A Glycine-Dependent Riboswitch That Uses Cooperative Binding to Control Gene Expression

See allHide authors and affiliations

Science  08 Oct 2004:
Vol. 306, Issue 5694, pp. 275-279
DOI: 10.1126/science.1100829

This article has a correction. Please see:


We identified a previously unknown riboswitch class in bacteria that is selectively triggered by glycine. A representative of these glycine-sensing RNAs from Bacillus subtilis operates as a rare genetic on switch for the gcvT operon, which codes for proteins that form the glycine cleavage system. Most glycine riboswitches integrate two ligand-binding domains that function cooperatively to more closely approximate a two-state genetic switch. This advanced form of riboswitch may have evolved to ensure that excess glycine is efficiently used to provide carbon flux through the citric acid cycle and maintain adequate amounts of the amino acid for protein synthesis. Thus, riboswitches perform key regulatory roles and exhibit complex performance characteristics that previously had been observed only with protein factors.

Genetic control by riboswitches located within the noncoding regions of mRNAs is widespread among bacteria (13). About 2% of the genes in Bacillus subtilis are regulated by these metabolite-binding RNA domains (4). All riboswitches discovered thus far use a single highly structured aptamer as a sensor for their corresponding target molecules. Selective binding of metabolite by the aptamer causes allosteric modulation of the secondary and tertiary structures of the mRNA 5′-untranslated region (5′-UTR), which changes gene expression by one or more mechanisms that influence transcription termination (5, 6), translation initiation (7, 8), or mRNA processing (9, 10).

The existence of riboswitches in modern cells implies that RNA molecules have considerable potential for forming intricate structures that are comparable to protein receptors. Furthermore, riboswitches do not have an obligate need for additional protein factors to carry out their gene control tasks and thus serve as economical genetic switches that sense and respond to changes in metabolite concentrations. However, higher-ordered functions exhibited by some protein factors have not been observed with natural riboswitches. For example, many protein enzymes, receptors, and gene control factors make use of cooperative binding to provide the cell with a means to rapidly respond to small changes in ligand concentrations [e.g., (1113)].

We recently reported the identification of highly conserved RNA motifs in numerous bacterial species that have features similar to known riboswitches (14). One of these motifs, termed gcvT (Fig. 1A), is found in many bacteria, where it typically resides upstream of genes that express protein components of the glycine cleavage system. In B. subtilis, a three-gene operon (gcvT-gcvPA-gcvPB) codes for components of this protein complex, which catalyzes the initial reactions for use of glycine as an energy source (15, 16).

Fig. 1.

Type I and type II gcvT motifs are natural RNA aptamers for glycine. (A) Consensus nucleotides present in more than 80% (black) and 95% (red) of representative sequences were identified by bioinformatics (17) (fig. S1). Circles and thick lines represent nucleotides whose base identities are not conserved. P1 through P4 identify common base-paired elements. ORF, open reading frame. (B) Patterns of spontaneous cleavage that occur with VC I-II in the absence and presence of glycine are depicted. Numbers adjacent to sites of changing spontaneous cleavage correspond to gel bands denoted with asterisks in (C) and data sets in (D). (C) Spontaneous cleavage products of VC I-II upon separation by polyacrylamide gel electrophoresis (PAGE) (7, 8) (fig. S2). NR, T1, and OH represent no reaction, partial digest with RNase T1, and partial digest with alkali, respectively. Pre, precursor RNA. Some fragment bands corresponding to T1 digestion (cleaves after G residues) are labeled. Numbered asterisks identify locations of major structural modulation in response to glycine. The two rightmost lanes carry 1 mM of the amino acids noted. Brackets labeled I and II identify RNA fragments that correspond to cleavage events in the type I and type II aptamers, respectively. (D) Plots of the extent of spontaneous cleavage products versus increasing concentrations of glycine for aptamer I (sites 1 through 3), aptamer II (sites 5 through 7), and the linker sequence (site 4). C, concentration.

Two forms of the gcvT RNA motif, type I and type II (Fig. 1A), had been identified on the basis of differences in the sequences that flank their conserved cores (14). More sensitive computational scans (17) revealed that both motif types reside adjacent to each other, as represented by the architecture of the region immediately upstream of the VC1422 gene (a putative sodium and alanine symporter) from Vibrio cholerae (Fig. 1B). Individually, the type I and type II elements appear to represent separate aptamer domains, wherein each presumably binds a separate target molecule. Furthermore, the linker sequence between the two aptamers exhibits some conservation of both sequence and length, suggesting that the aptamers are functionally coupled (fig. S1).

The metabolite-binding capabilities of V. cholerae RNAs were assessed by using a method termed inline probing (18), which can reveal metabolite-induced changes in aptamer structure by monitoring changes in the levels of spontaneous RNA cleavage (4, 6, 79). For example, the addition of glycine at 1 mM caused changes in the pattern of spontaneous cleavage of a 226-nucleotide RNA construct (VC I-II) that carries both aptamer types (Fig. 1C), whereas 1 mM l-alanine did not induce change.

Similar results were observed when a 105-nucleotide RNA (VC II) carrying the type II aptamer alone was used for inline probing (fig. S2). Because both type I and type II domains undergo similar structural changes upon introduction of glycine and because VC II alone exhibits ligand-dependent structural change, we conclude that each domain serves as a separate glycine binding aptamer. Furthermore, all three sections of the VC I-II construct (aptamer I, linker, and aptamer II) responded to glycine equally at various concentrations (Fig. 1D). This concerted response to glycine suggests the two aptamers either have perfectly matched affinities for glycine or bind glycine in a highly cooperative manner.

The molecular recognition specificity of VC I-II was examined by using inline probing with a variety of glycine analogs. The RNA exhibited measurable structural modulation with the methyl ester and tertiary butyl ester analogs of glycine but rejected all other analogs when tested at 1 mM (Fig. 2A). The concentrations of ligand needed to cause half-maximal structure modulation of VC II are about 10 μM for glycine, 100 μM for glycine methyl ester, 1 mM for glycine tertiary butyl ester, and 1 mM for glycine hydroxamate (19). Specificity for glycine also was observed by using equilibrium dialysis. For example, when an equilibrium dialysis system is preequilibrated with either VC II or VC I-II RNAs, excess glycine restored an equal distribution of 3H-glycine upon subsequent incubation (Fig. 2B). However, the addition of either l-alanine or l-serine failed to restore equal distribution, confirming that the RNAs serve as precise sensors for glycine.

Fig. 2.

Ligand specificity of VC II and VC I-II RNAs. (A) Inline probing of VC I-II in the absence (–) or presence of glycine (compound 1) or the analogs l-alanine (2), d-alanine (3), l-serine (4), l-threonine (5), sarcosine (6), mercaptoacetic acid (7), β-alanine (8), glycine methyl ester (9), glycine tert-butyl ester (10), glycine hydroxamate (11), glycinamide (12), aminomethane sulfonic acid (13), and glycyl-glycine (14). Other notations are as described in the legend to Fig. 1C. (B) Equilibrium dialysis data for VC II and VC I-II (100 μM) in the absence (–) or presence (+) of excess (1 mM) unlabeled glycine, alanine, or serine as indicated. Fraction of 3H-glycine in chamber b reflects the amount of glycine bound by RNA plus half the total amount of free glycine in chambers a and b versus the total amount of 3H-glycine. i to iii, separate experiments where RNA and 3H are equilibrated (left) and competitor is subsequently added.

We explored the stoichiometry of glycine binding to these RNAs by using equilibrium dialysis with high glycine concentrations. When three equivalents of the amino acid were present versus one equivalent of VC II RNA (100 μM), we observed a shift in glycine distribution (19) that indicates ∼0.8 equivalents (1 expected) of glycine were bound by RNA. In contrast, when one equivalent of the VC I-II RNA was present (two aptamer equivalents), there is a ∼1.6-fold increase (2 expected) in the amount of glycine that was bound by RNA. These data provide preliminary evidence for a stoichiometry of 1:1 between glycine and each individual aptamer.

Our laboratory created an engineered allosteric RNA construct with a tandem aptamer configuration that demonstrated cooperative binding of multiple ligands (20), thus providing a precedent for this more sophisticated form of RNA switch. If the two aptamers of VC I-II function cooperatively, then structural changes in the RNA should be atypically responsive to increasing glycine concentrations compared with those of a single glycine aptamer. The ligand-dependent modulation of VC II structure by glycine (Fig. 3) was typical of that observed for single aptamer domains of known riboswitches (4, 69, 2124). The change from ∼10% to ∼90% ligand-bound VC I RNA occurred over a ∼100-fold increase in glycine concentration, which corresponds with the response predicted for a receptor that binds a single ligand (fig. S3).

Fig. 3.

Cooperative binding of two glycine molecules by the VC I-II RNA. Plot depicts the fraction of VC II (open) and VC I-II (solid) bound to ligand versus the concentration of glycine. The constant, n, is the Hill coefficient for the lines as indicated that best fit the aggregate data from four different regions (fig. S3). Shaded boxes demark the dynamic range (DR) of glycine concentrations needed by the RNAs to progress from 10%- to 90%-bound states.

In contrast, VC I-II underwent the same change in ligand occupancy over only a ∼10-fold increase in glycine concentration (Fig. 3). This reduction in the dynamic range for the glycine-mediated response is consistent with the hypothesis that glycine binding at one site substantially improves the affinity for glycine binding to the other site. The Hill coefficient (25, 26) calculated for VC I-II is 1.64, whereas the maximum value for two binding sites is 2. In comparison, the Hill coefficient for the oxygen-carrying protein hemoglobin is 2.8 (27), whereas the maximum value for four binding sites is 4. Thus, the degree of cooperativity per binding site with the two VC I-II aptamers is equal to or greater than that derived for each of the four sites in hemoglobin.

A cooperative mechanism for ligand binding is further supported by the observation that single-point mutations made to either of the conserved cores of VC I-II cause substantial loss of glycine-binding affinity to the mutated aptamer and also cause a dramatic loss of affinity to the unaltered aptamer (fig. S4). Thus, the binding of glycine at one site induces the adjacent site to exhibit an improvement in ligand binding affinity by ∼100- to ∼1000-fold.

Tandem aptamer architecture (Fig. 4A) and selective glycine recognition (19) are also observed with RNA corresponding to the putative 5′-UTR of the gcvT operon from B. subtilis. This provided us with a construct that is more amenable to experiments that assess whether the gcvT RNA is important for genetic control. We used single-round transcription assays (17) to determine whether a DNA construct corresponding to the intergenic region (IGR) upstream of the B. subtilis gcvT operon yields transcripts whose termination sites are influenced by glycine. In the absence of glycine, only ∼30% of the RNA products generated by in vitro transcription were full-length (Fig. 4B). The remaining ∼70% were premature termination products that correspond in length to that expected if RNA polymerase stalls at a putative intrinsic terminator (28, 29) that partially overlaps the second glycine aptamer (also fig. S5).

Fig. 4.

Control of B. subtilis gcvT RNA expression in vitro and in vivo. (A) The IGR between the yqhH and gcvT genes of B. subtilis encompassing both aptamers I and II was used for in vitro transcription and in vivo expression assays. Inline probing results were mapped, and mutations used to assess riboswitch function are indicated with red boxes. Orange shading identifies the putative intrinsic terminator stem, which is expected to exhibit mutually exclusive formation of aptamer II when bound to glycine. nt, nucleotide. (B) Single-round in vitro transcription assays demonstrating that full-length (Full) transcripts are favored when >10 μM glycine is added to the transcription mixture, whereas serine and most glycine analogs (fig. S5) are rejected by the riboswitch. The line reflects a best-fit curve to an equation reflecting cooperative binding with a Hill coefficient of 1.4 (19). An additional transcription product, termed “+,” appears to be due to spurious transcription initiation (17). (C) Plot of the expression of a β-galactosidase reporter gene fused to wild-type (WT) gcvT IGR or to a series of mutant IGRs (M1–M6). Data reflect the averages of three assays with two replicates each. Error bars indicate ± two standard deviations.

The addition of glycine caused a substantial increase in the amount of full-length RNA transcript relative to the amount of truncated RNA (Fig. 4B). This improvement is induced only by glycine or by other analogs that cause RNA structure modulation. Compounds such as serine, alanine, and other analogs that do not induce modulation also failed to trigger an increase in the production of full-length transcripts (fig. S5) (19).

Furthermore, the glycine-dependent increase in the yield of full-length transcripts corresponded with that expected for a cooperative RNA switch requiring two ligand binding events. Fitting the transcription data yields a curve that corresponded to cooperative ligand binding, with a Hill coefficient of 1.4 (Fig. 4B). Therefore, transcription control by the gcvT 5′-UTR of B. subtilis responds to glycine with characteristics that parallel those observed when conducting inline probing of the cooperative VC I-II RNA.

To assess whether glycine binding and in vitro transcription control correspond to genetic control events in vivo, we generated reporter constructs by fusing the IGR up stream of the gcvT operon from B. subtilis to a β-galactosidase reporter gene and integrated them into the bacterial genome (17). The reporter fusion construct carrying the wild-type IGR expresses a high amount of β-galactosidase when glycine is present in the growth medium, whereas a low amount of gene expression results when alanine is present (Fig. 4C). These results indicate that the gcvT motif is part of a glycine-responsive riboswitch with a default state that is off. Glycine binding is required to activate gene expression, as was also observed with the in vitro transcription assays (Fig. 4B).

The importance of several conserved features of the motif were examined by mutating the P1 and P2 stems of the first aptamer domain to disrupt (variants M1 and M3, respectively) and restore (M2 and M4, respectively) base pairing (Fig. 4A). Resulting gene expression levels from constructs carrying the mutant IGRs are consistent with base-paired elements predicted from phylogenetic analyses (14) (fig. S1). Furthermore, the introduction of mutations into the conserved cores of either aptamer I or aptamer II (variants M5 and M6, respectively) caused a complete loss of reporter gene activation. This latter result suggests that glycine binding to both aptamers is necessary to trigger gene activation, which is consistent with a model wherein cooperative glycine binding is important for riboswitch function.

The glycine-dependent riboswitch is a remarkable genetic control element for several reasons. First, glycine riboswitches form selective binding pockets for a ligand composed of only 10 atoms and thus bind the smallest organic compound among known natural and engineered RNA aptamers. This observation is consistent with the hypothesis that RNA has sufficient structural potential to selectively bind a wide range of biomolecules.

Second, the 5′-UTR of the B. subtilis gcvT operon is a genetic on switch, and thus joins the adenine riboswitch (23) as a rare type of RNA that has been proven to harness ligand binding and activate gene expression. In most instances, riboswitches cause repression of their associated genes, which is to be expected because many of these genes are involved in biosynthesis or import of the target metabolites. However, the glycine riboswitch from B. subtilis controls the expression of three genes required for glycine degradation. A ligand-activated riboswitch would be required to determine whether sufficient amino acid substrate is present to warrant production of the glycine cleavage system, thereby providing a rationale for why this rare on switch is used.

Third, this is the only known metabolite-binding riboswitch class that regularly makes use of a tandem aptamer configuration. In both V. cholerae and B. subtilis, the juxtaposition of aptamers enables the cooperative binding of two glycine molecules. For the B. subtilis riboswitch, this characteristic is expected to result in unusually rapid activation and repression of genes encoding the glycine cleavage system in response to rising and falling concentrations of glycine, respectively. Given the prevalence of the tandem architecture of glycine riboswitches, this more “digital” switch likely gives the bacterium an important selective advantage by controlling gene expression in response to small changes in glycine.

Supporting Online Material

Materials and Methods

Figs. S1 to S5

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article