©2004 Nature Publishing Group
The GlmS enzyme, which is involved in GlcN6P synthesis, is translated from mRNA containing a ribozyme sequence. GlcN6P activates the ribozyme cleaving the mRNA sequence and turning enzyme production off. (Redrawn from T.R. Cech,
Coming from various directions, a number of research groups stumbled upon a naturally occurring mechanism for gene control that depends solely on RNA and environmental cues. Though the working details for this largely prokaryotic device are just now being revealed, so called riboswitches have thus far intrigued and impressed. "It's a completely new, unsuspected way of controlling gene expression," says David Lilley, professor of molecular biology, University of Dundee, Scotland
With implications and applications for basic research and therapeutics, riboswitches may represent another chisel chipping away at the now crumbling notion of RNA as an intermediary between gene and protein. While some disagree as to whether...
TPP OF THE ICEBERG
The standard model of gene repression in bacteria involves proteins that sense the presence of a small molecule such as a hormone or amino acid, and then bind DNA or RNA. A riboswitch, on the other hand, bypasses proteins altogether; RNAs activate or inactivate themselves by changing shape when bound by specific molecules.
Researchers are beginning to elucidate the details of how some riboswitches work. In one example, mRNAs that encode thiamine-synthesis enzymes in
Evidence for the existence of natural riboswitches has been generated both in silico and in vivo. In 2002, Mikhail Gelfand and colleagues at the Institute for Information Transmission Problems, Moscow, used computational methods to predict the molecular mechanism of riboswitch-mediated regulation of riboflavin biosynthesis in various bacteria.2 Also that year, New York University scientist Evgeny Nudler and coworkers published in vitro evidence of a riboflavin riboswitch in
Ron Breaker of Yale University, a prominent force in riboswitch awareness, has been generating synthetic RNA-based molecular sensors in his laboratory for several years. Breaker hypothesized that a system that works so well in the test tube might be exploited naturally in the cell. He assumed that scientists had already uncovered instances of riboswitches and that potential candidates could be found in published experiments.
So his research group searched the literature for situations in which investigators looked for, but could not find, a protein sensor controlling specific gene regulation. "We simply looked for those stories in the literature, and we found six major examples of genetic mysteries that ultimately were solved by proving that a riboswitch was controlling the gene," Breaker explains. These included mechanisms controlling B-vitamin synthesis, riboflavin processing, and purine metabolism. He even found a riboswitch sequence for thiamine biosynthesis in
In a recent example, Breaker and colleagues discovered a bacterial ribozyme switch, a molecule that catalyzes its own cleavage in response to a binding metabolite.4 "Not only is this a new class of ribozyme but it also is the first example of a ribozyme that directly senses a metabolite for the purpose of genetic control," says Breaker.
Naturally occurring riboswitches have been found in archaea and some eukaryotes, but by far the majority of riboswitches identified to date are bacterial. Few eukaryotic riboswitches have been found, in part because researchers simply have not been looking for them, but the search is also difficult. Eukaryotic genomes are larger and less available than those of bacteria, so investigators have less to work with when scanning genomes for likely candidates.
Nudler speculates that most eukaryotic riboswitches will have specialized functions and that their sequences are not as evolutionarily conserved as those of bacteria. "We don't think that the majority of riboswitches that exist in higher eukaryotes are similar to ancient riboswitches [that] we find in bacteria," says Nudler. Uncovering them will require that investigators look for alternatives to sequence comparison, he adds.
OLD WORLD RELICS
Some say that riboswitch sequences could unravel secrets to life's origins; some view riboswitches as relics of an RNA world. "The modern switches that we see in cells today might be directly descended from RNAs that were present some three or four billion years ago in the RNA world," says Breaker, who cites three pieces of evidence to support this argument. First, the riboswitches that have been identified are widespread in the organisms in which they are found, suggesting they were present in ancient precursors of these organisms. Riboswitches also represent a simple form of molecular sensing, one that does not require DNA or proteins. "So you wouldn't have to invoke a more complex metabolic state in order for these things to function," says Breaker. Finally, contemporary riboswitches sense metabolites that are essential to modern and likely ancient cells.
©2002 Nature Publishing Group
In the proposed mechanism for thymine pyrophosphate (TPP)-dependent deactivation of thiM translation, P8* pairing occurs between the anti-Shine-Delgarno (SD) element and the anti-anti-SD element when TPP is absent. This allows the SD sequence to interact with the ribosome. TPP causes obligate formation of the P1 stem which sequesters part of the anti-anti-SD element, and the full P8 stem forms limiting ribosome access.
Gerald Joyce of The Scripps Research Institute, La Jolla, Calif., says that while the principle of RNA adopting a regulatory role is relevant to the RNA-world hypothesis, he doubts that any extant ribo-switches are fossils. Rather, they may be clever solutions to contemporary biological problems. Certainly the metabolic pathways involving the synthesis and processing of modern riboswitch-binding metabolites did not exist, he says, because these pathways involve proteins. "The slate may have been cleaned once protein-based metabolism started operating," says Joyce. Existing riboswitches, he adds, "are part and parcel of a biology of a protein-based metabolism."
Still, studying riboswitches across a range of organisms can lead to insights about the evolution of regulatory mechanisms. In inpress work, Gelfand and colleagues have examined the evolution of methionine-biosynthesis regulation in Gram-positive bacteria. Here a riboswitch shared by common ancestors was replaced by other regulatory mechanisms. "In certain bacteria, this system ... was eliminated and another regulatory system took the role," Gelfand speculates.
In recent years, multiple labs have uncovered surprising biological roles for RNA, and molecules implicated in RNA interference have been tapped for their potential as therapeutic agents. Riboswitches may be next. "If you've got elements in bacterial RNA that are selectively binding small molecules, it doesn't take that huge a jump to imagine that chemists could get working to design drugs that look somewhat like the effectors but bind irreversibly," thereby creating an effective antibiotic, says Lilley. The Breaker lab has uncovered such molecules: In
Alternatively, riboswitches themselves could be designed to target specific molecules for use as drugs, research tools, or even interfaces to nanoscale biological sensors. Unlike proteins, which have complex tertiary structures that are difficult to tease out from amino acid sequence alone, RNAs have little tertiary structure and are relatively easy to engineer, says researcher Andrew Ellington at the University of Texas, Austin. "Basically, if you know the two-dimensional structure of RNA ... you have a really good idea of its overall structure," says Ellington. Indeed, a group of scientists at Friedrich Alexander University in Erlangen, Germany, published one of the first examples of a rationally designed riboswitch earlier this year.6
INSIDE A SECRET SOCIETY
Two major questions still need to be addressed, according to Breaker. The first is whether riboswitches play a fundamental role in genetic regulatory processes or simply represent a biological oddity. Further, scientists need to understand precisely how the switches work. "I think what we have to do is to identify what appears to be many different mechanisms for how riboswitches function, and then begin to dismantle each switch and understand precisely how metabolite binding triggers a change in gene expression," Breaker says. Understanding these mechanisms could help molecular engineers design novel gene-control elements for gene-therapy vectors or diagnostic tools.
RNA experts emphasize that riboswitches are just part of the recent explosion of research implicating RNA in a host of regulatory mechanisms. "A few years ago, before RNAi, siRNAs, microRNAs, and riboswitches, RNA was [seen as] the passive messenger that carried the information dutifully from the DNA to the protein-synthesizing machinery," says Joyce. Now, he adds, "there's this whole secret society that's really running quite a bit of the show within the cells."
Further probes into structured RNAs may reveal that the road does not end at riboswitches. "Each time we learn something more, says Ellington, "it seems as though there's an even greater network of complexity that we didn't appreciate previously. I'm not sure where it will stop."
Aileen Constans can be contacted at