Molecular Microbiology (2010) 78(6), 1327–1331 䊏
doi:10.1111/j.1365-2958.2010.07428.x
First published online 26 October 2010
MicroCommentary
Small RNAs promote mRNA stability to activate the synthesis
of virulence factors
mmi_7428 1327..1331
Dimitri Podkaminski and Jörg Vogel*
Institute for Molecular Infection Biology, University of
Würzburg, 97080 Würzburg, Germany.
Summary
Although most bacterial small RNAs act to repress
target mRNAs, some also activate messengers. The
predominant mode of activation has been seen in
‘anti-antisense’ regulation whereby a small RNA prevents the formation of an inhibitory 5⬘ mRNA structure that otherwise impairs translational initiation and
protein synthesis. The translational activation might
also stabilize the target yet this was considered a
secondary effect in the examples known thus far. Two
recent papers in Molecular Microbiology investigate
post-transcriptional activation of collagenase mRNA
by Clostridium VR-RNA, and streptokinase mRNA by
Streptococcus FasX RNA, to suggest that small RNAs
exert positive regulation of virulence genes primarily
at the level of mRNA stabilization.
Small RNAs (sRNAs) are the largest class of posttranscriptional regulators in bacteria known to date and
the vast majority of them regulate mRNAs by base-pairing
mechanisms. Regardless of whether sRNAs are encoded
in cis or trans relative to their target mRNAs, and recognize them by long or short antisense pairing, they most
commonly decrease mRNA translation or stability. As
such, the bacterial sRNAs very much like their prominent
eukaryotic counterparts – microRNAs – are primarily
known as negative regulators of gene expression (Waters
and Storz, 2009; Papenfort and Vogel, 2010).
However, unlike in eukaryotes where positive action by
sRNAs has rarely been seen, studies in diverse bacteria
have also identified numerous sRNAs that function as
mRNA activators (Fröhlich and Vogel, 2009). Most
Accepted 5 October, 2010. *For correspondence. E-mail joerg.
[email protected]; Tel. (+49) 931 3182 576; Fax (+49) 931
3182 578.
© 2010 Blackwell Publishing Ltd
positive regulations take place at the level of protein synthesis, following a common theme (Fig. 1, upper panel):
under normal circumstances – no sRNA present – the
translation of the target mRNA is intrinsically inhibited
by a RNA structure that folds after transcription of the
5′ untranslated region (UTR) and occludes the Shine–
Dalgarno sequence (SD) and/or AUG start codon, the two
essential elements of the ribosome binding site (RBS) for
translational initiation. The sRNA exerts its positive activity
as a structural competitor that frees the occluded RBS,
usually by pairing with the anti-SD or anti-AUG sequences
of the inhibitory structure. As translation and ribosome
occupancy increase, ribonucleases might stand a lesser
chance for cleavage, whereby the stability and steadystate levels of the activated mRNA might also increase.
This elegant mechanism, commonly referred to as the
‘anti-antisense mechanism’, was first proposed for Staphylococcus aureus RNAIII (Morfeldt et al., 1995), but
only brought to a comprehensive understanding through
work on the rpoS mRNA, which is translationally activated
by no fewer than three sRNAs (ArcZ, DsrA and RprA)
in Escherichia coli, Salmonella and likely many other
Gram-negative bacteria (Papenfort et al., 2009; Mandin
and Gottesman, 2010; Soper et al., 2010). There is also in
vitro work to suggest that the anti-antisense mechanism
only requires sRNA pairing for stimulation of translation,
as demonstrated for DsrA-rpoS mRNA and 30S ribosome
association (Worhunsky et al., 2003) and GlmZ-glmS
mRNA and 70S ribosome translation (Urban and Vogel,
2008). Notably, this basic mechanism of translational activation is also common in cis-regulatory systems such as
RNA thermometers or riboswitches where instead of a
trans-acting sRNA, temperature-induced melting or ligand
binding followed by conformational changes in the 5′ UTR
resolve inhibitory RBS structures (Waters and Storz,
2009).
Regarding antisense pairing, there is in principle no
conceptual difference between sRNAs that activate or
repress targets, and some sRNAs such as ArcZ or DsrA
indeed work in both ways on different mRNAs (Sledjeski
& Gottesman, 1995; Papenfort et al., 2009; Mandin and
1328 D. Podkaminski and J. Vogel 䊏
no sRNA
sRNA present
general anit-antisense
Ribosome
Ribosome
5'
5'
3'
sRNA
RBS
RBS
CDS
5'
3'
CDS
mRNA
3'
mRNA
colA activation
RNase
5'
RBS
RNase
3'
CDS
5'
3'
RBS
VR-RNA
RNase
CDS
colA mRNA
3'
RBS
5'
CDS
3'
colA mRNA
colA mRNA
5'
ska activation
FasX sRNA
RNase
5'
RBS
5'
RBS
RNase
CDS
3'
ska mRNA
5'
CDS
3'
ska mRNA
3'
Fig. 1. Small RNAs activate protein synthesis. Top: General ‘anti-antisense’ mechanism: In the absence of a sRNA an intrinsic stem-loop
structure occludes the RBS, inhibiting ribosome initiation. sRNA binding to the anti-SD and/or the anti-AUG sequence prevents the
intramolecular interaction, enabling ribosome association. Middle: colA activation: An inhibitory stem-loop structure prevents the expression of
colA mRNA in absence of VR-RNA. VR-RNA binding to the 5′ UTR of colA induces an endonucleolytic cleavage that generates a new mRNA
species with an accessible RBS and a stabilizing 5′ stem-loop structure. Bottom: ska activation: When FasX sRNA is absent ska mRNA is
rapidly degraded. FasX binding to the extreme 5′ end of ska generates a 5′ proximal stabilizing structure, which protects ska from nucleolytic
attack.
Gottesman, 2010). In fact, the only common denominator
for activation has been that sRNA pairing in the 5′ UTR
occurs upstream of the general ribosome contact region
on bacterial mRNAs, i.e. upstream of the approximate -20
position relative to AUG; in contrast, most sRNAs that
inhibit translation bind at the SD or AUG sequences.
However, as more sRNAs are being discovered and scrutinized for cellular targets and molecular mechanisms, it
appears that translational repression is not confined to
the RBS but can also take place in the upstream 5′ UTR
or coding sequence (Waters and Storz, 2009; Papenfort
and Vogel, 2010). It also emerges that tinkering with transcript stability rather than impairing translation is another
viable option for sRNAs to repress target mRNAs (Caron
et al., 2010). May the same diversity of mechanisms hold
true for positive regulations in the 5′ UTR?
Two recent papers in Molecular Microbiology by
Obana et al. (2010) and Ramirez-Peña et al. (2010)
suggest that sRNA-mediated stability control is the
crucial element of activation of trans-encoded mRNAs.
The first paper seeks to understand the mechanism by
which VR-RNA promotes that expression of collagenase
(colA) mRNA in Clostridium perfringens (Obana et al.,
2010). VR-RNA was originally identified in a screen for
target genes of the VirR/VirS two-component system
(Banu et al., 2000), a key regulator of gene expression
in this human pathogen. Earlier studies had identified a
requirement of the 3′ end of VR-RNA for colA mRNA
accumulation, and excluded potential reading frames of
the 386 nt VR-RNA transcript in the process, collectively
suggesting that colA was activated by a non-coding
RNA (Shimizu et al., 2002). Yet, the underlying mechanism remained elusive.
Using a combination of truncation analysis, biocomputational prediction and in vitro RNA assays, Obana et al.
(2010) now demonstrate that the extreme 3′ end of
VR-RNA binds to the 5′ UTR of colA mRNA via an imperfect ~35 bp interaction. Intriguingly, binding of VR-RNA
appears to cause endonucleolytic cleavage in the colA
mRNA, just downstream of the predicted RNA duplex
(Fig. 1, middle panel). As the cleavage removes the first
~60 nucleotides, a shorter and much more stable colA
mRNA species is produced whose structure is predicted
to differ from the full-length mRNA in two ways. First, the
RBS is no longer sequestered in a long 5′ hairpin, hence
permitting ribosome access as would be expected from
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 1327–1331
Small RNAs promote mRNA stability 1329
the anti-antisense mechanism. Second, and more importantly, the new 5′ end of colA mRNA is a shorter hairpin
without a protrusion that could be an easy catch for
nucleases. Thus, the endonucleolytic cleavage induced
by VR-RNA might increase both translation and mRNA
stability yet independently of each other. Which of the two
events makes the major contribution is not clear at this
point; mutation of the RBS shows that without ribosome
binding, even the shorter colA mRNA is very unstable
(the mRNA cleavage itself does not seem to rely on ribosome binding). Nonetheless, the dramatic increase in
colA mRNA stability and steady-state levels induced by
VR-RNA nicely correlates with ColA protein production
whereas mere translational activation usually results in
much stronger protein synthesis than changes at the
mRNA level.
The other study by Ramirez-Peña et al. (2010) probes
the underlying cause of an almost complete loss of
ska mRNA expression in the absence of FasX sRNA in
group A Streptococcus (GAS). The secreted streptokinase
(SKA) is a key pathogenicity factor of GAS (Sun et al.,
2004), and aids the dissemination of this dreadful human
pathogen by activating host plasminogen to cause breakdown of fibrin fibres in a blood clot. FasX originates
from downstream of fasBCA (an operon encoding a
two-component regulator) and is transcribed in a FasAdependent manner (Kreikemeyer et al., 2001).
Similar to the colA mRNA in the absence of VR-RNA,
the ska mRNA is almost undetectable in a fasX deletion
strain of GAS, and Ramirez-Peña et al. (2010) show
that FasX impacts ska mRNA stability rather than transcription. The authors also confirm an earlier speculation
that FasX might be a non-coding RNA (Kreikemeyer
et al., 2001), showing that nonsense mutations meant
to disrupt all potential reading frames of the 205 nt long
FasX transcript do not abrogate ska mRNA stabilization.
A single-nucleotide mutagenesis screen then identifies
crucial positions in FasX for ska mRNA stabilization.
Knowledge of these positions guides the prediction of
a short (nine base pair) FasX RNA interaction with the
very 5′ end of ska mRNA, which is solidly confirmed by
successful compensatory base pair changes in the two
partners. Importantly, FasX pairing renders the formerly
single-stranded 5′ end of the ska message doublestranded, as if generating a 5′ hairpin (Fig. 1, lower
panel). Although FasX binds dangerously close to the
RBS of the target – around 30 nucleotides upstream of
the AUG where some sRNAs repress translation
(Sharma et al., 2007) – the net result is a dramatic
increase in SKA protein synthesis that again neatly correlates with increased ska mRNA stability and steadystate levels.
Several lines of evidence support that the 5′ end
targeting of ska mRNA is indeed a novel mechanism of
activation that relies primarily on mRNA stabilization. The
labile ska mRNA can be stabilized by a short artificial 5′
extension that generates a hairpin structure akin to the
bound state of FasX RNA. In contrast, a 5′ extension of
scrambled sequence destabilizes the message, regardless of FasX binding, collectively suggesting that FasX
must bind to the very 5′ end and prevent the target from
having a single-stranded protrusion.
Different from VR-RNA activity on colA, FasX stabilizes
the ska mRNA without an intermediate step of target
cleavage. However, the common denominator in both
systems is the generation of double-stranded 5′ ends of
the targets without protrusions. The function of paired 5′
ends has been well-established in diverse bacteria, starting with elegant work in E. coli where a 5′ stem-loop was
shown to be critical for the intracellular stability of ompA
mRNA (Emory et al., 1992). Reminiscent of the above
experiment with FasX and ska, unpaired nucleotides at
the 5′ end nullified the stability gain by the 5′ stem-loop of
ompA. The current model of mRNA protection in E. coli
implies that a paired 5′ end hinders the activity of two
enzymes, namely RppH and RNase E, of which RppH
converts the 5′ tri-phosphate ends of primary transcripts
to 5′ mono-phosphates, the latter of which are the preferred substrate for the major mRNA decay factor, RNase
E (Belasco, 2010).
Following the work in E. coli, 5′ proximal elements were
also shown to stabilize mRNAs in Bacillus subtilis (Sharp
and Bechhofer, 2005), a Gram-positive organism whose
repertoire of RNA processing factors and nucleases
would be more akin to C. perfringens and GAS. Instead
of RNase E, mRNA decay in Gram-positive appears to be
dominated by the 5′-to-3′ exo/endoribonucleases J1 and
J2, and endonuclease Y (Belasco, 2010). Again, however,
biochemical studies have shown that RNase J prefers
single-stranded 5′ ends (Mathy et al., 2007), which could
explain how colA mRNA after cleavage by VR-RNA,
or ska mRNA when paired to FasX, become protected
against degradation and hence stabilized. However,
exactly which nuclease is outwitted by VR-RNA or FasX is
unknown at this point. Ramirez-Peña et al. (2010) made
the commendable effort to look into GAS mutant strains of
RNase Y or PNPase, but did not observe changes in ska
mRNA stability. RNase J1 or J2 remain on their list, but
are harder to test because they are encoded by essential
GAS genes.
Thus, future work on colA, ska and other potential
targets of VR-RNA and FasX might greatly benefit from
in vitro systems in which sRNA-promoted cleavage and
protection from mRNA decay can be recapitulated upon
the addition of purified proteins, or in bacterial extracts
depleted for the factors of interest. Similarly, in vitro
assays will permit to address the relative contribution of
a free RBS and hence enhanced translation to the highly
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 1327–1331
1330 D. Podkaminski and J. Vogel 䊏
elevated synthesis of ColA and SKA from their activated
mRNAs. For example, despite the strong evidence for
ska mRNA stabilization by protection against nucleolytic
attack, a positive effect of FasX on translational initiation
cannot be ruled out at this point; FasX binding at the 5′
end might render the RBS more accessible to ribosomes
by promoting structural changes of the 5′ UTR. Thus, it
will also be essential to use the well-established methodology (Geissmann et al., 2009) and determine mRNA
structures experimentally, in order to validate the current
in silico predictions that underlie the presented models of
regulations.
It is interesting to note that mRNA processing has also
begun to be addressed for those sRNAs that use the
anti-antisense mechanism and translational control. For
example, the duplex of DsrA and its rpoS target mRNA
was shown to trigger an alternative mRNA cleavage by
RNase III in E. coli, and similar to FasX, DsrA was predicted to stay bound to the rpoS mRNA and endow it with
an alternative 5′ stem-loop (Resch et al., 2008; Vecerek
et al., 2010). Beyond the 5′ end, the E. coli GadY sRNA
constitutes an example of positive regulation at the other
end of mRNA. GadY is transcribed oppositely to the
intergenic region of gadXW mRNA, and its base-pairing
promotes processing and 3′ end formation to yield an
abundant monocistronic gadX mRNA; the exact mechanism is yet to be understood (Opdyke et al., 2004).
One may argue that, given the mechanistic diversity
of sRNAs and the importance of regulated RNA turnover
in bacteria, mRNA activation at the stability level was not
beyond the fringes of the imagination. However, the two
papers by Obana et al. (2010) and Ramirez-Peña et al.
(2010) are important in providing model systems with
well-detectable sRNAs and targets that could foster a
comprehensive understanding of such mechanisms.
VR-RNA and FasX also exemplify the increasing appreciation of the roles of sRNAs in the control of virulence
factor production in bacterial pathogens (Papenfort and
Vogel, 2010). What is more, GAS inasmuch as other
important bacterial pathogens seem to lack the common
sRNA chaperone Hfq (Chao and Vogel, 2010), and
perhaps lend themselves as model organisms to discover
other auxiliary proteins for sRNA-mediated regulations.
Moreover, as more bacterial pathogens are screened for
sRNAs by deep sequencing approaches (Martin et al.,
2010; Sharma et al., 2010), not only the list of riboregulators of virulence factors but also the diversity of sRNA
mechanisms is expected to grow.
These present studies in Gram-positive bacteria might
also impact our understanding of target regulation by the
intensely investigated Hfq-dependent sRNAs of Gramnegative bacteria. An increasing number of Salmonella
and E. coli mRNAs turn out to be targeted in the upstream
5′ UTR (Sharma et al., 2007; Vecerek et al., 2007; Bal-
bontin et al., 2010; Holmqvist et al., 2010) and regulation
of mRNA stability is also becoming more common (Pfeiffer
et al., 2009; Caron et al., 2010). Furthermore, some
sRNA binding sites overlap with the 5′ hairpin of stable
mRNAs, for example, RybB sRNA sites in the Salmonella
ompC or ompF messages (D. Podkaminski and J. Vogel,
unpublished results from our laboratory). Thus, it is tempting to speculate that some sRNAs work in a reciprocal
manner to VR-RNA or FasX, and destabilize targets by
competing with the formation of a protective 5′ hairpin
structure.
Acknowledgements
Work in the Vogel lab is supported by the DFG Priority
Program SPP1258 Sensory and Regulatory RNAs in
Prokaryotes and BMBF NGFN+ grant RNomics of infection.
References
Balbontin, R., Fiorini, F., Figueroa-Bossi, N., Casadesus, J.,
and Bossi, L. (2010) Recognition of heptameric seed
sequence underlies multi-target regulation by RybB
small RNA in Salmonella enterica. Mol Microbiol 78: 380–
394.
Banu, S., Ohtani, K., Yaguchi, H., Swe, T., Cole, S.T.,
Hayashi, H., and Shimizu, T. (2000) Identification of novel
VirR/VirS-regulated genes in Clostridium perfringens. Mol
Microbiol 35: 854–864.
Belasco, J.G. (2010) All things must pass: contrasts and
commonalities in eukaryotic and bacterial mRNA decay.
Nat Rev Mol Cell Biol 11: 467–478.
Caron, M.P., Lafontaine, D.A., and Massé, E. (2010) Small
RNA-mediated regulation at the level of transcript stability.
RNA Biol 7: 140–144.
Chao, Y., and Vogel, J. (2010) The role of Hfq in bacterial
pathogens. Curr Opin Microbiol 13: 24–33.
Emory, S.A., Bouvet, P., and Belasco, J.G. (1992) A
5′-terminal stem-loop structure can stabilize mRNA in
Escherichia coli. Genes Dev 6: 135–148.
Fröhlich, K.S., and Vogel, J. (2009) Activation of gene
expression by small RNA. Curr Opin Microbiol 12: 674–
682.
Geissmann, T., Marzi, S., and Romby, P. (2009) The role of
mRNA structure in translational control in bacteria. RNA
Biol 6: 153–160.
Holmqvist, E., Reimegard, J., Sterk, M., Grantcharova, N.,
Römling, U., and Wagner, E.G. (2010) Two antisense
RNAs target the transcriptional regulator CsgD to inhibit
curli synthesis. EMBO J 29: 1840–1850.
Kreikemeyer, B., Boyle, M.D., Buttaro, B.A., Heinemann, M.,
and Podbielski, A. (2001) Group A streptococcal growth
phase-associated virulence factor regulation by a novel
operon (Fas) with homologies to two-component-type
regulators requires a small RNA molecule. Mol Microbiol
39: 392–406.
Mandin, P., and Gottesman, S. (2010) Integrating anaerobic/
aerobic sensing and the general stress response through
the ArcZ small RNA. EMBO J 29: 3094–3107.
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 1327–1331
Small RNAs promote mRNA stability 1331
Martin, J., Zhu, W., Passalacqua, K.D., Bergman, N., and
Borodovsky, M. (2010) Bacillus anthracis genome organization in light of whole transcriptome sequencing. BMC
Bioinformatics 11 (Suppl. 3): S10.
Mathy, N., Benard, L., Pellegrini, O., Daou, R., Wen, T., and
Condon, C. (2007) 5′-to-3′ exoribonuclease activity in bacteria: role of RNase J1 in rRNA maturation and 5′ stability
of mRNA. Cell 129: 681–692.
Morfeldt, E., Taylor, D., von Gabain, A., and Arvidson, S.
(1995) Activation of alpha-toxin translation in Staphylococcus aureus by the trans-encoded antisense RNA, RNAIII.
EMBO J 14: 4569–4577.
Obana, N., Shirahama, Y., Abe, K., and Nakamura, K. (2010)
Stabilization of Clostridium perfringens collagenase mRNA
by VR-RNA-dependent cleavage in 5′ leader sequence.
Mol Microbiol 77: 1416–1428.
Opdyke, J.A., Kang, J.G., and Storz, G. (2004) GadY, a
small-RNA regulator of acid response genes in Escherichia
coli. J Bacteriol 186: 6698–6705.
Papenfort, K., and Vogel, J. (2010) Regulatory RNA in bacterial pathogens. Cell Host Microbe 8: 116–127.
Papenfort, K., Said, N., Welsink, T., Lucchini, S., Hinton, J.C.,
and Vogel, J. (2009) Specific and pleiotropic patterns of
mRNA regulation by ArcZ, a conserved, Hfq-dependent
small RNA. Mol Microbiol 74: 139–158.
Pfeiffer, V., Papenfort, K., Lucchini, S., Hinton, J.C., and
Vogel, J. (2009) Coding sequence targeting by MicC RNA
reveals bacterial mRNA silencing downstream of translational initiation. Nat Struct Mol Biol 16: 840–846.
Ramirez-Peña, E., Treviño, J., Liu, Z., Perez, N., and Sumby,
P. (2010) The group A Streptococcus small regulatory
RNA FasX enhances streptokinase activity by increasing
the stability of the ska mRNA transcript. Mol Microbiol 78:
1332–1347.
Resch, A., Afonyushkin, T., Lombo, T.B., McDowall, K.J.,
Bläsi, U., and Kaberdin, V.R. (2008) Translational activation
by the noncoding RNA DsrA involves alternative RNase III
processing in the rpoS 5′-leader. RNA 14: 454–459.
Sharma, C.M., Darfeuille, F., Plantinga, T.H., and Vogel, J.
(2007) A small RNA regulates multiple ABC transporter
mRNAs by targeting C/A-rich elements inside and upstream
of ribosome-binding sites. Genes Dev 21: 2804–2817.
Sharma, C.M., Hoffmann, S., Darfeuille, F., Reignier, J., Findeiss, S., Sittka, A., et al. (2010) The primary transcriptome
of the major human pathogen Helicobacter pylori. Nature
464: 250–255.
Sharp, J.S., and Bechhofer, D.H. (2005) Effect of 5′-proximal
elements on decay of a model mRNA in Bacillus subtilis.
Mol Microbiol 57: 484–495.
Shimizu, T., Yaguchi, H., Ohtani, K., Banu, S., and Hayashi,
H. (2002) Clostridial VirR/VirS regulon involves a regulatory RNA molecule for expression of toxins. Mol Microbiol
43: 257–265.
Sledjeski, D., and Gottesman, S. (1995) A small RNA acts as
an antisilencer of the H-NS-silenced rcsA gene of Escherichia coli. Proc Natl Acad Sci USA 92: 2003–2007.
Soper, T., Mandin, P., Majdalani, N., Gottesman, S., and
Woodson, S.A. (2010) Positive regulation by small RNAs
and the role of Hfq. Proc Natl Acad Sci USA 107: 9602–
9607.
Sun, H., Ringdahl, U., Homeister, J.W., Fay, W.P., Engleberg,
N.C., Yang, A.Y., et al. (2004) Plasminogen is a critical host
pathogenicity factor for group A streptococcal infection.
Science 305: 1283–1286.
Urban, J.H., and Vogel, J. (2008) Two seemingly homologous
noncoding RNAs act hierarchically to activate glmS mRNA
translation. PLoS Biol 6: e64.
Vecerek, B., Moll, I., and Bläsi, U. (2007) Control of
Fur synthesis by the non-coding RNA RyhB and ironresponsive decoding. EMBO J 26: 965–975.
Vecerek, B., Beich-Frandsen, M., Resch, A., and Bläsi, U.
(2010) Translational activation of rpoS mRNA by the
non-coding RNA DsrA and Hfq does not require ribosome
binding. Nucleic Acids Res 38: 1284–1293.
Waters, L.S., and Storz, G. (2009) Regulatory RNAs in
bacteria. Cell 136: 615–628.
Worhunsky, D.J., Godek, K., Litsch, S., and Schlax, P.J.
(2003) Interactions of the non-coding RNA DsrA and RpoS
mRNA with the 30S ribosomal subunit. J Biol Chem 278:
15815–15824.
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 1327–1331