RNAs: Regulators of bacterial virulence

Abstract

RNA-based pathways that regulate protein expression are much more widespread than previously thought. Regulatory RNAs, including 5' and 3' untranslated regions next to the coding sequence, cis-acting antisense RNAs and trans-acting small non-coding RNAs, are effective regulatory molecules that can influence protein expression and function in response to external cues such as temperature, pH and levels of metabolites. This Review discusses the mechanisms by which these regulatory RNAs, together with accessory proteins such as RNases, control the fate of mRNAs and proteins and how this regulation influences virulence in pathogenic bacteria.

Full text

To survive and thrive in an often hostile environment,
a bacterium has to monitor its surroundings and adjust
its gene expression and physiology accordingly. This
is especially important for pathogenic bacteria, which
continuously interact with the host during an infection.
RNAs are excellent regulatory molecules that can carry
out a plethora of regulatory tasks1 (FIG. 1). For instance,
RNAs can directly sense environmental cues, such as
differences in temperature, pH and nutrient availabil-
ity, through regulatory regions that lie upstream of the
coding sequence on the same transcript, leading to an
altered transcriptional read-through or translation initia-
tion of that downstream coding sequence. Furthermore,
bacteria can regulate transcript expression through cis-
acting RNAs, which function in an antisense manner
and control the expression of mRNAs encoded on the
opposite DNA strand, or trans-acting small non-coding
RNAs (sRNAs), which function at a distance to alter the
expression of target RNAs through an antisense mecha-
nism. The fate of RNA transcripts can also be control-
led by proteins, including RNases and RNA chaperones,
which can degrade both cis- and trans-acting antisense
regulatory RNAs and their target mRNAs or can facili-
tate the interaction of trans-acting sRNAs with the target
mRNAs, respectively.
Because pathogenic bacteria encounter diverse envi-
ronmental conditions, they require rapid regulatory cir-
cuits to survive, making regulatory RNAs particularly
suitable for controlling bacterial virulence. Together with
regulatory proteins and two-component systems, regula-
tory RNAs integrate environmental stimuli into outputs
that are important for pathogenicity. In fact, there is evi-
dence that regulatory RNAs are more suitable than pro-
teins for controlling gene expression. First, the energy
cost of transcription (that is, generating regulatory and
target RNAs) is much lower than that of translating reg-
ulatory proteins. Second, regulatory RNAs can control
gene expression much faster than regulatory proteins;
this is especially true for 5′ untranslated regions (UTRs),
which directly dictate the expression of downstream
mRNAs on sensing an environmental cue. Third, regula-
tory RNAs are generally much less stable than regulatory
proteins, which allows their rapid clearance when they
are no longer needed. Fourth, many regulatory RNAs
act at the post-transcriptional level and can therefore
modify mRNAs that have already been expressed; they
can thereby dictate and possibly overcome effects at the
transcriptional level.
Studies in the past few years have shown that regula-
tion mediated by RNAs and their associated proteins is
more common than previously thought. Genome-wide
analyses based on tiling arrays and high-throughput
RNA-sequencing technologies (BOXES 1,2) have revealed
the transcriptomes of several bacteria, including patho-
genic bacteria2. Both techniques allow the unbiased
visualization of all RNA molecules transcribed during
specific conditions (for example, during stress, high
osmolarity and low oxygen) without taking into account
the positions of annotated ORFs. Therefore, the posi-
tions of all RNAs transcribed in a cell — namely, mRNAs
(including operons), tRNAs, ribosomal RNAs, cis-
acting antisense RNAs and trans-acting sRNAs — can be
mapped with 1-nucleotide resolution. These studies have
shown that the transcriptional landscape of all organisms
is much more complex than expected.
This Review focuses on 5′ and 3′ UTRs, cis-acting
antisense RNAs and trans-acting sRNAs, as well as on
proteins that affect the expression of virulence genes and
*Department of Molecular
Biology and Laboratory
for Molecular Infection
Medicine Sweden (MIMS),
Umeå University,
90187 Umeå, Sweden.
‡Instituto de Agrobiotecnología,
Universidad Pública de
Navarra–CSIC–Gobierno de
Navarra, 31006 Pamplona,
Spain.
Correspondence to J.J.
e-mail: jorgen.johansson@
molbiol.umu.se
doi:10.1038/nrmicro2457
RNAs: regulators of bacterial virulence
Jonas Gripenland*, Sakura Netterling*, Edmund Loh*, Teresa Tiensuu*,
Alejandro Toledo-Arana‡ and Jörgen Johansson*
Abstract | RNA-based pathways that regulate protein expression are much more widespread
than previously thought. Regulatory RNAs, including 5′ and 3′ untranslated regions next to
the coding sequence, cis-acting antisense RNAs and trans-acting small non-coding RNAs,
are effective regulatory molecules that can influence protein expression and function in
response to external cues such as temperature, pH and levels of metabolites. This Review
discusses the mechanisms by which these regulatory RNAs, together with accessory proteins
such as RNases, control the fate of mRNAs and proteins and how this regulation influences
virulence in pathogenic bacteria.
REVIEWS
NATURE R EVIEWS
|
MICROBIOLOGY VOLUME 8
|
DECEMBER 2010
|
857
© 20 Macmillan Publishers Limited. All rights reserved10

Nature Reviews | Microbiology
Gene X
DNA
SD
SD
5′ UTR
Trans-acting sRNA
Cis-acting
antisense RNA
3′ UTR
6S RNA
σ70
5′3′
5′
3′
mRNA
RNase
RNase
RNase
Riboswitch
An mRNA control element
that changes conformation in
response to the binding of
a metabolite (for example,
glycine, lysine or coenzyme
B12) and influences gene
expression.
Shine–Dalgarno sequence
A sequence that is located 5′
of the AUG (start) codon on
bacterial mRNAs and functions
as the binding motif of the 30S
subunit of the ribosome. The
consensus sequence is
AGGAGG.
Aptamer
An RNA domain, either
engineered or natural,
that forms a precise
three-dimensional structure
and selectively binds a target
molecule.
therefore promote virulence. For scientific interest, one
example of ‘non-pathogenic’ RNA regulation (a lysine
riboswitch) is also discussed.
5′ regulatory UTRs of mRNAs
The region between the transcriptional start site and
the start codon of an mRNA is known as the 5′ UTR.
This region can vary substantially in length, ranging
from only a few to several hundred bases. Transcription
can begin at various promoters, allowing the formation
of many potential 5′ UTRs and, hence, complex possi-
bilities of post-transcriptional regulation3,4. 5′ UTRs are
used by pathogenic bacteria, among other organisms,
to modify gene expression on the basis of changes in
temperature, pH and the presence of metabolites.
Temperature control. Bacteria have developed thermo-
sensor mechanisms that act at the protein, DNA and
RNA levels to directly detect changes in temperature,
although most act at the RNA level5–7. Such sensing
mechanisms are especially important for pathogens,
which need to fine-tune gene expression in response to
host temperature.
The food-borne human pathogen Listeria monocy-
togenes, which causes various brain and maternofetal
infections, possesses such an RNA thermosensor. The
116-nucleotide 5′ UTR upstream of the prfA mRNA
coding sequence forms a secondary structure at lower
temperatures, masking the Shine–Dalgarno sequence (SD
sequence) and thereby inhibiting translation. An alterna-
tive secondary structure is formed at human body tem-
perature (37 °C), exposing the SD region and allowing
translation of the ORF, which encodes the transcrip-
tional activator listeriolysin regulatory protein (PrfA)8.
Once present, PrfA, which is essential for L. mono-
cytogenes virulence, activates the expression of virulence
genes encoding adhesins, phagosome-escaping factors
and immune-modulating factors9,10. For instance, PrfA
activates the expression of hly (encoding listeriolysin O,
which is essential for bacterial escape from the phago-
some) and actA (encoding actin assembly-inducing
protein (ActA), which is essential for listerial intracellular
movement).
Thermosensors that function similarly to prfA have
been identified in other pathogenic bacteria, such as
Yersinia spp. and Salmonella spp.11,12. Yersinia pestis,
the causative agent of plague, has a thermosensor that
controls the expression of the transcriptional virulence
regulator LcrF, ensuring its expression only at 37 °C.
LcrF is required for Y. pe s t i s virulence13 because it acti-
vates the expression of YopE, which blocks phagocyto-
sis14,15. lcrF and other thermosensors of the four-uracil
family (but not prfA) are thought to achieve this con-
trol through their anti-SD sequence (made up of four
uracils)16. Specifically, at low temperatures the anti-SD
sequence sequesters the SD sequence, whereas at human
body temperature the SD site dissociates from the anti-
SD sequence, allowing the ribosome to bind to the SD
sequence and translation to progress.
pH sensing. The alx gene in Escherichia coli encodes a
putative transporter implicated in resistance against the
antimicrobial agent tellurite17 and is expressed in highly
alkaline conditions18. The 5′ UTR in front of the alx
mRNA has been shown to be a pH-responsive element17.
Under normal growth conditions (pH 7.0), transcrip-
tion of the 5′ UTR proceeds uninterrupted, leading to
the formation of a translationally inactive structure. By
contrast, under alkaline conditions an alternative struc-
ture that is translationally active is formed; in this case,
the RNA polymerase pauses at two different sites in the
5′ UTR, preventing the formation of the inactive struc-
ture. Instead, an open structure is formed, exposing the
SD sequence and allowing binding of the ribosome and,
hence, translation. It is possible that the basic regulatory
mechanism controlling alx expression (that is, a differ-
ence in transcription speed generating alternative RNA
structures) is used to regulate the expression of other
genes as well.
Metabolite sensing. One group of 5′ UTRs is the ribo-
switches19–22. These are metabolite-sensing regulatory
RNA structures that function as sensors and regula-
tors of various metabolic pathways in bacteria. Each
class of riboswitch binds a specific metabolite through
its aptamer domain, and this interaction induces a struc-
tural change in the riboswitch regulatory domain that
causes altered transcription or translation. For ribo-
switches acting at the transcriptional level, binding of
the metabolite to the aptamer domain generally induces
the formation of a terminator structure in the regula-
tory domain to prevent transcription of the downstream
gene. For riboswitches acting at the translational level,
Figure 1 | Control of mRNA activity and stability. The fate of an mRNA is controlled
by several factors. 5′ untranslated regions (UTRs) lie upstream of the coding sequences
and include thermosensors, pH sensors and riboswitches. 3′ UTRs lie at the other end
of coding RNAs and, in many cases, determine transcription termination and stability.
Cis-acting antisense RNAs are expressed from the opposite strand of the DNA.
Trans-acting small non-coding RNAs (sRNAs) are expressed from a different location on
the chromosome and most commonly bind to the Shine–Dalgarno (SD) sequence of a
target mRNA in an antisense manner. Trans-acting sRNAs can also sequester target
proteins; for example, 6S RNA sequesters the housekeeping RNA polymerase σ70 (also
known as RpoD). The fate of transcripts can also be controlled by RNases, which function
either exonucleolytically (orange), by targeting transcripts from the 5′ or 3′ end (such
as RNase J1 and RNase J2, which are found in some Gram-positive bacteria) or
endonucleolytically (purple), by targeting sequences in the middle of the transcript.
REVIEWS
858
|
DECEMBER 2010
|
VOLUME 8 www.nature.com/reviews/micro
© 20 Macmillan Publishers Limited. All rights reserved10

Nature Reviews | Microbiology
RNA isolation
mRNA enrichment
Amplify and sequence
Analyse gene expression
Binding of cDNA to beads
Reverse transcribe
fragments (cDNA)
Cyclic-di-GMP
A second messenger that is
generated by diguanylate
cyclases and hydrolysed by
phosphodiesterase A.
Rho-independent
transcriptional terminator
A strong secondary RNA
structure followed by several
uracils that destabilizes the
RNA–DNA duplex so that
the RNA polymerase falls off.
Normally found after the
coding sequence of an mRNA.
binding of the metabolite to the aptamer domain
alters the interaction between the SD and an anti-SD
sequence, either preventing or allowing binding of the
ribosome to the SD sequence19.
Recently, one riboswitch class was identified that
controls the expression of downstream genes — includ-
ing genes important for DNA uptake, motility and viru-
lence — in many pathogenic bacteria (Vibrio cholerae,
the causative agent of cholera, Clostridium difficile, an
intestinal opportunistic pathogen, and Bacillus cereus,
a major cause of food poisoning) by binding the sec-
ond messenger cyclic-di-GMP23. Binding of c-di-GMP
to the riboswitch aptamer domain induces a structural
alteration that can affect transcription termination or
translation initiation, depending on the bacterial spe-
cies. Interestingly, different riboswitches in the same
bacterial species might respond differently to c-di-GMP,
either stimulating or repressing expression of the
downstream gene.
One riboswitch that is negatively regulated by
c-di-GMP lies in front of gbpA, which encodes
N-acetylglucosamine-binding protein A (GbpA), a pro-
tein important for V. cholerae intestinal attachment24.
GbpA is expressed only under specific growth condi-
tions, including when V. cholerae enters the intestine.
This is at least partly because the concentration of
c-di-GMP drops in the intestine, and this is sensed by
the riboswitch, thereby allowing the expression of GbpA
and the consequent attachment of the bacteria to the
intestine25.
3′ regulatory UTRs of mRNAs
In eukaryotes, regulatory UTRs located at the 3′ end of
the transcript are known to control translation26,27. The
3′ UTRs of bacterial mRNAs are thought to mainly har-
bour transcription termination structures, which might
prevent access of exonucleases to the 3′ end of the tran-
script but have no clear regulatory function. However,
the recent advances in bacterial transcriptome analyses
(BOXES 1,2) have indicated regulatory roles for 3′ UTRs,
albeit not specifically with regard to virulence.
One such example is the lysine riboswitch of L. mono-
cytogenes4 (FIG. 2), which lies in the 5′ UTR, similarly to
other bacterial riboswitches. The lysine riboswitch regu-
lates the expression of the downstream gene, lmo0798,
which encodes a lysine transporter. In the presence of
lysine, an intrinsic Rho-independent transcriptional terminator
is formed and the downstream gene is not expressed. By
contrast, when lysine is absent, a structure that blocks
termination (an anti-terminator structure) is formed and
the downstream gene is expressed (FIG. 2). Interestingly,
in addition to this well-known type of riboswitch regu-
lation, the lysine riboswitch can also control the termi-
nation of the upstream gene, which encodes a putative
blue-light receptor. The mRNA encoding the upstream
gene harbours a long 3′ UTR containing the lysine
ribo switch but no other terminator. The presence of
lysine terminates transcription, whereas the absence
of lysine allows co-expression of the upstream and
downstream gene on the same transcript (FIG. 2).
Other RNA structures located in the 3′ UTR with
putative regulatory functions have been identified in
Bacillus subtilis28. Specifically, the transcripts of nine
genes were found to have long 3′ UTRs that have high
sequence similarity and form stable RNA secondary
structures. Because most of these genes encode proteins
that associate with the membrane and are involved in
cell wall synthesis and transport, the 3′ UTR structures
might be involved in the translation and/or localiza-
tion of the proteins to specific compartments in the
cell. Alternatively, the long 3′ UTRs might protect from
3′-to-5′ exonucleolytic degradation through their sec-
ondary structures. In addition to these structures, long
putative 3′ regulatory UTRs have been observed next
to transcripts encoding certain virulence factors in the
human pathogen Staphylococcus aureus29. This bacte-
rium has several long 3′ UTRs that give rise to sRNAs30.
These examples suggest that large bacterial 3′ UTRs
might have a role in mRNA stability, mRNA localiza-
tion, the generation of sRNAs or regulation by binding
trans-acting factors.
Cis-acting antisense RNAs
Cis-acting antisense RNAs consist of two subtypes (FIG. 3):
bona fide antisense RNAs, which are present on the com-
plementary strand to one or several ORFs; or overlapping
UTRs, which consist of a long 5′ or 3′ UTR of an mRNA
Box 1 | RNA sequencing
Isolated total RNA is enriched for mRNA
(see the figure; green) through the
removal of ribosomal RNA (rRNA) and
tRNA (black) using several techniques.
Capture of rRNA involves the use of
probes that bind 16S and 23S rRNA.
Processed RNA is removed by 5′-to-3′
exonucleases that recognize
monophosphorylated 5′ ends: 16S and
23S rRNAs are processed and thus
carry a monophosphorylated 5′ end,
so these RNAs are degraded, whereas
primary transcripts usually carry a
triphosphorylated 5′ end. Samples can
also be enriched for mRNA though the
addition of an artificial poly(A) tail to
mRNAs followed by fishing with an
oligo(dT). The mRNA is converted into
cDNA (red) by reverse transcription.
The lack of natural poly(A) tails enforces
alternative priming approaches, such
as random hexamer priming, addition
of artificial poly(A) tails followed by
priming with oligo(dT), or priming from
RNA probes ligated to the mRNAs. The
cDNA is amplified (normally when
attached to beads) and sequenced by
one of many different platforms (such as
454, Ilumina and Solid). Normally, each
sequence read is 30–400 bases, and this
is mapped on the reference genome. The
number of reads for each region gives a
direct measurement of RNA expression
and stability. The total number of
sequence reads can reach 1 × 10 9.
REVIEWS
NATURE R EVIEWS
|
MICROBIOLOGY VOLUME 8
|
DECEMBER 2010
|
859
© 20 Macmillan Publishers Limited. All rights reserved10

Nature Reviews | Microbiology
RNA isolation
Conversion to cDNA
Analyse gene expression
Hybridize to microarray
5′3′
Genomic DNA
Tiling array
Partially overlapping
probes
Response regulator
A bacterial gene-regulatory
protein that controls gene
expression in response to
external signals. Most response
regulators consist of two
domains: a regulatory domain,
the activity of which is
modulated (indirectly) by
the external signal, and a
DNA-binding domain.
that overlaps with the mRNA encoded by the other DNA
strand. For example, an overlapping antisense 5′ UTR
can be expressed from a promoter located on the oppo-
site strand of an ORF. A long 3′ UTR, generated either by
the absence of a transcriptional terminator near the stop
codon or by termination read-through events, can lead
to overlapping antisense 3′ UTRs, as the transcription of
that gene would end in a position located inside or after
the gene encoded on the opposite strand2 (FIG. 3).
Until a few years ago, our understanding of cis-
acting antisense RNAs was limited to studies in bac-
teriophages, plasmids and transposons. Although the
number of identified cis-acting antisense RNAs in bac-
teria increased soon after, those identified were limited
to sRNAs that acted on regions near the SD, owing to
the biased nature of the methodologies used7,19. Recent
technological advancements (BOXES 1,2) have shown that
antisense transcription occurs in all species, including
bacteria. Indeed, thousands of transcripts that origi-
nate from antisense RNAs to genes or from intergenic
regions that were previously thought to be silent have
been identified in eukaryotes. This unanticipated level
of complexity is known as ‘pervasive’ transcription, as
transcripts are not restricted to well-defined features
such as ORFs31,32.
Pervasive transcription has also been found in bac-
teria. For example, a study examining the primary tran-
scriptome of the gastric pathogen Helicobacter pylori
found that around 46% of all ORFs were associated with
at least one antisense transcription start site33. Similarly,
more than 1,000 antisense transcription start sites have
been mapped in E. coli34, supporting previous results35.
Many antisense transcripts have also been found in
Mycoplasma pneumoniae, Pseudomonas syringae,
L. monocytogenes, B. subtilis and the cyanobacterium
Synechocystis sp. PCC 6803 (REFS 4,28,36–38). It there-
fore seems that cis-acting antisense RNAs are extremely
abundant. Interestingly, the size of the antisense tran-
scripts found in a cell can sometimes range from a few
bases to several kilobases, so one particular antisense RNA
may overlap several genes, as shown in L. monocytogenes
and B. subtilis4,28.
Although the mechanisms involved in pervasive
antisense transcription are still unclear, some insights
have been obtained33,34,36,37,39–41. For example, the
1.2 kb antisense RNA AmgR was shown to regulate
Salmonella enterica virulence. AmgR is complemen-
tary to the mgtC portion of the polycistronic mgtCBR
operon, which encodes MgtC, MgtB and MgtR. MgtC
is an inner-membrane protein that is present in several
bacterial pathogens, in which it is required for survival
in macrophages, virulence in mice and growth at low
Mg2+ concentrations. A variation in the expression of
AmgR promotes changes in MgtC protein levels, thereby
affecting virulence. Inactivation of the amgR promoter
derepresses the expression of MgtC, which makes this
S. enterica mutant more virulent than the wild-type
strain; by contrast, overexpression of AmgR attenuates
S. enterica virulence owing to a decrease in MgtC levels.
Thus, S. enterica modulates its proliferation inside host
tissues by regulating MgtC levels through the action of
the AmgR cis-acting antisense RNA42. Interestingly, the
expression of both the antisense AmgR and the mgtCBR
operon is controlled by the same response regulator, PhoP,
under low Mg2+ concentrations. However, PhoP binds
the amgR promoter with less affinity than it binds the
mgtC promoter; this results in a regulatory loop in which
low levels of active PhoP (caused by a slight increase in
Mg2+ concentrations) induce mgtC expression, whereas
high levels of active PhoP (caused by a high concentra-
tion of Mg2+) induce AmgR expression, leading to mgtC
inactivation42. Thus, the antisense RNA might act as
a ‘timing device’, allowing a dynamic change in target
mRNA levels in a PhoP-dependent manner.
Similarly, in H. pylori several cis-acting antisense
RNAs and their corresponding mRNA targets (encod-
ing proteins required for acid resistance) are induced
by the same signal (in this case, acid stress)33. This is
in contrast to most trans-acting sRNAs, which are typi-
cally controlled by a different regulator than their mRNA
target19 (see below).
Another example of a cis-acting antisense RNA affect-
ing a crucial bacterial process is the L. mono cytogenes
1.7 kb 5′ UTR of the mRNA encoding MogR4, the nega-
tive regulator of flagellum biosynthesis. mogR is tran-
scribed from two alternative promoters, P1 and P2, which
are located at 1,697 and 45 nucleotides upstream of the
start codon, respectively. Consequently, P1 generates a
long 5′ UTR that overlaps the first three genes (lmo0675,
lmo0676 and lmo0677) of the operon required for flag-
ellum biosynthesis, which is encoded on the opposite
strand. The transcription of this long overlapping 5′ UTR
depends on the stress-activated transcriptional regulator
RNA polymerase σB, and it is therefore overexpressed
Box 2 | Tiling array
Each tiling array in bacteria
usually contains >105
oligonucleotides. Total RNA
(black and green) is isolated and
converted into labelled cDNA
(red) before hybridization, and
the array is then read under a
fluorescence microscope. RNA
might also be spotted directly
onto the array and visualized
by antibodies recognizing
DNA–RNA duplexes. The
signal strength of each
oligonucleotide reflects
the relative amount of that
particular mRNA in each of
the tested samples. In newly
designed tiling arrays, the
entire genome is covered by
overlapping oligonucleotides,
allowing non-coding regions
of the chromosome (such as
intergenic regions and antisense
strands) to be read.
REVIEWS
860
|
DECEMBER 2010
|
VOLUME 8 www.nature.com/reviews/micro
© 20 Macmillan Publishers Limited. All rights reserved10

Nature Reviews | Microbiolog
y
Imo0799LysRS Imo0798
Lys
3′5′
5′3′
DNA
Anti-terminator
– Lysine
5′3′
Lys
Terminator
+ Lysine
mRNA
PPT
3′5′
Transcriptional interference
The negative impact that one
transcriptional activity can
have on another transcriptional
activity in cis.
Quorum sensing
The phenomenon in which the
accumulation of signalling
molecules enables a single
cell to sense the number of
bacteria that are present (the
cell density); the purpose is to
coordinate certain behaviours
or actions between bacteria.
during the stationary growth phase. Absence of σB
increases bacterial motility, whereas overexpression of
the P1-derived, σB-dependent transcript decreases motil-
ity. Although the precise mechanism is not known, it is
expected that bacteria modulate the expression of flag-
ellum proteins by integrating environmental signals in
a fast manner. Thus, the RNA levels of the flagellum
operon can be controlled by post-transcriptional RNA
processing that depends on the levels of antisense RNA,
which are in turn controlled by σB.
In general, it is conceivable that only mRNA that
is present at higher levels than its cis-acting antisense
antagonist is translated, so translation of certain genes
starts only when the mRNA concentration reaches a
certain level2. Furthermore, it cannot be excluded that
transcriptional regulation of mRNA might begin while
the cis-acting antisense RNA is being transcribed and,
consequently, transcriptional interference might occur
between both elongation complexes43. Because antisense
transcription has been shown to be a genome-wide phe-
nomenon, further investigations are needed to deter-
mine whether the main function of cis-acting antisense
RNAs is to control gene expression at the transcriptional
level (by transcriptional interference) or at the post-
transcriptional level (by mRNA processing or translational
inhibition), or whether both regulatory mechanisms
can coexist.
Trans-acting sRNAs
RNA molecules acting in trans on distant targets are
commonly denoted as trans-acting sRNAs and are per-
haps the best-studied form of regulatory RNA19. Tran s -
acting sRNAs, traditionally identified in intergenic
regions, are encoded distally from their targets and func-
tion either by binding RNAs, leading to downregulation
of target mRNA activity, typically through degradation,
or by binding target proteins and affecting their activity.
Most trans-acting sRNAs are involved in responding to
rapidly changing environmental conditions, and only a
few have been shown to have roles in different infection
models4,44–51.
One example of a trans -acting sRNA with a role
in virulence has been identified in the Gram-positive
opportunistic human pathogen S. aureus. S. aureus har-
bours the accessory gene regulator (agr) locus, which
encodes an autoactivating quorum sensing system52
and is composed of two transcriptional units that are
transcribed from the divergent P2 and P3 promoters.
Expression of the four genes encoding the quorum sens-
ing two-component system (agrBDCA) is initiated at P2,
whereas expression from P3 results in a transcript that
can act as a trans-acting 514-nucleotide sRNA, known
as RNAIII53, and can encode δ-haemolysin. RNAIII is
structurally conserved among staphylococcal species
and regulates several mRNAs involved in the patho-
genicity of S. aureus. The expression of RNAIII peaks
at late logarithmic and stationary phases and is upregu-
lated by AgrA, the response regulator of the agr locus52.
Because it can function as both an activator and a
repressor, RNAIII can control the switch between the
expression of secreted factors and the inhibition of sur-
face proteins53. For example, RNAIII regulates the expres-
sion of α-haemolysin (hla; also known as hly) mRNA54
by binding, through its 5′ domain, to a region upstream
of the hla coding sequence that normally sequesters the
hla SD sequence and inhibits translation; the interac-
tion releases the SD sequence of hla, and translation
can begin55.
Furthermore, RNAIII negatively regulates the early-
expressed virulence factors immunoglobulin G-binding
protein A and fibrinogen-binding protein56,57 and the
transcriptional regulator Rot58. In this case RNAIII
binds the target RNA through its 3′ UTR and central
domain, thereby inhibiting translation and inducing
target mRNA degradation. Finally, RNAIII has been
found to regulate coa mRNA, which encodes staphylo-
coagulase, another early-expressed virulence factor
that causes coagulation of human plasma and enables
S. aureus to hide from the host immune system59. Two
hairpins of RNAIII can directly interact with the coa
Figure 2 | Untranslated region-mediated regulation. In the presence of lysine, an
intrinsic Rho-independent terminator (T) is formed in the lysine riboswitch (LysRS) mRNA
sequence, preventing transcription of the downstream gene, lmo0798. In this case, a
small non-coding transcript is generated that corresponds to LysRS. The LysRS
terminator also functions as a terminator for the upstream gene, lmo0799, generating a
transcript that consists of the lmo0799 sequence and LysRS. In the absence of lysine, the
intrinsic terminator is not formed, allowing transcription to proceed into the downstream
gene. Because there is no termination of lmo0799 expression in the absence of lysine, a
long transcript comprising lmo0799, LysRS and lmo0798 is also generated from the
lmo0799 promoter. P, promoter.
REVIEWS
NATURE R EVIEWS
|
MICROBIOLOGY VOLUME 8
|
DECEMBER 2010
|
861
© 20 Macmillan Publishers Limited. All rights reserved10

ORF1 ORF2
RNA
Small antisense RNA
Long antisense RNA
Antisense RNAs
Overlapping UTRs
ORF3 ORF4
ORF9 ORF10ORF11
ORF6 ORF7
Overlapping 5′ UTRs
Overlapping 3′ UTRs
ORF8
ORF5
ORF12
Nature Reviews | Microbiology
Promoter
Terminator
SD sequence and with parts of the coding sequence,
thereby inhibiting translation initiation and promoting
RNase III-dependent degradation of the coa transcript
together with RNAIII60.
In L. monocytogenes, approximately 50 trans-acting
sRNAs have been discovered4,61–64, but functional and
mechanistic data are still limited. Two trans-acting
sRNAs, RliB and Rli38 (which are absent in non-
pathogenic strains of L. monocytogenes), were recently
shown to have a role in L. monocytogenes pathogenic-
ity4. RliB is 360 nucleotides long and contains five
repeats of 29 nucleotides interspaced by 35–36 nucle-
otides62. Rli38 is 369 nucleotides long and has homology
to putative sRNAs in other pathogenic bacteria, such
as S. aureus and Enterococcus faecalis. Interestingly,
Rli38 is differentially expressed during various envi-
ronmental stress conditions and is markedly induced
when bacteria are exposed to human blood4. Infection
experiments using rliB- or rli38 -knockout bacteria
showed that the rli38-knockout strain had a lower
ability to colonize various mouse organs than the wild-
type strain. Surprisingly, the rliB-knockout strain had
a higher coloni zation ability than the wild-type strain;
these findings highlight the complexity of the regula-
tion exerted by these trans-acting sRNAs with regard
to L. monocytogenes virulence.
The complexity of RNA regulation was further
increased by the recent finding that riboswitches can
have dual regulatory functions in L. monocytogenes. A
S-adenosylmethionine (SAM) riboswitch can function
both as a classical cis-acting riboswitch, controlling the
expression of its downstream gene by transcriptional
termination, and as a trans-acting sRNA, directly bind-
ing to a target mRNA65. On binding to its metabolite
(SAM), the riboswitch adopts a terminator structure,
terminating its transcription and generating the trans-
acting sRNA. This transcriptionally terminated SAM
riboswitch, termed SreA (SAM riboswitch element A;
227 nucleotides), directly base-pairs with the 5′ UTR of
the prfA transcript (the distal part of the RNA thermo-
sensor), thereby blocking its translation (FIG. 4a). This
interaction occurs only at high temperatures (~37 °C),
when the thermosensor has a more open conforma-
tion, and in nutrient-rich conditions when the levels
of SreA are high (that is, when SAM levels are high).
Intriguingly, the expression of SreA RNA is also control-
led by PrfA: when PrfA levels are high, PrfA switches
off its own expression, thus downregulating SreA. In a
sreA mutant background, the virulence gene hly (which
is regulated by PrfA, as mentioned above) is upregulated
because PrfA levels increase, suggesting that SreA antag-
onizes L. monocytogenes virulence and therefore inhibits
pathogenesis. It therefore seems that SreA functions as
a sensor of the metabolic state in bacteria, preventing
the expression of PrfA if conditions are rich in the host
cytoplasm (when SAM levels are high and SreA is ter-
minated). The observation that terminated riboswitches
can have a function by themselves markedly increases
the number of putative trans-acting sRNAs.
Six trans-acting sRNAs were recently discovered in
the Gram-negative opportunistic pathogen Legionella
pneumophila, and among these was a homologue of the
well-studied E. coli 6S RNA50. In E. coli, 6S RNA associates
with and sequesters the RNA polymerase holoenzyme
containing σ70 (also known as RpoD; the housekeeping
sigma factor that facilitates the binding of RNA polymer-
ases to specific promoter sequences) but not σS (also
known a s σ38 or RpoS; the stationary phase sigma factor).
The 6S RNA–σ70 interaction leads to increased transcrip-
tion of σS-dependent genes66. Because 6S RNA expression
is increased at stationary phase, this allows the bacte-
rium to switch from σ70-dependent to σS-dependent gene
expression. Two sizes of 6S RNA have been identified
in L. pneumophila, one of 182 nucleotides and another
of 147 nucleotides. Interestingly, only the shorter,
3′-processed 6S RNA could co-immunoprecipitate with
σ70. During infection of THP-1 macrophage-like cells,
a strain with mutated 6S RNA (ssrS) was attenuated,
indicating the importance of 6S RNA during L. pneu-
mophila replication inside human cells. The 6S RNA
controls the expression of ~5% of L. pneumophila genes;
among the positively regulated genes are vipA and
legC5, which encode effector molecules exported by the
type IVB secretion system50. Data suggest that 6S RNA
in L. pneumophila could bind other, as-yet unknown,
targets, the expression of which depends on σ70 RNA
polymerases50.
Figure 3 | Cis-acting antisense RNAs. Schematic representation of the different types
of antisense RNA molecules. These include bona fide antisense RNAs and overlapping
5′ and 3′ untranslated regions (UTRs). The antisense RNA may be either a long antisense
RNA covering more than one ORF (in the example, the long antisense RNA overlaps
ORF1, ORF2 and ORF3) or a small antisense RNA that overlaps the Shine–Dalgarno
sequence (which lies between the promoter and the start codon) and affects mRNA
stability and/or protein translation (for example, the antisense RNA overlapping ORF4).
Overlapping 5′ UTRs are generated when the transcription of a certain gene (for example,
ORF5) starts from a promoter located on the DNA strand opposite divergent genes (in
this case, ORF6, ORF7 and ORF8). As a result, the ORF5 mRNA has a long 5′ UTR that
overlaps the ORF6, ORF7 and ORF8 mRNA. Overlapping 3′ UTRs are generated when
transcription of a certain gene ends in or after a gene located on the opposite DNA
strand; for example, the transcription of the operon encoding ORF9, ORF10 and ORF11
ends after ORF12, so the long 3′ UTR completely overlaps the ORF12 mRNA.
REVIEWS
862
|
DECEMBER 2010
|
VOLUME 8 www.nature.com/reviews/micro
© 20 Macmillan Publishers Limited. All rights reserved10

Nature Reviews | Microbiology
37 °C
+ SreA
Closed
a
b
Open
SreA
SreA
30 °C
prfA 5′ UTR
prfA
5′ UTR
Ribosome
VrrA
ompA
OmpA
Translation
inhibited
Vesicle
production
Virulence
Translation
inhibited
Translation
initiated
Translation
inhibited
3′
3′
3′
3′
5′
5′
5′
5′3′5′
3′
3′
5′
5′
SD
Degradosome
A complex of several proteins
involved in the degradation
and processing of various
transcripts in Gram-negative
bacteria.
So far, 36 trans-acting sRNAs have been identified in
V. cholerae67, at least nine of which are involved in quo-
rum sensing and biofilm formation68–70. The recently
discovered trans-acting sRNA VrrA (Vibrio regulatory
RNA of OmpA) is 140 nucleotides long and is conserved
among the Vibrio species48. Expression of VrrA is regu-
lated by σE, an alternative sigma factor involved in the
response to extracytoplasmic, high-temperature and oxi-
dative stress conditions. VrrA acts as a repressor of outer-
membrane protein A (OmpA) synthesis. On ultraviolet
irradiation, VrrA levels increase, repressing the expression
of OmpA by directly base-pairing with the ompA mRNA
and blocking its SD sequence. Indeed, an ompA mutant
showed increased survival during ultraviolet irradiation
compared with survival of the wild-type strain. This is
because reduction of OmpA levels leads to increased
production of outer-membrane vesicles (FIG. 4b), which
have been proposed to physically protect the bacterium
against ultraviolet light-induced damage. The absence of
VrrA increases V. cholerae virulence in mice through an
unknown mechanism, possibly owing to the increased
expression of OmpA and/or the reduced production of
outer-membrane vesicles. Another outer-membrane pro-
tein, OmpT, which is known to be involved in V. cholerae
bile resistance, has recently been identified as the second
target of VrrA71,72. VrrA directly base-pairs with the 5′
region of the ompT transcript, repressing its translation.
The interaction between VrrA and ompT depends on
Hfq, similarly to many other trans-acting RNA–target
mRNA interactions in E. coli and S. enterica71; interest-
ingly, however, the VrrA–ompA interaction does not
strictly depend on Hfq19,48.
Accessory proteins
Regulation by RNAs often requires an interaction with
accessory proteins. Below, we discuss the interactions of
these accessory proteins with the regulatory RNAs and
the mechanisms by which they exert their effects.
Hfq. If the mRNA is controlled by a trans-acting sRNA,
the RNA chaperone protein Hfq is often required to
facilitate a stable trans-acting sRNA–mRNA target
interaction1,19,73. Hfq is especially important if the level
of complementarity between the trans-acting sRNA and
the target mRNA is low. Hfq displays structural homol-
ogy to eukaryotic Sm proteins, which are involved in
RNA degradation and splicing, and acts as a hexamer,
forming a doughnut shape19. Recent data suggest that
distinct parts of Hfq bind the target mRNA and the
trans-acting sRNA74. However, the exact mechanism by
which Hfq functions in currently unclear.
Homologues of Hfq have been found in approxi-
mately half of the sequenced eubacteria examined to
date. Intriguingly, Hfq has been shown to participate
in virulence in many Gram-negative bacteria but is dis-
pensable for virulence in Gram-positive species75. One
exception is L. monocytogenes, in which a hfq strain
displays a lower bacterial count in mice than the wild-
type strain76. Hfq coordinates V. cholerae quorum sens-
ing by facilitating the antisense interaction between the
trans-acting sRNA Qrr (quorum regulatory RNA) and
the hapR mRNA, which encodes a major transcriptional
regulator, and this interaction allows virulence genes to
be expressed at low cell density77. Interestingly, in E. coli
Hfq was recently shown to interact more frequently with
the antisense strand of the protein-coding sequence
than with the sense strand, indicating that it is involved
in regulation mediated by both cis and trans antisense
RNAs78.
RNases. RNases are a class of enzymes that govern the
maturation and degradation of target mRNAs and trans-
acting sRNAs79. RNases are classified by their nature
of cleavage, which can be endonucleolytic (within the
transcript) or exonucleolytic (from the 5′ or, most com-
monly, the 3′ end of the transcript). RNases recognize
either single-stranded or structured double-stranded
RNA sequences. Many recent reviews80–84 have discussed
RNases, so this Review describes only a few RNases
involved in virulence.
RNase E is a central endonuclease in Gram-negative
bacteria that recognizes AU-rich segments as cleav-
age sites. It is also the major ribonuclease in the in vivo
degradosome, which also contains enolase (a glycolytic
enzyme), PNPase (polynucleotide phosphorylase) and
RhlB (an RNA helicase)81,85. The degradosome com-
ponents might vary depending on growth conditions,
Figure 4 | Trans-acting RNAs. a | Expression of listeriolysin regulatory protein (PrfA) in
Listeria monocytogenes is controlled by temperature and by the trans-acting riboswitch
S-adenosylmethionine (SAM) riboswitch element A (SreA; 227 nucleotides). The prfA
untranslated region (UTR) forms a closed stem loop structure at temperatures below
30 °C, masking the Shine–Dalgarno (SD) sequence. At 37 °C, the closed stem loop
structure is unstable, allowing the ribosome to bind to the SD and initiate translation.
SreA can function in trans by binding to the 5′ UTR of prfA. The SreA–prfA UTR
interaction leads to decreased expression of PrfA. b | The trans-acting small non-coding
RNA VrrA (Vibrio regulatory RNA of OmpA) negatively regulates the expression of
outer-membrane protein A (OmpA) in Vibrio cholerae. VrrA binds the ompA mRNA,
thereby blocking ribosome access to the SD sequence. The reduction of OmpA protein
levels leads to increased release of outer-membrane vesicles and decreased virulence in
an infant-mouse infection model.
REVIEWS
NATURE R EVIEWS
|
MICROBIOLOGY VOLUME 8
|
DECEMBER 2010
|
863
© 20 Macmillan Publishers Limited. All rights reserved10

and the activity of the degradosome is modulated by
regulator of ribosome activity A (RraA)86,87. RNase E
is thought to preferentially bind monophosphorylated
rather than triphosphorylated 5′ ends, although this
idea was recently challenged88–92. In most cases, RNase
E and Hfq act together to induce the degradation of
trans-acting sRNA–mRNA target complexes and have
been shown to interact directly with trans-acting sRNAs
in a complex that eventually degrades target mRNAs93.
However, RNase E has been shown to compete with Hfq
for some AU-rich single-stranded sequences, indicating
that Hfq can sometimes protect target mRNAs and trans-
acting sRNAs from endonucleolytic cleavage94. RNase
E also controls the degradation of the AmgR–mgtCBR
cis-acting antisense RNA–target mRNA complex in a
mechanism that is independent of RNase III and Hfq42.
RNase E-mediated processing is important for the
proper stoichiometric expression of different ORFs in
polycistronic mRNAs. One such example is observed
with the expression of the pyelonephritis-associated
pilus (composed of Pap proteins), which mediates
binding of uropathogenic E. coli to the kidney, causing
pyelonephritis95. The papBA bicistronic mRNA is proc-
essed by RNase E, which recognizes AU-rich sequences
between the coding RNAs96. After processing, the tran-
script encoding the regulatory PapB protein is rapidly
degraded, whereas papA, which encodes the major
structural component of the pilus, is expressed, resulting
in the formation of pili of the correct length.
Another example of an RNase with a role in viru-
lence is RNase III, which belongs to the family of
double-stranded RNA-specific endoribonucleases. It
cleaves phosphodiester bonds, creating 5′-phosphate
and 3′-hydroxyl termini with overhangs of 2 nucle-
otides97. Tra ns-acting sRNA-mediated gene regulation
has been shown to depend on RNase III activity in many
cases. One such example is S. aureus RNAIII (see above),
which requires RNase III to degrade the RNAIII–target
mRNA duplexes98.
Although most bacterial exoribonucleases process
RNA from the 3′ to the 5′ end, RNase J1 and RNase J2
in B. subtilis possess 5′-to-3′ exonuclease activity as well
as endonucleolytic activity99. RNase J1 and RNase J2
homologues are conserved in many Gram-positive
bacteria and seem to function similarly to the RNase E
of Gram-negative bacteria. As with RNase E, RNase J1
and RNase J2 are specific for AU-rich single-stranded
RNA segments and show preference for monophospho-
rylated 5′ ends100,101. Although RNase J1 and RNase J2
can work in a complex, only RNase J1 is essential for
growth in B. subtilis102; by contrast, both are essential for
growth in Streptococcus pyogenes103. Targets of RNase J1
and RNase J2 in S. pyogenes include the virulence fac-
tors hasA (which encodes hyaluronate synthase A,
involved in capsule synthesis), sagA (which encodes
streptolysin S) and streptodornase (sda; which encodes
a DNase)103.
Another endonucleolytic RNase in Gram-positive
bacteria is RNase Y (encoded by ymdA), which targets,
among other transcripts, SAM riboswitches in B. subti-
lis104. Interestingly, RNase Y of B. subtilis has been s hown
to associate with the RNA helicase CshA, enolase, PNPase
and RNase J1, possibly forming a degradosome-like com-
plex similar to the RNase E degradosome in E. coli105,106.
The RNase Y orthologue of S. pyogenes, CvfA, is induced
in the absence of carbohydrates and downregulates
genes involved in metabolism and transport. Strains of
S. pyogenes and S. aureus lacking CvfA show attenuated
virulence in mice and silkworms. Transcriptomics data
revealed that CvfA controls the expression of the S. pyo-
genes virulence factors speB (encoding streptopain, a
cysteine protease that is involved in the degradation of
host cellular proteins) and sagA (which is important for
rupturing host cells)107,108.
Concluding remarks
The recent advances in detailed RNA expression studies
carried out on a global scale have given us a vast amount
of information about how RNA regulation is exerted in
bacterial pathogens. What has become clear is that the
activity and stability of an mRNA (or an operon) can
be controlled by many factors. These regulatory factors
might be part of the transcript itself (the 5′ or 3′ end) and
can directly sense environmental cues such as tempera-
ture, pH and the concentration of specific metabolites.
They might also function in an antisense manner, being
true antisense transcripts (complementary to the cod-
ing RNA, like AmgR) or being expressed from locations
that are distal to target mRNAs (like RNAIII and VrrA).
These regulatory factors may interact with proteins such
as Hfq to ultimately control the stability and processing
of target transcripts.
One advantage of regulation by RNA structures is
the speed at which it can occur. As illustrated by the
pH-mediated regulation of alx, the speed of transcrip-
tion generating alternative RNA secondary structures
probably applies for most 5′ UTR-mediated regulation.
Similarly, with the help of the chaperone Hfq, trans-
acting sRNAs can be exposed to the correct target
mRNA, allowing the interaction to occur; this can also
be achieved by transcript maturation, which allows
exposure of the correct sequence to the trans-acting
sRNA. The speed of the interactions, along with the abil-
ity to react to environmental cues and for the mRNA to
be degraded as soon as the reaction is complete, makes
regulation by RNA structures ideal for pathogenic bacte-
ria. These bacteria encounter many different, often hos-
tile, environments (for example, the intestine, the blood
and the phagosome) during the course of infection. This
Review shows that we have started to grasp the role of
RNAs during infection, although many questions need
to be addressed. Are the regulatory RNAs expressed in
the whole population at the same time or are they only
expressed in some bacteria? This is particularly interest-
ing for cis-acting antisense RNAs and their targets on
the opposite strand because it might explain stochastic
gene expression in a population (that is, expression of
a particular gene in only a subset of a bacterial popula-
tion). At what stages during infection are the regulatory
RNAs active and expressed? Can regulatory RNAs inter-
act more directly with the host (can they be secreted and
function as microRNAs, downregulating host RNAs)?
REVIEWS
864
|
DECEMBER 2010
|
VOLUME 8 www.nature.com/reviews/micro
© 20 Macmillan Publishers Limited. All rights reserved10

It is also possible that the information we obtain on
RNA regulation could be used to develop new meth-
ods of treatment. Different analogues of riboswitch-
binding metabolites have already proved successful in
reducing the expression of bacterial growth genes and
decreasing pathogenesis in mice (for example, during
S. aureus infection)109,110. Similar unbiased chemical
biology experiments (blocking gene or protein function
with libraries of chemical compounds) could be used to
inhibit the function of certain 5′ UTR elements, making
them inactive and thereby preventing the expression of
the downstream gene product.
1. Beisel, C. L. & Storz, G. Base pairing small RNAs and
their roles in global regulatory networks. FEMS
Microbiol. Rev. 34, 866–882 (2010).
2. Toledo-Arana, A. & Solano, C. Deciphering the
physiological blueprint of a bacterial cell: revelations
of unanticipated complexity in transcriptome and
proteome. Bioassays 32, 461–467 (2010).
3. Loh, E., Gripenland, J. & Johansson, J. Control of
Listeria monocytogenes virulence by 5′-untranslated
RNA. Trends Microbiol. 14, 294–298 (2006).
4. Toledo-Arana, A. et al. The Listeria transcriptional
landscape from saprophytism to virulence. Nature
459, 950–956 (2009).
This study unveils new RNA-mediated regulation
occurring in bacteria by describing genome-wide
expression changes when L. monocytogenes
transits from a saprophytic to an infectious lifestyle.
5. Hurme, R. & Rhen, M. Temperature sensing in
bacterial gene regulation — what it all boils down to.
Mol. Microbiol. 30, 1–6 (1998).
6. Falconi, M., Colonna, B., Prosseda, G., Micheli, G. &
Gualerzi, C. O. Thermoregulation of Shigella and
Escherichia coli EIEC pathogenicity. A temperature-
dependent structural transition of DNA modulates
accessibility of virF promoter to transcriptional
repressor H-NS. EMBO J. 17, 7033–7043 (1998).
7. Johansson, J. & Cossart, P. RNA-mediated control of
virulence gene expression in bacterial pathogens.
Trends Microbiol. 11, 280–285 (2003).
8. Johansson, J. et al. An RNA thermosensor controls
expression of virulence genes in Listeria
monocytogenes. Cell 110, 551–561 (2002).
This work identifies an RNA thermosensor that
controls the expression of the master virulence
regulator in L. monocytogenes.
9. Freitag, N. E., Port, G. C. & Miner, M. D. Listeria
monocytogenes — from saprophyte to intracellular
pathogen. Nature Rev. Microbiol. 7, 623–628 (2009).
10. Hamon, M., Bierne, H. & Cossart, P. Listeria
monocytogenes: a multifaceted model. Nature Rev.
Microbiol. 4, 423–434 (2006).
11. Hoe, N. P. & Goguen, J. D. Temperature sensing in
Yersinia pestis: translation of the lcrf activator protein
is thermally regulated. J. Bacteriol. 175, 7901–7909
(1993).
12. Waldminghaus, T., Fippinger, A., Alfsmann, J. &
Narberhaus, F. RNA thermometers are common in
α- and γ-proteobacteria. Biol. Chem. 386, 1279–1286
(2005).
13. Flashner, Y. et al. Generation of Yersinia pestis
attenuated strains by signature-tagged mutagenesis in
search of novel vaccine candidates. Infect. Immun. 72,
908–915 (2004).
14. Von Pawel-Rammingen, U. et al. GAP activity of the
Yersinia YopE cytotoxin specifically targets the Rho
pathway: a mechanism for disruption of actin
microfilament structure. Mol. Microbiol. 36, 737–748
(2000).
15. Fällman, M. & Gustavsson, A. Cellular mechanisms of
bacterial internalization counteracted by Yersinia. Int.
Rev. Cytol. 246, 135–188 (2005).
16. Waldminghaus, T., Heidrich, N., Brantl, S. &
Narberhaus, F. FourU: a novel type of RNA
thermometer in Salmonella. Mol. Microbiol. 65,
413–424 (2007).
17. Nechooshtan, G., Elgrably-Weiss, M., Sheaffer, A.,
Westhof, E. & Altuvia, S. A pH-responsive
riboregulator. Genes Dev. 23, 2650–2662 (2009).
18. Taylor, D. E. Bacterial tellurite resistance. Trends
Microbiol. 7, 111–115 (1999).
19. Waters, L. S. & Storz, G. Regulatory RNAs in bacteria.
Cell 136, 615–628 (2009).
20. Roth, A. & Breaker, R. R. The structural and functional
diversity of metabolite-binding riboswitches. Annu.
Rev. Biochem. 78, 305–334 (2009).
21. Henkin, T. M. Riboswitch RNAs: using RNA to sense
cellular metabolism. Genes Dev. 22, 3383–3390
(2008).
22. Nudler, E. & Mironov, A. S. The riboswitch control of
bacterial metabolism. Trends Biochem. Sci. 29, 11–17
(2004).
23. Sudarsan, N. et al. Riboswitches in eubacteria sense
the second messenger cyclic di-GMP. Science 321,
411–413 (2008).
The authors show that bacterial riboswitches can
sense the bacterial second messenger c-di-GMP
and thereby control the expression of genes
involved in virulence and pilus formation.
24. Kirn, T. J., Jude, B. A. & Taylor, R. K. A colonization
factor links Vibrio cholerae environmental survival and
human infection. Nature 438, 863–866 (2005).
25. Tamayo, R., Tischler, A. D. & Camilli, A. The EAL
domain protein VieA is a cyclic diguanylate
phosphodiesterase. J. Biol. Chem. 280, 33324–33330
(2005).
26. Wilkie, G. S., Dickson, K. S. & Gray, N. K. Regulation of
mRNA translation by 5′- and 3′-UTR-binding factors.
Trends Biochem. Sci. 28, 182–188 (2003).
27. Sonenberg, N. & Hinnebusch, A. G. Regulation of
translation initiation in eukaryotes: mechanisms and
biological targets. Cell 136, 731–745 (2009).
28. Rasmussen, S., Nielsen, H. B. & Jarmer, H. The
transcriptionally active regions in the genome of
Bacillus subtilis. Mol. Microbiol. 73, 1043–1057
(2009).
29. ten Broeke-Smits, N. J. et al. Operon structure of
Staphylococcus aureus. Nucleic Acids Res. 38,
3263–3274 (2010).
30. Beaume, M. et al. Cartography of methicillin-resistant
S. aureus transcripts: detection, orientation and
temporal expression during growth phase and stress
conditions. PLoS ONE 5, e10725 (2010).
31. Jacquier, A. The complex eukaryotic transcriptome:
unexpected pervasive transcription and novel
small RNAs. Nature Rev. Genet. 10, 833–844
(2009).
32. Berretta, J. & Morillon, A. Pervasive transcription
constitutes a new level of eukaryotic genome
regulation. EMBO Rep. 10, 973–982 (2009).
33. Sharma, C. M. et al. The primary transcriptome of the
major human pathogen Helicobacter pylori. Nature
464, 250–255 (2010).
Using an improved high-throughput RNA-sequencing
methodology, the authors create a genome-wide
map of the transcriptional start sites of H. pylori,
revealing unexpected massive antisense RNA
transcription.
34. Dornenburg, J. E., DeVita, A. M., Palumbo, M. J. &
Wade, J. T. Widespread antisense transcription in
Escherichia coli. mBio 1, 1–4 (2010).
35. Selinger, D. W. et al. RNA expression analysis using a
30 base pair resolution Escherichia coli genome array.
Nature Biotech. 18, 1262–1268 (2000).
36. Guell, M. et al. Transcriptome complexity in a genome-
reduced bacterium. Science 326, 1268–1271
(2009).
37. Filiatrault, M. J. et al. Transcriptome analysis of
Pseudomonas syringae identifies new genes,
noncoding RNAs, and antisense activity. J. Bacteriol.
192, 2359–2372 (2010).
38. Georg, J. et al. Evidence for a major role of antisense
RNAs in cyanobacterial gene regulation. Mol. Syst.
Biol. 5, 1–17 (2009).
39. Giangrossi, M. et al. A novel antisense RNA regulates
at transcriptional level the virulence gene icsA of
Shigella flexneri. Nucleic Acids Res. 38, 3362–3375
(2010).
40. Padalon-Brauch, G. et al. Small RNAs encoded within
genetic islands of Salmonella typhimurium show host-
induced expression and role in virulence. Nucleic Acids
Res. 36, 1913–1927 (2008).
41. Stork, M., Di Lorenzo, M., Welch, T. J. & Crosa, J. H.
Transcription termination within the iron transport-
biosynthesis operon of Vibrio anguillarum requires
an antisense RNA. J. Bacteriol. 189, 3479–3488
(2007).
42. Lee, E. J. & Groisman, E. A. An antisense RNA that
governs the expression kinetics of a multifunctional
virulence gene. Mol. Microbiol. 76, 1020–1033
(2010).
43. Shearwin, K. E., Callen, B. P. & Egan, J. B.
Transcriptional interference – a crash course. Trends
Genet. 21, 339–345 (2005).
44. Santiviago, C. A. et al. Analysis of pools of targeted
Salmonella deletion mutants identifies novel genes
affecting fitness during competitive infection in mice.
PLoS Pathog. 5, 1 –12 (2009).
45. Sahr, T. et al. Two small ncRNAs jointly govern
virulence and transmission in Legionella pneumophila.
Mol. Microbiol. 72, 741–762 (2009).
46. Gillaspy, A. F. et al. Role of the accessory gene
regulator (agr) in pathogenesis of Staphylococcal
osteomyelitis. Infect. Immun. 63, 3373–3380
(1995).
47. Cheung, A. L. et al. Diminished virulence of a sar/agr
mutant of Staphylococcus aureus in the rabbit model
of endocarditis. J. Clin. Invest. 94, 1815–1822
(1994).
48. Song, T. et al. A new Vibrio cholerae sRNA modulates
colonization and affects release of outer membrane
vesicles. Mol. Microbiol. 70, 100–111 (2008).
49. Chabelskaya, S., Gaillot, O. & Felden, B. A
Staphylococcus aureus small RNA is required for
bacterial virulence and regulates the expression of an
immune-evasion molecule. PLoS Pathog. 6, 1–11
(2010).
50. Faucher, S. P., Friedlander, G., Livny, J., Margalit, H. &
Shuman, H. A. Legionella pneumophila 6S RNA
optimizes intracellular multiplication. Proc. Natl Acad.
Sci. USA 107, 7533–7538 (2010).
A comprehensive study covering the identification
of several sRNAs, including a 6S homologue.
51. Fortune, D. R., Suyemoto, M. & Altier, C. Identification
of CsrC and characterization of its role in epithelial cell
invasion in Salmonella enterica serovar Typhimurium.
Infect. Immun. 74, 331–339 (2006).
52. Novick, R. P. & Geisinger, E. Quorum sensing in
Staphylococci. Annu. Rev. Genet. 42, 541–564 (2008).
53. Novick, R. P. et al. Synthesis of Staphylococcal
virulence factors is controlled by a regulatory RNA
molecule. EMBO J. 12, 3967–3975 (1993).
54. Walev, I. et al. Staphylococcal alpha-toxin kills human
keratinocytes by permeabilizing the plasma-
membrane for monovalent ions. Infect. Immun. 61,
4972–4979 (1993).
55. Morfeldt, E., Taylor, D., Vongabain, A. & Arvidson, S.
Activation of alpha-toxin translation in Staphylococcus
aureus by the trans-encoded antisense RNA, RNAIII.
EMBO J. 14, 4569–4577 (1995).
56. Boisset, S. et al. Staphylococcus aureus RNAIII
coordinately represses the synthesis of virulence
factors and the transcription regulator Rot by an
antisense mechanism. Genes Dev. 21, 1353–1366
(2007).
A detailed study of the regulatory mechanisms
used by RNAIII.
57. Huntzinger, E. et al. Staphylococcus aureus RNAIII
and the endoribonuclease III coordinately regulate
spa gene expression. EMBO J. 24, 824–835
(2005).
58. Geisinger, E., Adhikari, R. P., Jin, R. Z., Ross, H. F. &
Novick, R. P. Inhibition of rot translation by RNAIII, a
key feature of agr function. Mol. Microbiol. 61,
1038–1048 (2006).
59. Kawabata, S., Morita, T., Miyata, T., Iwanaga, S. &
Igarashi, H. Isolation and characterization of
Staphylocoagulase chymotryptic fragment.
Localization of the procoagulant-binding and
prothrombin-bindi ng domain of this protein. J. Biol.
Chem. 261, 1427–1433 (1986).
60. Chevalier, C. et al. Staphylococcus aureus RNAIII
binds to two distant regions of coa mRNA to arrest
translation and promote mRNA degradation. PLoS
Pathog. 6, 1–12 (2010).
REVIEWS
NATURE R EVIEWS
|
MICROBIOLOGY VOLUME 8
|
DECEMBER 2010
|
865
© 20 Macmillan Publishers Limited. All rights reserved10

61. Oliver, H. F. et al. Deep RNA sequencing of
L. monocytogenes reveals overlapping and extensive
stationary phase and sigma B-dependent
transcriptomes, including multiple highly transcribed
noncoding RNAs. BMC Genomics 10, 641 (2009).
62. Mandin, P., Repoila, F., Vergassola, M., Geissmann, T.
& Cossart, P. Identification of new noncoding RNAs in
Listeria monocytogenes and prediction of mRNA
targets. Nucleic Acids Res. 35, 962–974 (2007).
63. Christiansen, J. K. et al. Identification of small
Hfq-binding RNAs in Listeria monocytogenes.
RNA 12, 1383–1396 (2006).
64. Nielsen, J. S., Olsen, A. S., Bonde, M., Valentin-
Hansen, P. & Kallipolitis, B. H. Identification of a
σB-dependent small noncoding RNA in Listeria
monocytogenes. J. Bacteriol. 190, 6264–6270
(2008).
65. Loh, E. et al. A trans-acting riboswitch controls
expression of the virulence regulator PrfA in Listeria
monocytogenes. Cell 139, 770–779 (2009).
The first example of a riboswitch acting as a
trans-acting sRNA by binding a target mRNA
through direct base-pairing.
66. Wassarman, K. M. 6S RNA: a regulator of
transcription. Mol. Microbiol. 65, 1425–1431
(2007).
67. Livny, J., Fogel, M. A., Davis, B. M. & Waldor, M. K.
sRNAPredict: an integrative computational approach
to identify sRNAs in bacterial genomes. Nucleic Acids
Res. 33, 4096–4105 (2005).
68. Hammer, B. K. & Bassler, B. L. Quorum sensing
controls biofilm formation in Vibrio cholerae. Mol.
Microbiol. 50, 101–114 (2003).
69. Kovacikova, G. & Skorupski, K. Regulation of virulence
gene expression in Vibrio cholerae by quorum sensing:
HapR functions at the aphA promoter. Mol. Microbiol.
46, 1135–1147 (2002).
70. Zhu, J. et al. Quorum-sensing regulators control
virulence gene expression in Vibrio cholerae. Proc.
Natl Acad. Sci. USA 99, 3129–3134 (2002).
71. Song, T., Sabharwal, D. & Wai, S. N. VrrA mediates
Hfq-dependent regulation of OmpT synthesis in Vibrio
cholerae. J. Mol. Biol. 400, 682–688 (2010).
72. Wibbenmeyer, J. A., Provenzano, D., Landry, C. F.,
Klose, K. E. & Delcour, A. H. Vibrio cholerae OmpU
and OmpT porins are differentially affected by bile.
Infect. Immun. 70, 121–126 (2002).
73. Soper, T. J. & Woodson, S. A. The rpoS mRNA leader
recruits Hfq to facilitate annealing with DsrA sRNA.
RNA 14, 1907–1917 (2008).
74. Link, T. M., Valentin-Hansen, P. & Brennan, R. G.
Structure of Escherichia coli Hfq bound to
polyriboadenylate RNA. Proc. Natl Acad. Sci. USA
106, 19286–19291 (2009).
75. Chao, Y. J. & Vogel, J. The role of Hfq in bacterial
pathogens. Curr. Opin. Microbiol. 13, 24–33 (2010).
76. Christiansen, J. K., Larsen, M. H., Ingmer, H.,
Sogaard-Andersen, L. & Kallipolitis, B. H. The RNA-
binding protein Hfq of Listeria monocytogenes: role in
stress tolerance and virulence. J. Bacteriol. 186,
3355–3362 (2004).
77. Lenz, D. H. et al. The small RNA chaperone Hfq and
multiple small RNAs control quorum sensing in
Vibrio harveyi and Vibrio cholerae. Cell 118 , 69–82
(2004).
78. Lorenz, C. et al. Genomic SELEX for Hfq-binding RNAs
identifies genomic aptamers predominantly in
antisense transcripts. Nucleic Acids Res. 38,
3794–3808 (2010).
79. Anderson, K. L. & Dunman, P. M. Messenger RNA
turnover processes in Escherichia coli, Bacillus
subtilis, and emerging studies in Staphylococcus
aureus. Int. J. Microbiol. 2009, 525491 (2009).
80. Picard, F., Dressaire, C., Girbal, L. & Cocaign-
Bousquet, M. Examination of post-transcriptional
regulations in prokaryotes by integrative biology.
C. R. Biol. 332, 958–973 (2009).
81. Carpousis, A. J. The RNA degradosome of
Escherichia coli: an mRNA-degrading machine
assembled on RNase E. Annu. Rev. Microbiol. 61,
71–87 (2007).
82. Arraiano, C. M. et al. The critical role of RNA
processing and degradation in the control of gene
expression. FEMS Microbiol. Rev. 34, 883–923
(2010).
83. Condon, C. Maturation and degradation of RNA in
bacteria. Curr. Opin. Microbiol. 10, 271–278 (2007).
84. Belasco, J. G. All things must pass: contrasts and
commonalities in eukaryotic and bacterial mRNA
decay. Nature Rev. Mol. Cell Biol. 11, 467–478
(2010).
85. Nurmohamed, S., Vaidialingam, B., Callaghan, A. J. &
Luisi, B. F. Crystal structure of Escherichia coli
polynucleotide phosphorylase core bound to RNase E,
RNA and manganese: implications for catalytic
mechanism and RNA degradosome assembly. J. Mol.
Biol. 389, 17–33 (2009).
86. Khemici, V., Toesca, I., Poljak, L., Vanzo, N. F. &
Carpousis, A. J. The RNase E of Escherichia coli has at
least two binding sites for DEAD-box RNA helicases:
functional replacement of RhlB by RhlE. Mol.
Microbiol. 54, 1422–1430 (2004).
87. Gorna, M. W. et al. The regulatory protein RraA
modulates RNA-binding and helicase activities of the
E. coli RNA degradosome. RNA 16, 553–562
(2010).
88. Jiang, X. Q. & Belasco, J. G. Catalytic activation
of multimeric RNase E and RNase G by
5′-monophosphorylated RNA. Proc. Natl Acad. Sci.
USA 101, 9211–9216 (2004).
89. Koslover, D. J. et al. The crystal structure of the
Escherichia coli RNase E apoprotein and a mechanism
for RNA degradation. Structure 16, 1238–1244
(2008).
90. Jiang, X. Q., Diwa, A. & Belasco, J. G. Regions of
RNase E important for 5′-end-dependent RNA
cleavage and autoregulated synthesis. J. Bacteriol.
182, 2468–2475 (2000).
91. Mackie, G. A. Ribonuclease E is a 5′-end-dependent
endonuclease. Nature 395, 720–723 (1998).
92. Kime, L., Jourdan, S. S., Stead, J. A., Hidalgo-Sastre, A.
& McDowall, K. J. Rapid cleavage of RNA by RNase E
in the absence of 5′-monophosphate stimulation. Mol.
Microbiol. 76, 590–604 (2009).
93. Morita, T., Maki, K. & Aiba, H. RNase E-based
ribonucleoprotein complexes: mechanical basis of
mRNA destabilization mediated by bacterial
noncoding RNAs. Genes Dev. 19, 2176–2186 (2005).
94. Moll, I., Afonyushkin, T., Vytvytska, O., Kaberdin, V. R.
& Blasi, U. Coincident Hfq binding and RNase E
cleavage sites on mRNA and small regulatory RNAs.
RNA 9, 1308–1314 (2003).
95. Hernday, A., Braaten, B. & Low, D. The intricate
workings of a bacterial epigenetic switch. Adv. Syst.
Biol. 547, 83–89 (2004).
96. Nilsson, P., Naureckiene, S. & Uhlin, B. E. Mutations
affecting mRNA processing and fimbrial biogenesis in
the Escherichia coli pap operon. J. Bacteriol. 178,
683–690 (1996).
97. Court, D. in Control of mRNA Stability (eds
Brawerman, G. and Belasco, J.) 70–116 (Academic
Press, New York, 1993).
98. Chevalier, C. et al. Staphylococcus aureus
endoribonuclease III purification and properties.
Methods Enzymol. 447, 309–327 (2008).
99. Mathy, N. et al. 5′-to-3′ exoribonuclease activity in
Bacteria: role of RNase J1 in rRNA maturation and 5′
stability of mRNA. Cell 129, 681–692 (2007).
The first example of a bacterial exonuclease
degrading RNA in a 5′-to-3′ orientation.
100. Even, S. et al. Ribonucleases J1 and J2: two novel
endoribonucleases in B. subtilis with functional
homology to E. coli RNase E. Nucleic Acids Res. 33,
2141–2152 (2005).
101. Deikus, G. & Bechhofer, D. H. Bacillus subtilis trp
leader RNA RNase J1 endonuclease cleavage
specificity and PNPase processing. J. Biol. Chem. 284,
26394–26401 (2009).
102. Mathy, N. et al. Bacillus subtilis ribonucleases J1 and
J2 form a complex with altered enzyme behaviour.
Mol. Microbiol. 75, 489–498 (2010).
103. Bugrysheva, J. V. & Scott, J. R. The ribonucleases J1
and J2 are essential for growth and have independent
roles in mRNA decay in Streptococcus pyogenes. Mol.
Microbiol. 75, 731–743 (2010).
104. Shahbabian, K., Jamalli, A., Zig, L. & Putzer, H.
RNase Y, a novel endoribonuclease, initiates
riboswitch turnover in Bacillus subtilis. EMBO J. 28,
3523–3533 (2009).
105. Commichau, F. M. et al. Novel activities of glycolytic
enzymes in Bacillus subtilis: interactions with essential
proteins involved in mRNA processing. Mol. Cell
Proteomics 8, 1350–1360 (2009).
106. Lehnik-Habrink, M. et al. The RNA degradosome in
Bacillus subtilis: identification of CshA as the major
RNA helicase in the multiprotein complex. Mol.
Microbiol. 77, 958–971 (2010).
107. Kang, S. O., Caparon, M. G. & Cho, K. H. Virulence
gene regulation by CvfA, a putative RNase: the CvfA-
enolase complex in Streptococcus pyogenes links
nutritional stress, growth-phase control, and virulence
gene expression. Infect. Immun. 78, 2754–2767
(2010).
108. Kaito, C. et al. Silkworm pathogenic bacteria infection
model for identification of novel virulence genes. Mol.
Microbiol. 56, 934–944 (2005).
109. Mulhbacher, J. et al. Novel riboswitch ligand analogs
as selective inhibitors of guanine-related metabolic
pathways. PLoS Pathog. 6, e1000865 (2010).
110. Kim, J. N. et al. Design and antimicrobial action of
purine analogues that bind guanine riboswitches. ACS
Chem. Biol. 4, 915–927 (2009).
Acknowledgements
J.G., S.N. and E.L. are supported by the J. C. Kempe founda-
tion. A.T.-A. has a JAE-DOC research contract from Consejo
Superior de Investigaciones Científicas (CSIC; the Spanish
National Research Council). J.J. is supported by Umeå
University, Sweden, by the Swedish Research Council grants
2008-58X-15144-05-3 and 621-2009-5677, and by
European Research Council Starting Grant number 260764.
We apologize to colleagues whose work could not be cited
owing to space limitations.
Competing interests statement
The authors declare no competing financial interests.
FURTHER INFORMATION
Jörgen Johansson’s homepage: http://www.molbiol.umu.se/
english/research/researchers/jorgen-johansson
ALL LINKS ARE ACTIVE IN THE ONLINE PDF
REVIEWS
866
|
DECEMBER 2010
|
VOLUME 8 www.nature.com/reviews/micro
© 20 Macmillan Publishers Limited. All rights reserved10