Point of
VIEW
Mix-and-Match Riboswitches Colby D Stoddard and Robert T. Batey*
Department of Chemistry and Biochemistry, University of Colorado, Boulder, Campus Box 215, Boulder, Colorado 80309-0215
R
egulation of gene expression is a tightly controlled process that allows organisms to adapt to a continuously fluctuating physical and chemical environment. Activation or repression of genes promoting specific responses to changing conditions must occur in a precise and efficient manner. Until recently, protein factors have dominated the known gene expression regulatory strategies; however, RNA-based control (also called riboregulation) is now shown to play a significant role in all domains of life (1–3). One form of riboregulation, called riboswitches, is broadly distributed within the 5=-untranslated region of bacterial messenger RNAs (mRNAs) and controls transcription or translation via direct binding of small-molecule metabolites or metal ions (reviewed in ref 3). This control element consists of two functional components, a highly structured ligandbinding aptamer domain that directly binds a specific ligand followed by an expression platform that directly controls the expression of downstream genes (Figure 1, panel a). In transcriptional control, transduction of intracellular ligand concentration into changes in gene expression is achieved through the interplay of mutually exclusive RNA structures (Figure 1, panel a, orange boxes) that direct RNA polymerase to continue or abort mRNA synthesis. Structural analysis of the purine (4, 5), thiamine pyrophosphate (TPP) (6, 7), and S-adenosylmethionine (SAM) riboswitch (8) aptamer domains revealed a common mechanism by which this occurs. Communication between
www.acschemicalbiology.org
the aptamer domain and the expression platform is mediated by a sequence at the 3=-end of a P1 helix (Figure 1, panel a, red) that joins the 5=- and 3=-ends of the aptamer domain, a feature common to most riboswitches. Ligand binding establishes an intricate network of hydrogen bonds that stabilize formation of the P1 helix at the expense of one of two mutually exclusive secondary structures in the expression platform. In the structure of the SAM riboswitch (8), this is achieved by encapsulation of the ligand to form a series of interactions between the 3=-end of the P1 helix and other elements of the RNA (Figure 1, panel b). Direct sensing of metabolites and the absence of protein involvement allow riboswitch-mediated regulation of gene expression to be an economical and fast-reacting means of regulation (9). In their simplest form, riboswitches appear to function through feedback inhibition with control effected by a single ligand. However, recent studies have uncovered different arrangements of the basic riboswitch, presumably to create more sophisticated responses to ligands. The first example of this was found in an mRNA containing two complete TPP riboswitches positioned tandemly upstream of the thiSGHFE operon in several Desulfovibrio species (10) (Figure 2, panel a). The result of this arrangement is likely that gene expression is influenced by each riboswitch in an independent but additive fashion. Cooperative ligand binding, on the other hand, is observed in the glycine riboswitch (11). This regulatory element consists of two aptamer domains influencing a single
A B S T R A C T Riboswitches are noncoding RNA elements found in the 5=-untranslated region of messenger RNA (mRNA) that mediate gene expression in a cis fashion in the absence of protein. This common regulatory strategy in bacteria is achieved through the interplay of two distinct domains: an aptamer domain responsible for sensing intracellular concentrations of a specific metabolite and a domain containing a secondary structural switch directly controlling expression. In a recent study, riboswitches have been discovered that are capable of regulating transcription by using an RNA architecture mimicking a Boolean NOR logic gate. Tandem arrangement of elements that recognize S-adenosylmethionine and coenzyme B12 yields an mRNA that is only expressed when both metabolites are in low concentration in the cell.
*Corresponding author,
[email protected].
Published online December 15, 2006 10.1021/cb600458w CCC: $33.50 © 2006 by American Chemical Society
VOL.1 NO.12 • ACS CHEMICAL BIOLOGY
751
a
Aptamer domain Ligand
b
Expression platform
Term
P1 Gene
5′
SAM
P3 helix
P1 helix
5′
Anti-term 5′
3′
Gene
To expression platform
Figure 1. The basic form of the riboswitch. a) The basic riboswitch design consisting of an aptamer domain that binds ligand and an expression platform that directs gene expression. The “switching sequence” (shown in red) primarily directs secondary structural switching between the terminator (Term) and the anti-terminator (Anti-term) elements. b) The tertiary structure of the SAM riboswitch (PDB code 2GIS) emphasizing the relationship between ligand binding and stabilization of the switching sequence in the P1 helix.
riboswitches were identified (12). Using bioinformatics methods, the authors discovered that the 5=-untranslated region of the metE gene in six different isolates of Bacillus clausii was controlled by both the SAM (13, 14) and adenosylcobalamin (AdoCbl) riboswitches (15). In this gram-positive bacterium, conversion of homocysteine to methionine by the MetE protein requires the methyl-group donor methyltetrahydrofolate, whereas MetH utilizes methylcobalamin (MeCbl) to perform the same function in a more energy-efficient manner. It was proposed that because the MetE and MetH isozymes have redundant function it may be advantageous to repress the less-efficient MetE enzyme when sufficient concentrations of MetH coenzyme are present
expression platform downstream of the second aptamer (Figure 2, panel b). Glycine binding is a cooperative event in which binding of glycine to the first domain promotes binding to the second, serving to create a more digital regulation in which the on/off switch occurs over a narrower change in ligand concentration. From a structural perspective, it remains unclear how these domains communicate, whether by a limited secondary structure rearrangement or through interacting tertiary structures similar to protein subunit–subunit interactions. A further advance in our understanding of the complexity of riboswitches emerges from a recent study from the Breaker laboratory in which two heterogeneous tandem
a
TPP-1
Ligand
b
TPP-2
Term
P1 5′
Ligand
(16–18). MeCbl is derived from AdoCbl, so it is clear why AdoCbl can exert control over the metE locus. Control of metE by SAM is a form of feedback inhibition because methionine produced by MetE and MetH can be subsequently converted to SAM. This is similar to the switching between MetE and MetH in other bacteria, which is controlled by the MetR activator that indirectly senses AdoCbl and the MetJ repressor that directly senses SAM (19). Thus, the same strategy is achieved with either RNA- or protein-based regulatory elements. Independent yet coupled regulation of metE by SAM and AdoCbl riboswitches is achieved through a linear arrangement of two complete riboswitches (Figure 3). In vitro in-line probing experiments demonstrate that each ligand influences only the structure associated with its respective riboswitch and that they do not influence each other’s affinity for the mRNA (12). Further experiments using in vivo approaches strongly support this mechanism. The tandem riboswitch architecture allows metE expression to be controlled as a natural Boolean NOR logic gate in which distinct ligand inputs can independently control gene repression (Figure 3), as long as the ligand is present in saturating amounts. In general, an output that is negatively controlled (i.e., turned off) by the presence of one OR another of two signals is a Boolean NOR logic gate, also termed a joint denial gate. Thus, the SAM–AdoCbl tandem riboswitch generates an output of turning off gene
Cooperative ligand binding
Ligand
Term
P1 Gene
Ligand P1
5′
Term disrupted Gene
Figure 2. Arrangement of tandem riboswitches. a) Two independent TPP riboswitches are tandemly arranged to yield additive control. b) Cooperative ligand binding as observed in the glycine riboswitch mediated by two nearly identical aptamer domains that communicate with each other as well as the expression platform. This is a rare example of ligand binding activating gene expression.
752
VOL.1 NO.12 • 751–754 • 2006
STODDARD AND BATEY
www.acschemicalbiology.org
Point of
VIEW +SAM Ligand 5′
Term
P1 Gene +AdoCbl Ligand
−SAM
Term
P1 Gene Anti-term
−AdoCbl
5′ Anti-term Gene
Figure 3. The riboswitch NOR gate. The presence of SAM and/or AdoCbl always results in the off state, whereas only the absence of both results in gene expression being turned on.
expression if either SAM or AdoCbl binds its respective aptamer domain. This condition of independent control by each riboswitch is achieved only at either extreme, high or low, in ligand concentrations. A more physiologically relevant scenario is that each ligand fluctuates around an intermediate concentration range, a reflection of the observation of synergistic control in vivo (12). Natural heterogeneous tandem architecture is not surprising in light of in vitro generated RNAs that are capable of multi-input ribozyme activation (20). Tandem arrangement of the in vitro selected theophylline and flavin mononucleotide (FMN) aptamers yielded an RNA in which the binding of theophylline significantly promotes FMN binding. Cooperative ligand binding is subsequently detected via activation of an attached self-cleaving hammerhead ribozyme. This linear arrangement of aptamers was the first direct demonstration of an RNAbased Boolean logic gate, an AND gate in this case, in which two distinct ligands are required for an output (ribozyme cleavage). These experiments, along with natural tandem riboswitches, suggest that a large repertoire may exist of complex RNA-based www.acschemicalbiology.org
aptamer, which lies between the guanine aptamer and the expression platform. This element is speculated to override the decision-making process of the guanine riboswitch, creating a Boolean “implication gate”. This override gate would activate gene expression if the guanine is high, normally an off situation (Figure 4, middle), and the unknown ykkC aptamer ligand is high (Figure 4, bottom). Similar to observations in the glycine riboswitch, the lack of an expression platform separating the guanine and ykkC aptamer also may be evidence that the aptamers display some form of cross talk concerning the state of ligand binding. Little direct evidence exists for this hypothesis, so it is still possible that this riboswitch regulates gene expression via another mechanism. The implication of these latest discoveries is that like proteins, RNA can exploit both the 1D and 3D properties of the macromolecule to create new functions out of existing modules. It has long been established that a standard means of readily generating proteins with new biological functions is to mix
control elements that are yet to be discovered. As more bacterial genomes are sequenced, we are certain to observe even stranger arrangements of riboswitch elements yielding increasingly sophisticated regulatory responses. In the same paper by Breaker and co-workers, a novel riboswitch is presented that is likely responsive to guanine and + − Pairing another unknown effector mol5′ Gene ecule in a yet-to-be-defined fashion (12). This mRNA ykkC Guanine element, found in Thermoanaerobacter tengcongensis, controls Ligand the ykkCD operon that encodes Term 5′ a multi-drug-resistance efflux Gene pump of broad specificity (21). In this RNA, the guanine aptamer domain is predicted to ? Ligand Ligand either bind guanine/hypoxan“Override” 5′ thine to turn off expression or, in Gene the absence of ligand, pair with sequences in the terminator Figure 4. The riboswitch implication gate. In the absence stem to allow transcription (Figure 4, top and middle). Com- of guanine, regardless of the ykkC ligand, the gene is turned on, akin to a standard riboswitch (Figure 1). In the plicating this, however, is the presence of guanine, gene expression is suppressed, also presence of a second conserved as is standard, except if the ykkC ligand is present, which stem-loop structure, the ykkC overrides the off state.
VOL.1 NO.12 • 751–754 • 2006
753
and match small, modular domains in novel combinations, akin to “beads on a string”. Similarly, riboswitches can arrange different patterns of either aptamer domains or a completely functional aptamer/expression module to create new regulatory responses that could not be achieved by the simple module alone. Acknowledgment: This work was supported by a Research Scholar Grant from the American Cancer Society to R.T.B.
REFERENCES 1. Gottesman, S. (2005) Micros for microbes: noncoding regulatory RNAs in bacteria, Trends Genet. 21, 399–404. 2. Mattick, J. S., and Makunin, I. V. (2006) Non-coding RNA, Hum. Mol. Genet. 15, R17–R29. 3. Winkler, W. C., and Breaker, R. R. (2005) Regulation of bacterial gene expression by riboswitches, Annu. Rev. Microbiol. 59, 487–517. 4. Batey, R. T., Gilbert, S. D., and Montange, R. K. (2004) Structure of a natural guanine-responsive riboswitch complexed with the metabolite hypoxanthine, Nature 432, 411–415. 5. Serganov, A., Yuan, Y. R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A. T., Hobartner, C., Micura, R., Breaker, R. R., and Patel, D. J. (2004) Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs, Chem. Biol. 11, 1729–1741. 6. Serganov, A., Polonskaia, A., Phan, A. T., Breaker, R. R., and Patel, D. J. (2006) Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch, Nature 441, 1167–1171. 7. Thore, S., Leibundgut, M., and Ban, N. (2006) Structure of the eukaryotic thiamine pyrophosphate riboswitch with its regulatory ligand, Science 312, 1208–1211. 8. Montange, R. K., and Batey, R. T. (2006) Structure of the S-adenosylmethionine riboswitch regulatory mRNA element, Nature 441, 1172–1175. 9. Nudler, E., and Mironov, A. S. (2004) The riboswitch control of bacterial metabolism, Trends Biochem. Sci. 29, 11–17. 10. Rodionov, D. A., Dubchak, I., Arkin, A., Alm, E., and Gelfand, M. S. (2004) Reconstruction of regulatory and metabolic pathways in metal-reducing deltaproteobacteria, Genome Biol. 5, R90. 11. Mandal, M., Lee, M., Barrick, J. E., Weinberg, Z., Emilsson, G. M., Ruzzo, W. L., and Breaker, R. R. (2004) A glycine-dependent riboswitch that uses cooperative binding to control gene expression, Science 306, 275–279. 12. Sudarsan, N., Hammond, M. C., Block, K. F., Welz, R., Barrick, J. E., Roth, A., and Breaker, R. R. (2006) Tandem riboswitch architectures exhibit complex gene control functions, Science 314, 300–304. 13. Grundy, F. J., and Henkin, T. M. (1998) The S box regulon: a new global transcription termination control system for methionine and cysteine biosynthesis genes in Gram-positive bacteria, Mol. Microbiol. 30, 737–749.
754
VOL.1 NO.12 • 751–754 • 2006
14. Winkler, W. C., Nahvi, A., Sudarsan, N., Barrick, J. E., and Breaker, R. R. (2003) An mRNA structure that controls gene expression by binding S-adenosylmethionine, Nat. Struct. Biol. 10, 701–707. 15. Nahvi, A., Sudarsan, N., Ebert, M. S., Zou, X., Brown, K. L., and Breaker, R. R. (2002) Genetic control by a metabolite binding mRNA, Chem. Biol. 9, 1043. 16. Banerjee, R. V., Frasca, V., Ballou, D. P., and Matthews, R. G. (1990) Participation of cob(I) alamin in the reaction catalyzed by methionine synthase from Escherichia coli: a steady-state and rapid reaction kinetic analysis, Biochemistry 29, 11101– 11109. 17. Gonzalez, J. C., Peariso, K., Penner-Hahn, J. E., and Matthews, R. G. (1996) Cobalamin-independent methionine synthase from Escherichia coli: a zinc metalloenzyme, Biochemistry 35, 12228–12234. 18. Pejchal, R., and Ludwig, M. L. (2005) Cobalaminindependent methionine synthase (MetE): a face-toface double barrel that evolved by gene duplication, PLoS Biol. 3, e31. 19. Wu, W. F., Urbanowski, M. L., and Stauffer, G. V. (1992) Role of the MetR regulatory system in vitamin B12-mediated repression of the Salmonella typhimurium metE gene, J. Bacteriol. 174, 4833– 4837. 20. Jose, A. M., Soukup, G. A., and Breaker, R. R. (2001) Cooperative binding of effectors by an allosteric ribozyme, Nucleic Acids Res. 29, 1631–1637. 21. Jack, D. L., Storms, M. L., Tchieu, J. H., Paulsen, I. T., and Saier, M. H., Jr. (2000) A broad-specificity multidrug efflux pump requiring a pair of homologous SMR-type proteins, J. Bacteriol. 182, 2311–2313.
STODDARD AND BATEY
www.acschemicalbiology.org
