Mimicking nature: Ron Breaker (left) and his team are developing synthetic RNA switches.

One day in 1999, molecular biologist Ron Breaker interrupted his lab's weekly meeting to issue a challenge: to discover whether their synthetic RNA biosensors were original, or whether nature got there first.

Breaker's team, at Yale University in New Haven, Connecticut, was crafting molecular sensors out of RNA that changes shape when it binds to a specific molecule. The team's aim was to produce miniature devices that could, for example, monitor the biochemical secretions of living cells. These sensors were working so well that Breaker couldn't believe that evolution had overlooked the same mechanism.

Breaker was remarkably prescient. Similar sensors, it turned out, had been staring biologists in the face for nearly 30 years. In an eye-opening series of experiments, Breaker's team has revealed that these RNA sensors, dubbed 'riboswitches', represent a previously overlooked mechanism of gene regulation. “We can explain a lot of the unexplained findings that have appeared in the literature over the years,” Breaker says.

Until Breaker's team revealed its results, biologists had assumed that the regulation of gene activity in response to environmental cues was mediated only by proteins. In the classic model of gene regulation, cells monitor their environment through a variety of specialized sensor proteins deployed either on their surfaces or internally. For example, a protein in the bacterium Escherichia coli detects the presence of the amino acid tryptophan, allowing E. coli to shut down its own tryptophan production when a ready supply of this essential nutrient exists, thus avoiding waste.

Detective work

Sensor proteins typically trigger signalling pathways, a series of protein interactions that results in a protein binding to regulatory sequences of DNA that control genes — as the tryptophan detector protein of E. coli does. Binding of the protein to DNA influences the activity of another group of proteins that controls the expression of genes to produce messenger RNA (mRNA). In the case of tryptophan production in E. coli, these mRNAs code for the enzymes that manufacture the amino acid.

Riboswitches are mRNAs that sense the environment directly, shutting themselves down in response to particular chemical cues. For example, bacterial genes for enzymes that direct the synthesis of vitamin B12 use a riboswitch1. The mRNAs transcribed from these genes fold into a specialized shape that creates a binding pocket for coenzyme B12 — the activated form of the vitamin. When coenzyme B12 lands in this pocket, the mRNA shifts its shape in such a way as to mask a nearby sequence that would otherwise tell the cell's protein-manufacturing factories, known as ribosomes, to 'start reading here'. When the coenzyme B12 is abundant, these sequences are hidden, and enzymes for B12 synthesis are no longer produced.

Turned off: in the absence of its target molecule (left) this riboswitch gives gene expression the green light to go ahead, but when the molecule binds, expression is halted. Credit: B. BOESE

Seven types of riboswitch have been found in bacteria so far. These include switches controlling the manufacture of the vitamins B1 and B2 (refs 2–4) and the nucleotide guanine5. By searching databases of nucleotide sequence, Breaker has found candidate riboswitches in plants and fungi6. “Higher organisms might be riddled with riboswitches,” he says.

Some of the biosynthetic pathways in which riboswitches are now implicated have been studied intensely for nearly three decades. So why were these control elements missed? One reason is that techniques for working with RNA were not good enough until recently. But another has to do with a long-standing prejudice against mRNA: for years it was considered as nothing more than a carrier of genetic instructions.

Role reversal

In the 1980s, however, researchers learned that certain RNA molecules, dubbed ribozymes, could catalyse biochemical reactions, a job previously thought to be the exclusive province of enzymes, which are proteins7,8. Later, two research groups independently found that RNA could behave like another type of protein: antibodies. The teams synthesized RNA molecules — now known as aptamers — that, like antibodies, can latch on tightly to specific target molecules9,10.

Researchers then began to speculate about the existence of aptamers in living organisms. “We used to say it would be surprising if nature didn't take advantage of this,” says Gary Stormo, who studies RNA at Washington University in St Louis, Missouri. Several years later, this is exactly what Breaker found: a riboswitch is a specialized type of aptamer in which the act of binding to its target alters a gene's activity.

Unbelievable

Many biologists initially resisted the idea of natural aptamers, because it conflicted with traditional notions of what RNA can do. “I still give talks where scientists walk out of the room shaking their heads,” says Larry Gold, who led one of the teams that first made synthetic aptamers while he was at the University of Colorado at Boulder. “The entire idea of aptamers seems completely unbelievable to them.”

Nevertheless, aptamers have useful applications. Gold is now chief science officer and chairman at SomaLogic, a company in Boulder that is developing aptamer-based proteomic 'chips' to study the profile of proteins produced in particular cells or tissues.

Other companies are exploring the use of aptamers as drugs. Breaker stumbled upon natural riboswitches while working on a biotech application for synthetic aptamers. And in May 2001, he and Andrew Ellington of the University of Texas at Austin helped to found Archemix, a company based in Cambridge, Massachusetts, that is developing ways to use riboswitches in research and medicine. One application couples an aptamer to a ribozyme so that the ribozyme is only active when the aptamer binds to its target. Ellington is also interested in using riboswitches in gene therapy, allowing patients to take pills to switch introduced genes on or off.

More fundamentally, some researchers believe that natural riboswitches represent an evolutionary throwback to a time near the dawn of life, before DNA and proteins, when RNA reigned supreme. In this world, RNA would have handled both the storage of genetic information — DNA's job in most organisms today — and its expression, which is now the role of proteins. RNA-only genomes still exist in some viruses. And ribozymes are now known to catalyse a slew of important biochemical reactions. What's been missing from the 'RNA-world' hypothesis is a mechanism whereby organisms react in precise ways to their environment — which is where riboswitches come in.

What might riboswitches have controlled in the ancient RNA world? After all, there was no protein synthesis for them to block. Breaker suggests that riboswitches might have controlled ribozymes directly — as do the RNA sensors that he has designed. A nucleotide might have shut down the ribozymes that catalysed nucleotide synthesis, for example.

Back to the future

Other researchers are not convinced that Breaker's riboswitches are the direct descendants of sensors from the RNA world. Ellington agrees that riboswitches played a role in early evolution, but he suspects that since then they have emerged several times over the history of life on Earth. If modern riboswitches were direct descendants from the RNA world, he argues, they would be much more widespread.

Perhaps riboswitches are more widespread than we realize. Although sequence comparisons have so far turned up candidates in only a few species of plants and fungi7, the number discovered could be limited by our knowledge of what ribo-switches look like. If they are very common, their sequences could be as varied as the molecules they bind to, says Evgeny Nudler, who heads a team working on riboswitches at New York University in Manhattan.

The big question now is whether riboswitches represent an occasional oddity, or a widespread and biologically important mechanism of gene control. If they start turning up in less ancient biochemical pathways throughout the plant and animal kingdoms, biologists will certainly sit up and take notice. Gerald Joyce, a ribozyme researcher at the Scripps Research Institute in La Jolla, California, agrees, “If we found them in a real blood-and-guts metabolic pathway, that would be really incredible.”