explain our preliminary
observations, such as the fact that diverse sS-regulated
genes in E. coli can also act independently of RpoS. Based
on a review of the published literature, we analysed the G.
sulfurreducens, Borrelia burgdorferi and B. subtilis sS
responses. From this analysis we found that the repertoire
of sS-dependent genes in these organisms represents a low
proportion of the total repertoire of genes per organism, as
observed for E. coli, i.e. 1.7 % of the total genes in B.
subtilis, 7.5 % in Borrelia burgdorferi and 9.4 % in G.
sulfurreducens are under the regulation of sS. Additionally,
in these organisms, a large proportion of genes are
positively regulated by sS: 70 % in G. sulfurreducens and
80 % in Borrelia burgdorferi. These data suggest that the
observed negative regulation is an indirect effect of sS on
the other genes (30 % in G. sulfurreducens and 20 % in
Borrelia burgdorferi) through the regulation of global
regulators. In B. subtilis, diverse genes associated with this
sigmulon are also regulated by additional s factors (sA) or
TFs, as occurs in E. coli. In summary, sS response genes are
1396
associated with additional regulatory systems, suggesting
that this response is integrated by a large diversity of
regulons that can act independently of the stationary-phase
response.
Functional categories associated with the
regulated genes show high diversity
In order to acquire information concerning functions
common to genes regulated by sS, we classified them into
four global classes according to the COG database: (i)
information storage and processing, (ii) cellular processes
and signalling, (iii) metabolism and (iv) poorly characterized (Fig. 2). This functional classification shows that
metabolism, cellular processes and signalling are the most
prominent categories associated with sS-regulated genes in
E. coli and G. sulfurreducens. However, cellular processes
and signalling was the most abundant category associated
with genes under the control of sS in Borrelia burgdorferi,
while in B. subtilis, genes regulated by sB fell into the
metabolism category. Thus, the cellular processes and
signalling group is the most prominent functional category
associated with genes under the control of sS in the four
bacterial species analysed, followed by the metabolism
category. This finding suggests that genes are functionally
devoted to contending with metabolic challenges rather
than processes associated with information storage and
processing, probably as a consequence of bacterial adaptation. When we dissected the whole repertoire of RpoSregulated genes into smaller classes, we found that the most
abundant categories identified in E. coli were energy
production and conversion (19 % of the genes fell into
this category), and transcription processes (17 % of the
genes). In G. sulfurreducens, the distribution of the most
abundant categories was as follows: 14.3 % of the genes
were included in the energy production and conversion
class, and 13 % in cell wall/membrane/envelope biogenesis.
In contrast, in Borrelia burgdorferi, 22 % of the genes were
included in the cell cycle control category, and the cell
division, chromosome-partitioning category. Finally, for B.
subtilis, 14 % of the sS-regulated genes fell into the signal
transduction mechanism category, and this was the most
abundant functional category for this bacterium. Therefore, we suggest that this slight overrepresentation of
metabolism-related genes in some of the four bacteria
species could be attributed to the niches in which these
organisms live. Therefore, an exhaustive sequence comparison of the repertoire of genes in all the genomes could
contribute to an understanding of how variable or conserved the genes belonging to this sigmulon are across the
evolutionary process.
Taxonomic distribution of the RpoS sigmulon
Based on a taxonomic distribution of the sS-regulated
genes, we evaluated how this sigmulon has been assembled
during evolution and how versatile the stress response is
in different bacteria. In this regard the repertoire of
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Microbiology 157
Genomic analysis of the RpoS sigmulon
Fig. 2. Functional categories of RpoScontrolled genes. On the x axis, genes are
grouped according to their functional categories in the COG database. The y axis shows the
proportion of genes in each category.
sS-regulated genes in all four bacterial species was used to
identify orthologous genes in 679 complete genome
sequences, from the Domains Bacteria, Archaea and
Eukarya. We considered that a wide distribution and high
number of orthologues in relation to the bacterial species
used as a framework would strongly suggest conservation
of the sigmulon in all species considered, whereas a narrow
distribution of orthologues would suggest a lineage-specific
evolution.
In the first stage of our analysis, we determined the
fractions of the total numbers of sS-dependent genes of E.
coli, G. sulfurreducens, B. subtilis and Borrelia burgdorferi
that were related by orthology. From this analysis we
identified a small proportion of orthologues shared among
them (Table 1); for instance, only 16 genes of 195 under the
control of sS in E. coli exhibited an orthologue in Borrelia
burgdorferi. These findings suggest that genes under the
control of sS are likely lost to a great extent at such
phylogenetic distances. These observations gave rise to
several questions concerning the evolutionary and functional conservation of the sS sigmulon among these
bacterial genomes. Thus, to gain insights into the commonalities and differences in gene regulation among
prokaryotic species from the perspective of the sS
sigmulon, we used the complete repertoires of genes under
the regulation of RpoS in E. coli, G. sulfurreducens, B.
subtilis and Borrelia burgdorferi to scan whole genome
sequences and determine the distributions of these genes in
different taxa.
The orthologue distribution identified in bacteria suggests
that the sS response is a consequence of diverse gene
recruitment events involving multiple and independent
responses. We found that 70 % of the genes associated with
sS in Borrelia burgdorferi were lineage-specific, i.e. most of
the genes associated with sS in this bacterium do not
exhibit orthologues in other taxonomic groups, except in
the Spirochaetes; indeed, sS not only is the master
regulator of the general stress response in this organism
but also appears to act as a stress-responsive activator of a
subset of virulence determinants that encompass the
spirochaete genetic programs required for mammalian
host adaptation (Caimano et al., 2004). From these data,
we suggest that bacteria belonging to the Spirochaetes have
a specific sS response and an evolutionary pathway
independent from the other taxonomic groups analysed
in this work (Figs 3 and 4). We also suggest that a lack of
conservation beyond the Spirochaetes and even Borrelia
burgdorferi is a consequence of the large diversity of
organisms included in this phylum. In G. sulfurreducens
and E. coli we found that most of the sS-regulated genes
Table 1. RpoS-regulated genes and orthologues in the four bacterial species studied
Species
Number of RpoS-regulated genes* and number of orthologues in other species
E. coli
Borrelia burgdorferi
G. sulfurreducens
B. subtilis
E. coli
Borrelia burgdorferi
G. sulfurreducens
195
16
49
47
130
29
10
342
10
B. subtilis
68
*Values shown in bold type indicate the number of RpoS-regulated genes in that species.
http://mic.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
1397
A. Santos-Zavaleta, S. Gama-Castro and E. Pérez-Rueda
Fig. 3. Orthologous distributions by phylum from the perspective of E. coli, G. sulfurreducens, Borrelia burgdorferi and B.
subtilis RpoS-regulated genes. The bars show the relative abundance of orthologues by phylum, with a value of 1
corresponding to 100 % presence and 0 representing absence; values were calculated as follows: relative abundance by
phylum5(total number of orthologues identified)/(total number of representative organisms by phylum).
(around 90 %) shared at least one orthologue in another
taxonomic group. A similar result was found with B.
subtilis, in which around 80 % the sB-regulated genes had
an orthologue in another taxonomic group. Another
interesting result was the fact that the Phylum
Synergitetes had few orthologues among the bacterial
species studied: one orthologue with B. subtilis, 11 with G.
sulfurreducens, none with Borrelia burgdorferi and four with
E. coli, suggesting that in this phylum, the stress response is
based on completely different elements. Similar results
were found in other bacteria, such as Chlamydia,
Fusobacterium and hyperthermophilic bacteria. In the
archaeon Nanoarchaeum equitans, we could not find
orthologues that were likely to be associated with this
response. We believe that in these organisms, diverse genes
have been lost as a consequence of their lifestyles (i.e. they
are mostly pathogens or endosymbionts whose environments are relatively stable).
A more exhaustive analysis based on the distribution of
orthologues of Borrelia burgdorferi was conducted (Fig. 4),
because this species displays a considerable genome
reduction, in which minimal genetic information has been
preserved. From the data illustrated in Fig. 4, we identified
two main clusters of proteins with a widely distributed
pattern, suggesting that their distribution has been highly
preserved throughout evolution in prokaryotes and
eukaryotes. The first cluster contains proteins mainly
associated with glycerol metabolism. Functionally, glycerol
is an important intermediate of energy metabolism: it can
1398
control osmotic activity and crystal formation and then act
as a cryoprotective agent (Brisson et al., 2001). In the
second cluster, diverse periplasmic proteins, hypothetical
proteins and methyl-accepting chemotaxis proteins were
included. Methyl-accepting chemotaxis proteins are
important integral membrane elements that undergo
reversible methylation during adaptation of bacterial cells
to environmental attractants and repellants (Kehry &
Dahlquist, 1982). Therefore, the main result of this analysis
is that a cluster of genes related to glycerol metabolism is
conserved throughout all organisms, and also that there are
a large number of lineage-specific genes (specific to
spirochaetes). A similar result was obtained when E. coli,
G. sulfurreducens and B. subtilis were considered (see
Supplementary Fig. S1): a main cluster of conserved
proteins was identified, mainly associated with carbon
source assimilation, e.g. succinyl-CoA synthetase, which
catalyses the only substrate-level phosphorylation in the
TCA cycle (Weitzman, 1981).
Conclusions
In this study we analysed and compared the stress
responses mediated by sS in complete bacterial genome
sequences and their conservation in bacteria from diverse
taxonomic groups. E. coli, G. sulfurreducens, Borrelia
burgdorferi and B. subtilis were considered as models for
the analysis of this response because there is a large amount
of information concerning their sS-regulated genes, and
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Microbiology 157
http://mic.sgmjournals.org
1399
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Genomic analysis of the RpoS sigmulon
Fig. 4. Clustering of orthologues from the perspective of Borrelia burgdorferi. A single linkage-clustering algorithm with no leaf order optimization was applied with Manhattan
distance as the similarity measure. The display clustering results were obtained by using the MeV program (http://www.tm4.org/mev/). The conserved groups across the
different taxonomic groups, mainly associated with glycerol metabolism, are indicated. Each column denotes an RpoS-regulated gene in Borrelia burgdorferi, whereas rows
denote taxonomic groups. The bar at the top of the figure indicates the relative abundance of orthologues per group, with a value of 1 corresponding to 100 % presence and 0
representing absence.
A. Santos-Zavaleta, S. Gama-Castro and E. Pérez-Rueda
because they represent a wide diversity of taxonomic
groups. In addition, our analysis included organisms from
three archaeal taxonomic groups, 23 bacterial taxonomic
groups, and 18 eukarya genomes; therefore, we believe that
our results are valid, allowing high confidence for a wide
range of organisms beyond those considered here. In all
these bacteria, the set of sS-regulated genes is involved in
important cellular processes, such as energy production
and conservation, and signal transduction mechanisms. In
this regard, the evolution of regulatory networks associated
with sS might be influenced by genomic variation at the
level of the TF repertoire as a consequence of their
lifestyles, as suggested by Lozada-Chávez et al. (2006).
S
In this regard, we found that s in Borrelia burgdorferi was
spirochaete-specific, i.e. most genes associated with RpoS
in this bacterium are class-specific, and this relates to the
results of Caimano et al. (2004), who determined that sS is
not central to the general stress response but serves only a
limited role in environmental stress adaptation. Those
results were probably due to the limited biosynthetic
capabilities of Borrelia burgdorferi: it does not exist in
nature as a free-living organism, and therefore is confined
to a narrow host range with only limited exposure to
environmental stresses (Caimano et al., 2004).
In addition, our findings were unexpected in light of
the taxonomic distribution of the orthologues of RpoSregulated genes, suggesting that the RpoS regulon has
evolved to adapt to diverse growth environments. A similar
result has recently been described (Chiang & Schellhorn,
2010) in a study that compared the RpoS regulons in
diverse bacterial species. In this regard, a high proportion
of orthologues was identified in other E. coli strains (more
than 85 % of the RpoS-regulated genes), whereas 90
orthologues were identified in P. aeruginosa (a similar
finding was obtained by Chiang & Schellhorn, 2010).
In summary, based on comparative genomics, we have
determined that the evolution of the sS sigmulon seems to
have involved diverse losses and gains of regulatory
interactions, with the regulons maintaining a large functional independence. In this regard, it is possible that large
portions of the regulatory network associated with RpoS
evolved through extensive genetic changes during the
evolution of the species studied. Indeed, it has been
reported that two elements contribute to many speciesspecific modifications of the regulons: (i) the frequent loss
of the RpoS sigma factor due to environmental selection,
and (ii) the absence of essential genes in the RpoS regulon
(Chen et al., 2004; Dong et al., 2009; Zambrano et al.,
1993). All these elements suggest that in each speciation
event, transcriptional regulation is highly flexible for
phenotypic adaptation.
grant from Dirección General Asuntos del Personal AcadémicoUniversidad Nacional Autónoma de México (DGAPA-UNAM)
(IN-217508).
REFERENCES
Benson, D. A., Karsch-Mizrachi, I., Lipman, D. J., Ostell, J. & Wheeler,
D. L. (2008). GenBank. Nucleic Acids Res 36 (Database issue), D25–D30.
Brisson, D., Vohl, M. C., St-Pierre, J., Hudson, T. J. & Gaudet, D.
(2001). Glycerol: a neglected variable in metabolic processes?
Bioessays 23, 534–542.
Caimano, M. J., Eggers, C. H., Hazlett, K. R. & Radolf, J. D. (2004).
RpoS is not central to the general stress response in Borrelia
burgdorferi but does control expression of one or more essential
virulence determinants. Infect Immun 72, 6433–6445.
Chen, G., Patten, C. L. & Schellhorn, H. E. (2004). Positive selection
for loss of RpoS function in Escherichia coli. Mutat Res 554, 193–203.
Chiang, S. M. & Schellhorn, H. E. (2010). Evolution of the RpoS
regulon: origin of RpoS and the conservation of RpoS-dependent
regulation in bacteria. J Mol Evol 70, 557–571.
Dong, T., Chiang, S. M., Joyce, C., Yu, R. & Schellhorn, H. E. (2009).
Polymorphism and selection of rpoS in pathogenic Escherichia coli.
BMC Microbiol 9, 118.
Dufour, Y. S., Kiley, P. J. & Donohue, T. J. (2010). Reconstruction of
the core and extended regulons of global transcription factors. PLoS
Genet 6, e1001027.
Finn, R. D., Mistry, J., Tate, J., Coggill, P., Heger, A., Pollington, J. E.,
Gavin, O. L., Gunasekaran, P., Ceric, G. & other authors (2010). The
Pfam protein families database. Nucleic Acids Res 38 (Database issue),
D211–D222.
Fitch, W. M. (1970). Distinguishing homologous from analogous
proteins. Syst Zool 19, 99–113.
Gama-Castro, S., Jiménez-Jacinto, V., Peralta-Gil, M., SantosZavaleta, A., Peñaloza-Spinola, M. I., Contreras-Moreira, B.,
Segura-Salazar, J., Muñiz-Rascado, L., Martı́nez-Flores, I. & other
authors (2008). RegulonDB (version 6.0): gene regulation model of
Escherichia coli K-12 beyond transcription, active (experimental)
annotated promoters and Textpresso navigation. Nucleic Acids Res 36
(Database issue), D120–D124.
Hales, L. M. & Shuman, H. A. (1999). The Legionella pneumophila rpoS
gene is required for growth within Acanthamoeba castellanii.
J Bacteriol 181, 4879–4889.
Hecker, M., Pané-Farré, J. & Völker, U. (2007). SigB-dependent
general stress response in Bacillus subtilis and related Gram-positive
bacteria. Annu Rev Microbiol 61, 215–236.
Hengge-Aronis, R. (1996). Back to log phase: sS as a global regulator
in the osmotic control of gene expression in Escherichia coli. Mol
Microbiol 21, 887–893.
Hengge-Aronis, R. (1999). Interplay of global regulators and cell
physiology in the general stress response of Escherichia coli. Curr Opin
Microbiol 2, 148–152.
Hengge-Aronis, R. (2000). A role for the sS subunit of RNA
polymerase in the regulation of bacterial virulence. Adv Exp Med Biol
485, 85–93.
Hengge-Aronis, R. (2002). Signal transduction and regulatory
mechanisms involved in control of the sS (RpoS) subunit of RNA
polymerase. Microbiol Mol Biol Rev 66, 373–395.
ACKNOWLEDGEMENTS
Janga, S. C. & Moreno-Hagelsieb, G. (2004). Conservation of
We thank Hilda Solano-Lira for drawing Fig. 1, illustrating the
complex network of the RpoS sigmulon. E. P.-R. was financed by a
1400
adjacency as evidence of paralogous operons. Nucleic Acids Res 32,
5392–5397.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
Microbiology 157
Genomic analysis of the RpoS sigmulon
Kehry, M. R. & Dahlquist, F. W. (1982). The methyl-accepting chemotaxis
Paget, M. S. & Helmann, J. D. (2003). The s70 family of sigma factors.
proteins of Escherichia coli. Identification of the multiple methylation sites
on methyl-accepting chemotaxis protein I. J Biol Chem 257, 10378–
10386.
Genome Biol 4, 203.
King, T., Seeto, S. & Ferenci, T. (2006). Genotype-by-environment
interactions influencing the emergence of rpoS mutations in
Escherichia coli populations. Genetics 172, 2071–2079.
Sierro, N., Makita, Y., de Hoon, M. & Nakai, K. (2008). DBTBS: a
database of transcriptional regulation in Bacillus subtilis containing
upstream intergenic conservation information. Nucleic Acids Res 36
(Database issue), D93–D96.
Tatusov, R. L., Koonin, E. V. & Lipman, D. J. (1997). A genomic
Liu, R. & Ochman, H. (2007). Stepwise formation of the bacterial
perspective on protein families. Science 278, 631–637.
flagellar system. Proc Natl Acad Sci U S A 104, 7116–7121.
Tramonti, A., De Canio, M., Bossa, F. & De Biase, D. (2003). Stability
Loewen, P. C., Hu, B., Strutinsky, J. & Sparling, R. (1998). Regulation
in the rpoS regulon of Escherichia coli. Can J Microbiol 44, 707–
717.
and oligomerization of recombinant GadX, a transcriptional activator
of the Escherichia coli glutamate decarboxylase system. Biochim
Biophys Acta 1647, 376–380.
Lozada-Chávez, I., Janga, S. C. & Collado-Vides, J. (2006). Bacterial
Völker, U., Maul, B. & Hecker, M. (1999). Expression of the sB-
regulatory networks are extremely flexible in evolution. Nucleic Acids
Res 34, 3434–3445.
dependent general stress regulon confers multiple stress resistance in
Bacillus subtilis. J Bacteriol 181, 3942–3948.
Martı́nez-Garcı́a, E., Tormo, A. & Navarro-Llorens, J. M. (2001).
Weber, H., Polen, T., Heuveling, J., Wendisch, V. F. & Hengge, R.
(2005). Genome-wide analysis of the general stress response network
in Escherichia coli: sS-dependent genes, promoters, and sigma factor
Further studies on RpoS in enterobacteria: identification of rpoS in
Enterobacter cloacae and Kluyvera cryocrescens. Arch Microbiol 175,
395–404.
selectivity. J Bacteriol 187, 1591–1603.
Masuda, N. & Church, G. M. (2003). Regulatory network of acid
Weitzman, P. D. (1981). Unity and diversity in some bacterial citric
resistance genes in Escherichia coli. Mol Microbiol 48, 699–712.
acid-cycle enzymes. Adv Microb Physiol 22, 185–244.
Monsieurs, P., De Keersmaecker, S., Navarre, W. W., Bader, M. W., De
Smet, F., McClelland, M., Fang, F. C., De Moor, B., Vanderleyden, J. &
Marchal, K. (2005). Comparison of the PhoPQ regulon in Escherichia
Woodsmall, R. M. & Benson, D. A. (1993). Information resources at
coli and Salmonella typhimurium. J Mol Evol 60, 462–474.
Núñez, C., Esteve-Núñez, A., Giometti, C., Tollaksen, S., Khare, T.,
Lin, W., Lovley, D. R. & Methé, B. A. (2006). DNA microarray and
proteomic analyses of the RpoS regulon in Geobacter sulfurreducens.
J Bacteriol 188, 2792–2800.
http://mic.sgmjournals.org
the National Center for Biotechnology Information. Bull Med Libr
Assoc 81, 282–284.
Zambrano, M. M., Siegele, D. A., Almirón, M., Tormo, A. & Kolter, R.
(1993). Microbial competition: Escherichia coli mutants that take over
stationary phase cultures. Science 259, 1757–1760.
Edited by: D. W. Ussery
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 02 Aug 2017 15:40:27
1401