• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Apr 3, 2012; 109(14): 5376–5381.
Published online Mar 19, 2012. doi:  10.1073/pnas.1112211109
PMCID: PMC3325659

Riboswitch control of Rho-dependent transcription termination


Riboswitches are RNA sensors that regulate gene expression upon binding specific metabolites or ions. Bacterial riboswitches control gene expression primarily by promoting intrinsic transcription termination or by inhibiting translation initiation. We now report a third general mechanism of riboswitch action: governing the ability of the RNA-dependent helicase Rho to terminate transcription. We establish that Rho promotes transcription termination in the Mg2+-sensing mgtA riboswitch from Salmonella enterica serovar Typhimurium and the flavin mononucleotide-sensing ribB riboswitch from Escherichia coli when the corresponding riboswitch ligands are present. The Rho-specific inhibitor bicyclomycin enabled transcription of the coding regions at these two loci in bacteria experiencing repressing concentrations of the riboswitch ligands in vivo. A mutation in the mgtA leader that favors the “high Mg2+” conformation of the riboswitch promoted Rho-dependent transcription termination in vivo and in vitro and enhanced the ability of the RNA to stimulate Rho's ATPase activity in vitro. These effects were overcome by mutations in a C-rich region of the mRNA that is alternately folded at high and low Mg2+, suggesting a role for this region in regulating the activity of Rho. Our results reveal a potentially widespread mode of gene regulation whereby riboswitches dictate whether a protein effector can interact with the transcription machinery to prematurely terminate transcription.

The 5′ leader regions of many bacterial messenger RNAs contain regulatory elements—termed riboswitches—that change expression of the associated protein-coding regions upon binding specific metabolites or ions (1, 2). Ligand binding promotes the formation of RNA structures that typically turn off gene expression by creating an intrinsic transcription terminator, or by occluding the ribosome binding site and/or start codon to inhibit translation initiation (1, 2). However, an increasing number of riboswitches lack these gene control elements, suggesting that they use alternative mechanisms to regulate gene expression (3). Here we identify riboswitches that function by controlling the activity of the transcription termination factor Rho.

The Mg2+-sensing mgtA riboswitch from Salmonella enterica serovar Typhimurium regulates transcription elongation into the mgtA coding region by adopting two mutually exclusive conformations in response to cytoplasmic Mg2+ levels (4). Stem-loops A and B are favored in high Mg2+, which hinders transcription elongation into the mgtA coding region, whereas stem-loop C forms in low Mg2+, which brings about transcription of the coding region of the Mg2+ transporter gene mgtA (4) (Fig. 1A). However, the mechanism by which this riboswitch operates has remained unknown because the mgtA leader mRNA sequence lacks the GC-rich hairpin followed by a series of U residues that is typical of intrinsic transcription terminators (5).

Fig. 1.
A Rho-dependent terminator in the Salmonella mgtA leader prevents transcription into the mgtA coding region during growth at high Mg2+. (A) Schematic showing the secondary structure of the mgtA leader mRNA. High Mg2+ promotes formation of stem-loops A ...

Recent genomic analyses have revealed that bacteria harbor small RNA (sRNA) species corresponding to the 5′ leader regions of transcripts containing known or predicted riboswitch ligand-binding domains but lacking intrinsic transcription terminators (68). For instance, the sRNA sroG originates from within the leader region of the Escherichia coli ribB gene, which includes a predicted flavin mononucleotide (FMN)-binding riboswitch (rfn-box) but lacks sequences resembling a canonical intrinsic terminator (9, 10). This raises questions about the origin and function of these sequences.

In contrast to intrinsic terminators, which can dissociate transcription complexes in the absence of accessory proteins, the second major class of bacterial transcription terminators is recognized only in the presence of Rho, a homohexameric RNA helicase responsible for termination of 20–50% of the transcription units in E. coli (5, 1113). Rho-dependent termination requires Rho binding to untranslated regions of the nascent transcript at sites that are generally rich in C residues and relatively unstructured (5, 13). This stimulates Rho's RNA-dependent ATPase activity, resulting in Rho translocation along the nascent RNA in the 5′ to 3′ direction until Rho makes interactions with RNA polymerase (RNAP) that promote dissociation of the transcription elongation complex, which typically takes place at sites of RNAP pausing (5, 13).

We now report that both the mgtA and ribB riboswitches govern transcription elongation into their respective downstream coding regions by controlling the ability of Rho to terminate transcription. This represents a distinct mode of gene regulation, which may be widespread given that other riboswitches are found in leader regions where transcription can terminate before the corresponding coding region but which lack obvious intrinsic terminators (68).

Results and Discussion

Rho-Dependent Terminator in the Salmonella mgtA Leader Inhibits Transcription into the mgtA Coding Region at High Mg2+.

Because the Salmonella mgtA leader region does not contain sequences typical of an intrinsic transcription terminator (Fig. 1A) (4), we hypothesized that transcription termination in this leader might require Rho. To test this hypothesis, we monitored the effect of the Rho-specific inhibitor bicyclomycin (BCM) (14) on the β-galactosidase activity produced by wild-type Salmonella harboring a plasmid-borne transcriptional fusion between the mgtA leader and a promoterless E. coli lacZ gene driven by the Mg2+-insensitive plac1–6 promoter (4). [We chose a BCM concentration of 20 μg/mL so as to limit effects on bacterial growth. However, the chosen concentration may allow some limited Rho-dependent transcription termination to take place because total inhibition of Rho is lethal (11, 13).]

BCM overcame the transcriptional silencing taking place during growth in high Mg2+ (i.e., BCM promoted a 10-fold increase in the β-galactosidase activity relative to cells grown in the absence of this inhibitor) (Fig. 1B). BCM had a much smaller effect on the β-galactosidase activity of bacteria grown in low Mg2+ (i.e., twofold; Fig. 1B), possibly reflecting that the leader in most mgtA transcripts would be in the “low Mg2+” conformation that favors transcription elongation into the coding region (4), and that Rho is expected to act on the transcript subpopulation that is in the “high Mg2+” conformation. Rho exerts its regulatory effects via the mgtA leader because BCM did not alter the β-galactosidase activity in bacteria carrying a control plasmid harboring the plac1–6-driven lacZ without the mgtA leader (Fig. 1B, Inset).

We could rule out that the observed derepression was an artifact arising from transcription initiated from a heterologous promoter or the use of a plasmid-encoded mgtA-lacZ fusion because BCM heightened β-galactosidase levels in a strain with a chromosomal mgtA-lac fusion transcribed from the native mgtA promoter, but not in one with a chromosomal mgtA-lac fusion not preceded by the mgtA leader (Fig. 1C). In both strains, transcription initiation from the mgtA promoter is controlled by the transcription factor PhoP in response to extracytoplasmic Mg2+ sensed by the PhoQ sensor kinase (15). Hence, transcription fails to initiate efficiently at 10 mM MgCl2 because the PhoP/PhoQ system is inactive. When bacteria are grown in 0.5 mM MgCl2, transcription initiates at the mgtA promoter but does not continue into the downstream coding sequence when the mgtA leader is present because it adopts the high Mg2+ conformation that results in transcription stopping within the leader (4). During growth in 0.01 mM MgCl2, the mgtA leader adopts the low Mg2+ conformation that permits transcription elongation into the coding region (4).

To directly monitor transcription termination in the mgtA leader, we carried out single-round in vitro transcription assays using purified RNAP. Without Rho, RNAP was unable to terminate transcription within the mgtA leader and instead entered an unusually long-lived paused state (Fig. 2A). Pausing results in initial accumulation of a truncated transcript that subsequently disappears as RNAP escapes the pause and elongates the RNA to produce a run-off transcript (Fig. 2A, Lower). Our data suggest that the previously reported truncated RNA generated during in vitro transcription of the mgtA leader (4, 16) results from an extended pause rather than termination of transcription.

Fig. 2.
Mg2+ stimulates Rho-dependent transcription termination in the mgtA leader by promoting an RNA conformation that enhances Rho's catalytic activity. (A) Upper: In vitro transcription of a template corresponding to the wild-type Salmonella mgtA leader driven ...

Addition of purified Rho protein to the in vitro transcription reactions stimulated termination at the same site as that corresponding to the transcriptional pause in the mgtA leader (Fig. 2A). In contrast to pausing, transcription termination generates a truncated transcript that remains for the duration of the reaction because it can no longer be elongated after dissociation from RNAP (Fig. 2A, Lower). The observed band corresponds to a true termination product because a high-salt wash removed it from transcription complexes immobilized on Ni-NTA beads (Fig. S1).

High Mg2+ enhanced Rho-dependent termination in the mgtA leader in vitro because termination was more efficient at 3.5 mM than at 0.35 mM Mg2+ (Fig. 2 B and C). The stimulation of Rho-dependent termination by high Mg2+ is specific to the mgtA leader because high Mg2+ did not enhance termination (and actually decreased it) at the well-characterized Rho-dependent terminator λtR1 (Fig. S2A), which was used as a negative control because it does not respond to changes in the Mg2+ concentration in vivo (Fig. S2B). Although the in vitro effect of Mg2+ on the mgtA leader is modest (Fig. 2 B and C), it is of similar magnitude to the Mg2+-stimulated increase in intrinsic transcription termination reported for the unrelated Bacillus subtilis mgtE leader, which harbors the only other known Mg2+ sensing riboswitch (17). For the mgtA leader, we observed readthrough values of 55–60% at 3.5 mM Mg2+ and 70–80% at 0.35 mM Mg2+ in the presence of Rho (Fig. 2 B and C). These are comparable to the readthrough values reported for the mgtE leader (17), which were ≈63% at 3 mM Mg2+ and ≈75% at 1.5 mM Mg2+ (the lowest Mg2+ concentration used in that study). [Note that the range over which the Mg2+ concentration can be varied in vitro is limited because the overall efficiency of transcription initiation and elongation is sensitive to Mg2+. This may explain, in part, why the effect of Mg2+ in vitro is smaller than that observed in vivo for both the mgtA (Figs. 1B and and2B)2B) and mgtE (17) riboswitches.] Taken together, our data establish that the mgtA leader contains a Rho-dependent terminator that prevents transcription elongation into the mgtA coding region during growth in high Mg2+.

High Mg2+ Promotes a Conformation of the mgtA Leader That Favors Interaction with Rho.

To terminate transcription, Rho must first associate with the nascent transcript (5, 13). Thus, the RNA structure promoted in high Mg2+ may stimulate Rho-dependent termination by facilitating interaction of Rho with the mgtA leader RNA. To test this hypothesis, we took a genetic approach that made use of a mutant mgtA leader that mimics the high Mg2+ conformer. This mutant contains the C145G substitution that creates a mismatch in stem-loop C—a structure that forms at low Mg2+ and is required for transcription elongation into the coding region (4)—but does not interfere with formation of stem-loops A and B (Fig. 1A).

The C145G mutation results in constitutive Rho-dependent termination in vivo (Fig. 1B) and enhances Rho-dependent termination in vitro (Fig. 2C). To test whether the C145G mutation directly affects the ability of the RNA to interact with Rho, we monitored Rho's ATPase activity after incubation of wild-type and C145G mutant mgtA leader RNAs with purified Rho protein in vitro. Rho's ATPase activity was promoted to higher levels by the C145G RNA than by the wild-type mgtA leader RNA (Fig. 2D), suggesting that the high Mg2+ conformer interacts more efficiently with Rho. The ATPase activity increased as the RNA concentration increased in reactions carried out with the C145G mutant RNA; however, a small decrease in ATPase stimulation was observed at the three highest RNA concentrations in assays performed with the wild-type mgtA leader RNA (Fig. 2D). This could reflect that when the concentration of the wild-type mgtA leader RNA is increased, there is less Mg2+ available to bind each RNA molecule, and this would result in a rise in the proportion of RNAs adopting the low Mg2+ conformation, which is less able to stimulate Rho's ATPase activity. By contrast, the C145G mutant RNA would not be sensitive to this decrease in Mg2+ availability because it is genetically locked into the high Mg2+ conformation. Taken together, our data suggest that high Mg2+ stimulates Rho-dependent termination in the mgtA leader by promoting a conformation in the leader mRNA that favors interaction with Rho.

Riboswitch Sequences Required for Rho-Dependent Termination in the mgtA Leader.

We sought to identify sequences within the mgtA leader RNA that are required for Rho-dependent transcription termination. Rho's RNA substrates are usually rich in C residues and relatively unstructured (5, 13). Thus, we targeted the C residues in the R1 and R2 regions flanking stem-loop B for mutagenesis because these regions are present within predicted stem-loop structures in the low Mg2+ conformer but are exposed in single-stranded regions in high Mg2+ (4) (Fig. 1A) and thus are available to interact with Rho; and also because the sequences downstream of +126 (the start of R1) are sufficient to promote Rho-dependent termination in vivo (Fig. S3A) and in vitro (Fig. S3B).

If the nucleotides in the R1 and R2 regions of the mgtA leader are essential for stimulation of Rho activity, mutations in R1 or R2 should overcome the silencing effect of the C145G substitution. As predicted, mutation of the R1 region enhanced expression 30- and 39-fold with respect to the parental C145G mutant after growth in low and high Mg2+, respectively (Fig. 1B); the effect of the R2 mutation was less pronounced (Fig. 1B). The R1 and R2 mutations interfere specifically with Rho-dependent transcription termination because their effects on expression relative to C145G were much smaller in cells that were treated with BCM (Fig. 1B). Importantly, mutation of the R1 region reduced the efficiency of transcription termination by Rho in vitro (Fig. 2C) but had no effect in the absence of Rho (Fig. 2C and Fig. S4), indicating that mutation of R1 affects termination but not pausing by RNAP. The R1 mutation hinders Rho-dependent termination by preventing the mgtA leader RNA from establishing a productive interaction with Rho because it decreased the ability of the C145G RNA to stimulate Rho's ATPase activity 2.5- to threefold in vitro, to values that were even lower than those promoted by the wild-type mgtA leader RNA (Fig. 2D). The effect of the R1 mutation is not specific to a leader that is genetically “locked” into the high Mg2+ conformation because mutation of the R1 region in an otherwise wild-type mgtA leader overcame the repressive effect of growth in high Mg2+ on transcription through the leader in vivo (Fig. 1B), inhibited Rho-dependent termination in vitro (Fig. 2C), and reduced the ability of the mgtA leader RNA to stimulate Rho's ATPase activity in vitro (Fig. 2D).

How the mgtA Riboswitch Controls Rho-Dependent Termination.

Our data support a model in which high Mg2+ favors an RNA conformation that permits Rho to interact with R1 and possibly neighboring single-stranded regions. This would stimulate ATP hydrolysis and Rho translocation along the nascent transcript, allowing Rho to interact with paused RNAP to terminate transcription before the mgtA coding region. By contrast, under low Mg2+ conditions, Rho does not seem to make the contacts that activate its ATPase activity because these interactions require the R1 region, which is sequestered into a stem-loop structure in the low Mg2+ conformer (Fig. 1A). Hence RNAP escapes from the downstream pause site before Rho can terminate transcription.

The mgtA leader harbors a short, proline-rich ORF—termed mgtL—the translation of which hinders transcription elongation into the mgtA coding region (Fig. 1A) (16, 18). When proline levels are limiting in the cytosol, ribosome stalling at the proline codons in mgtL favors transcription elongation into the mgtA coding region (18). This would inhibit Rho-dependent termination by the same mechanism as that taking place when cytosolic Mg2+ is low because ribosome stalling is predicted to favor formation of stem-loop C, thus sequestering the R1 region and preventing it from interacting with Rho. By contrast, complete translation of mgtL would promote Rho-dependent transcription termination by the same mechanism as that taking place when cytosolic Mg2+ is high because a ribosome translating the full mgtL ORF would occlude the left arm of stem-loop C, thereby favoring formation of stem-loop B and freeing the R1 region for interaction with Rho. Consistent with this idea, mutation of the R1 region in an otherwise wild-type mgtA leader resulted in derepression under noninducing Mg2+ conditions and in the presence of casamino acids (Fig. 1B) (i.e., conditions whereby mgtL would be fully translated because proline is plentiful).

Riboswitch in the E. coli ribB Leader Promotes Rho-Dependent Termination in the Presence of FMN.

We posited that the FMN-sensing riboswitch located in the E. coli ribB leader might also function by controlling Rho-dependent termination because this leader encodes a known sRNA but lacks an obvious intrinsic terminator (Fig. 3A) (9, 10). As hypothesized, BCM overcame the fivefold repression that the FMN precursor riboflavin exerted on β-galactosidase activity in a strain harboring a transcriptional fusion between wild-type ribB and a promoterless lacZ gene, but had little effect in the absence of riboflavin (Fig. 3B). Moreover, primer extension and quantitative real-time PCR data demonstrated that BCM treatment enhanced RNA levels corresponding to the ribB coding region eightfold (Fig. S5 A and B). [BCM treatment resulted in a smaller (i.e., 3.5-fold) increase in the level of the leader region of the ribB mRNA (Fig. S5B), suggesting readthrough from an upstream promoter.] FMN stimulates Rho-dependent termination directly, because addition of FMN to in vitro transcription reactions enhanced Rho's ability to terminate transcription in the wild-type ribB leader but had no effect in the absence of Rho (Fig. 3C and Fig. S6). Taken together, these data establish that the ribB leader contains a Rho-dependent terminator that inhibits transcription elongation into the ribB coding region in the presence of FMN. Our findings now explain why the distribution of RNAP shifted from the ribB leader into the ribB coding region when E. coli was treated with BCM (12).

Fig. 3.
The E. coli ribB leader contains a riboswitch that promotes Rho-dependent termination in the presence of FMN. (A) Schematic showing the predicted secondary structure of the ribB leader RNA (9). In the absence of FMN, nucleotides at positions 9–15 ...

To investigate whether alternate folding of the ribB riboswitch in response to the availability of FMN controls the ability of Rho to terminate transcription, we examined the effect of mutations that favor different riboswitch conformers on Rho-dependent termination in vivo and in vitro. Mutations in the M1 region, which are predicted to favor the “+FMN” conformation (Fig. 3A), stimulated Rho-dependent termination in vitro even in the absence of FMN (Fig. 3C) and reduced in vivo expression from a ribB-lacZ transcriptional fusion in the absence of riboflavin to the same level displayed by a strain with the wild-type leader when incubated in the presence of riboflavin (Fig. 3B). Conversely, mutations in the M2 region, which are predicted to favor the “-FMN” conformation (Fig. 3A), prevented FMN from stimulating Rho-dependent transcription termination in vitro (Fig. 3C) and hindered riboflavin-mediated repression in vivo (Fig. 3B). These data indicate that FMN inhibits transcription elongation into the ribB coding region by promoting an RNA conformation that enhances Rho-dependent transcription termination.

Mechanisms of Gene Control by FMN-Sensing Riboswitches.

It has been suggested that riboswitches primarily regulate transcription in Gram-positive bacteria but translation in Gram-negative organisms (9, 19). Indeed, the FMN-sensing riboswitch from the Gram-positive bacterium Bacillus subtilis controls expression of riboflavin biosynthetic genes via early intrinsic termination of transcription (20, 21), whereas FMN riboswitches in E. coli and other Gram-negative bacteria were postulated to work primarily via translational repression (9). Even though Rho has been proposed to play an indirect role in gene control by riboswitches that affect translation of downstream genes [i.e., by prematurely terminating transcription of those mRNAs that are not translated (2, 12)], our results demonstrate that the E. coli ribB riboswitch operates in a distinct manner. Apart from regulating translation, the ribB riboswitch directly controls transcription termination within the leader region because FMN clearly stimulated Rho-dependent termination (Fig. 3 B and C) in the absence of the translation machinery (Fig. 3C). Moreover, in the presence of FMN, Rho-dependent termination takes place upstream of the ribB start codon (Fig. S6).

Concluding Remarks.

We have identified a general mode of riboswitch-mediated gene control that, in contrast to those described previously, relies not only on RNA but also on the essential RNA helicase Rho. We determined that two riboswitches originating from separate species and sensing different ligands act by directly modulating Rho's ability to terminate transcription, which, in turn, dramatically affects expression of the associated coding regions.

Rho is a general termination factor in bacteria that travels with RNAP throughout the transcription cycle, interrupting transcription of unprotected RNAs (22, 23). Our data now suggest that, by monitoring the conformational state of the nascent RNA, Rho can function as the effector of riboswitch gene control. This expands the physiological functions ascribed to Rho, which include generation of transcript 3′ ends (24), establishment of transcriptional polarity (25), silencing of horizontally acquired DNA (11), resolution of R-loops (26), and chromosome protection from double-stranded DNA breaks (27, 28). Given that pervasive Rho-dependent termination has been detected in bacterial genomes (11, 12) and that sRNA species can arise from the 5′ leader regions of transcripts lacking canonical intrinsic terminators (68), we propose that riboswitches yet to be associated with a particular control mechanism, as well as those classified by computational methods as regulating translation initiation when intrinsic transcription termination signals were not found (19), may actually function by governing Rho-dependent transcription termination.

Materials and Methods

Bacterial Strains, Plasmid Constructs, Primers, and Growth Conditions.

Bacterial strains and plasmids used in this study are listed in Table S1. Details of plasmid constructions are described in SI Materials and Methods, and primers are listed in Table S2. Salmonella strains were grown in N-minimal medium (pH 7.4) (29) supplemented with 0.1% casamino acids, 38 mM glycerol, and the indicated concentrations of MgCl2. Chloramphenicol (20 μg/mL) was added for growth of strains harboring derivatives of plasmid pYS1000. E. coli strains were grown in LB medium. Ampicillin (50 μg/mL) was used for growth of strains harboring derivatives of pJEL250.

β-Galactosidase Assays.

To monitor expression of mgtA-lacZ fusions in Salmonella, bacteria were grown overnight in N-minimal medium containing 10 mM MgCl2, washed twice in medium lacking MgCl2, and used to inoculate fresh medium containing the indicated concentrations of MgCl2. After incubation for 4 h at 37 °C, cultures were split in two, BCM was added to one tube at 20 μg/mL, and incubation was continued for 30 min before measurement of β-galactosidase activity as described (4). To monitor expression of ribB-lacZ fusions in E. coli, bacteria were grown overnight at 30 °C in LB medium containing 50 μM riboflavin, washed in medium lacking riboflavin, and diluted 1:25 in fresh medium containing riboflavin and/or BCM at the specified concentrations. Cultures were incubated for 2–2.5 h at 30 °C (OD600 = 0.5–0.6) before measurement of β-galactosidase activity. Shown β-galactosidase activities correspond to mean values from at least three independent experiments. Error bars correspond to the SD.

In Vitro Transcription Assays.

For in vitro transcription of plac1–6-driven mgtA and λtR1 templates, 5 pmol His6-RNAP was incubated with 10 pmol of PCR-generated template DNA in 35 μL transcription buffer (TB) containing 1 mM MgCl2 for 10 min at 37 °C, followed by addition of 50 μL Ni-NTA resin (Qiagen) and incubation for 15 min at 37 °C. Beads were washed in TB (no MgCl2) and resuspended in 80 μL TB, followed by addition of MgCl2 (as indicated), 0.4 μM Rho, 48 μM GTP, CTP, and ATP, 15 μM UTP, 10 μg/mL rifampicin, and 10 μCi [α-32P]-UTP and incubation at 37 °C. Samples were removed at different times and analyzed by denaturing gel electrophoresis. For in vitro transcription of PCR-generated ribB templates, initial elongation complexes (EC) were obtained as previously described (22) using 10 μM GpCpU RNA primer and 25 μM CTP, UTP, and ATP. After radiolabeling with [α-32P]-CTP, the resulting start-up EC25 was washed with 1 mL TB50Mg3, divided into four aliquots (10 μL), mixed with 0.1 mM FMN and/or 0.4 μM Rho or mock solution, and incubated for 5 min at room temperature. Samples were chased by 10 μM NTPs and 1 mM ATP for 10 min at 37° and analyzed by denaturing gel electrophoresis. A detailed description of the in vitro transcription assays is provided in SI Materials and Methods.

Rho ATPase Assays.

RNA corresponding to wild-type or mutant derivatives of the mgtA leader was synthesized with the Megascript T7 kit (Ambion) according to the manufacturer's instructions, using T7 promoter-driven templates generated by PCR with primers 6712 and 11882 and plasmid pYS1010 or derivatives thereof. RNAs were purified from 6% TBE urea gels and ethanol precipitated. Rho ATPase activity was determined with the EnzChek Phosphate Assay kit (Invitrogen) using a modification of a published protocol (30) in the presence of different RNA substrates. Briefly, RNAs were heated for 3 min at 95 °C in the presence of 0.75 mM MgCl2 and renatured for 1 h at room temperature. RNA (at the specified concentration) was then mixed in a 96-well plate with 5 nM Rho in a reaction mixture containing 0.2 mM 2-amino-6-mercapto-7-methylpurine riboside and 1 U/mL purine nucleoside phosphorylase in ATPase buffer [50 mM KCl, 0.75 mM MgCl2, 20 mM Hepes (pH 7.9), 0.1 mM EDTA, and 0.1 mM DTT]. After 10 min at room temperature, reactions were initiated by addition of 1 mM ATP and incubated at 37 °C in a SpectraMax Plus microplate reader that allowed detection of inorganic phosphate release by measuring absorbance at 360 nm over time. ATPase activity was determined by calibration against a standard curve generated using incubations with different concentrations of inorganic phosphate, as described in the manufacturer's instructions. Data shown are averages from at least three independent experiments. Error bars show the SD.

Supplementary Material

Supporting Information:


We thank Barbara Stitt for supplying purified Rho protein, Max Gottesman for providing bicyclomycin, and Charles Turnbough and Max Gottesman for comments on a previous version of the manuscript. This work was supported in part by National Institutes of Health Grants AI49561 (to E.A.G.) and GM58750 (to E.N.), the Dynasty Foundation (E.N.), and Russian Foundation for Basic Research Grant 11-04-01005-a (to A.M.). E.A.G. is an investigator of the Howard Hughes Medical Institute.


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1112211109/-/DCSupplemental.


1. Henkin TM. Riboswitch RNAs: Using RNA to sense cellular metabolism. Genes Dev. 2008;22:3383–3390. [PMC free article] [PubMed]
2. Roth A, Breaker RR. The structural and functional diversity of metabolite-binding riboswitches. Annu Rev Biochem. 2009;78:305–334. [PubMed]
3. Bastet L, Dubé A, Massé E, Lafontaine DA. New insights into riboswitch regulation mechanisms. Mol Microbiol. 2011;80:1148–1154. [PubMed]
4. Cromie MJ, Shi Y, Latifi T, Groisman EA. An RNA sensor for intracellular Mg(2+) Cell. 2006;125:71–84. [PubMed]
5. Peters JM, Vangeloff AD, Landick R. Bacterial transcription terminators: The RNA 3′-end chronicles. J Mol Biol. 2011;412:793–813. [PMC free article] [PubMed]
6. Irnov I, Sharma CM, Vogel J, Winkler WC. Identification of regulatory RNAs in Bacillus subtilis. Nucleic Acids Res. 2010;38:6637–6651. [PMC free article] [PubMed]
7. Livny J, Waldor MK. Mining regulatory 5'UTRs from cDNA deep sequencing datasets. Nucleic Acids Res. 2010;38:1504–1514. [PMC free article] [PubMed]
8. Raghavan R, Groisman EA, Ochman H. Genome-wide detection of novel regulatory RNAs in E. coli. Genome Res. 2011;21:1487–1497. [PMC free article] [PubMed]
9. Vitreschak AG, Rodionov DA, Mironov AA, Gelfand MS. Regulation of riboflavin biosynthesis and transport genes in bacteria by transcriptional and translational attenuation. Nucleic Acids Res. 2002;30:3141–3151. [PMC free article] [PubMed]
10. Vogel J, et al. RNomics in Escherichia coli detects new sRNA species and indicates parallel transcriptional output in bacteria. Nucleic Acids Res. 2003;31:6435–6443. [PMC free article] [PubMed]
11. Cardinale CJ, et al. Termination factor Rho and its cofactors NusA and NusG silence foreign DNA in E. coli. Science. 2008;320:935–938. [PMC free article] [PubMed]
12. Peters JM, et al. Rho directs widespread termination of intragenic and stable RNA transcription. Proc Natl Acad Sci USA. 2009;106:15406–15411. [PMC free article] [PubMed]
13. Rabhi M, Rahmouni AR, Boudvillain M. In: RNA Helicases. Jankowsky E, editor. Cambridge, UK: Royal Society of Chemistry; 2010. pp. 243–271.
14. Kohn H, Widger W. The molecular basis for the mode of action of bicyclomycin. Curr Drug Targets Infect Disord. 2005;5:273–295. [PubMed]
15. Cromie MJ, Groisman EA. Promoter and riboswitch control of the Mg2+ transporter MgtA from Salmonella enterica. J Bacteriol. 2010;192:604–607. [PMC free article] [PubMed]
16. Zhao G, Kong W, Weatherspoon-Griffin N, Clark-Curtiss J, Shi Y. Mg2+ facilitates leader peptide translation to induce riboswitch-mediated transcription termination. EMBO J. 2011;30:1485–1496. [PMC free article] [PubMed]
17. Dann CE, 3rd, et al. Structure and mechanism of a metal-sensing regulatory RNA. Cell. 2007;130:878–892. [PubMed]
18. Park SY, Cromie MJ, Lee EJ, Groisman EA. A bacterial mRNA leader that employs different mechanisms to sense disparate intracellular signals. Cell. 2010;142:737–748. [PMC free article] [PubMed]
19. Barrick JE, Breaker RR. The distributions, mechanisms, and structures of metabolite-binding riboswitches. Genome Biol. 2007;8:R239. [PMC free article] [PubMed]
20. Mironov AS, et al. Sensing small molecules by nascent RNA: A mechanism to control transcription in bacteria. Cell. 2002;111:747–756. [PubMed]
21. Winkler WC, Cohen-Chalamish S, Breaker RR. An mRNA structure that controls gene expression by binding FMN. Proc Natl Acad Sci USA. 2002;99:15908–15913. [PMC free article] [PubMed]
22. Epshtein V, Dutta D, Wade J, Nudler E. An allosteric mechanism of Rho-dependent transcription termination. Nature. 2010;463:245–249. [PMC free article] [PubMed]
23. Mooney RA, et al. Regulator trafficking on bacterial transcription units in vivo. Mol Cell. 2009;33:97–108. [PMC free article] [PubMed]
24. Roberts JW. Termination factor for RNA synthesis. Nature. 1969;224:1168–1174. [PubMed]
25. Richardson JP, Grimley C, Lowery C. Transcription termination factor rho activity is altered in Escherichia coli with suA gene mutations. Proc Natl Acad Sci USA. 1975;72:1725–1728. [PMC free article] [PubMed]
26. Harinarayanan R, Gowrishankar J. Host factor titration by chromosomal R-loops as a mechanism for runaway plasmid replication in transcription termination-defective mutants of Escherichia coli. J Mol Biol. 2003;332:31–46. [PubMed]
27. Washburn RS, Gottesman ME. Transcription termination maintains chromosome integrity. Proc Natl Acad Sci USA. 2011;108:792–797. [PMC free article] [PubMed]
28. Dutta D, Shatalin K, Epshtein V, Gottesman ME, Nudler E. Linking RNA polymerase backtracking to genome instability in E. coli. Cell. 2011;146:533–543. [PMC free article] [PubMed]
29. Snavely MD, Miller CG, Maguire ME. The mgtB Mg2+ transport locus of Salmonella typhimurium encodes a P-type ATPase. J Biol Chem. 1991;266:815–823. [PubMed]
30. Rabhi M, et al. Mutagenesis-based evidence for an asymmetric configuration of the ring-shaped transcription termination factor Rho. J Mol Biol. 2011;405:497–518. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...