Three-D Images Shed Light On First Steps Of RNA Synthesis

These research findings of Rockefeller University scientists illuminate how the process of transcription begins in bacteria. Understanding this process may provide clues for developing new drugs to treat bacterial infections.

The DNA-transcribing machinery is called the RNA polymerase (RNAP). The RNAP comprises a core enzyme, which can synthesize RNA from a DNA template, but cannot initiate transcription by itself. Initiation requires another protein, the sigma subunit, which binds core to form the holoenzyme.

The sigma subunit "tells" RNAP where in the genomic DNA to start transcription. It does this by recognizing a section of the gene’s DNA called the promoter region and "melting" the DNA to expose one of the double-helical strands the RNAP will use as the template to synthesize RNA.

The Rockefeller researchers, led by Seth Darst, Ph.D., studied RNAP from a bacterium called Thermus aquaticus, a member of a family of organisms that thrive at high temperatures. They used a technique called X-ray crystallography to reveal the structure of the holoenzyme, the core enzyme bound to the sigma subunit. They also used this research tool to visualize the structure of the holoenzyme bound to a short piece of DNA containing a promoter sequence.

In X-ray crystallography, X-rays are fired at crystallized proteins or complexes of proteins bound to other molecules. Analysis of the data from the diffracted X-rays yields information about the three-dimensional structure of the molecule. Some molecules can be visualized at the resolution of a few angstroms (a typical atom is roughly two angstroms in diameter). But Darst notes that the crystals his group used did not diffract well enough to produce a high resolution structure. So, he and his colleagues used a novel hybrid approach.

"From previous and parallel work in our lab, we had high resolution structures of the RNA polymerase and parts of the sigma subunit," says Darst, who is the Jack Fishman Professor and head of laboratory at Rockefeller. "We fitted the high resolution structures into the lower resolution structure to obtain the structure of the holoenzyme. And since the structure of DNA is known, we used that to solve the structure of the holoenzyme bound to the promoter fragment."

The holoenzyme structure offers some clues to how transcription begins in bacteria. Research from Darst’s lab published recently in Molecular Cell revealed that the sigma subunit contains four protein "domains" connected to one another by short stretches of amino acids called "linkers." With information from the holoenzyme structure published in Science, the researchers show that the domains of the sigma subunit spread out and bind to one face of the core enzyme, which positions sigma to bind to DNA.

The most interesting finding, says Darst, involves the position of one of the linkers. This linker, containing about 33 amino acids, is much longer than the others. This linker goes inside the polymerase towards the active site and back out, producing a very unusual protein-protein interaction.

"These sigma domains sit on the surface of the RNA polymerase, but this very long linker is buried inside the polymerase, near the active site of the enzyme," he says. "In order to establish this complex, there need to be a lot of conformational changes to get the linker inside."

In bacteria, the first step in transcription is binding of RNA polymerase to a nucleotide, one of the four building blocks that make up DNA. The researchers hypothesize that the linker participates in binding the initiating nucleotide, and once the RNA polymerase begins synthesizing RNA, the sigma subunit is no longer needed.

Gene expression is regulated at every step, but probably the major focal point of regulation is at the initiation of transcription. According to Darst, the linker that is buried near the active site may play a role in telling the RNA polymerase when to start synthesizing RNA.

"The linker sticks inside the polymerase and actually blocks the path of the RNA," says Darst. "As the RNA transcript begins to be elongated, it has to push this linker out from a channel inside the polymerase. The linker’s not completely pushed out until the RNA becomes about 12 nucleotides long."

Scientists have known from biochemical experiments that a transition occurs when the RNA transcript reaches a length of 12 nucleotides: the complex becomes much more stable and the elongation phase begins.

"We think that signal to go into elongation occurs when the RNA becomes long enough to fill the channel," says Darst. "It pushes the linker peptide out, and then sigma begins to fall off."

This structure also explains a curious phenomenon called abortive initiation, in which transcription starts and stops. The polymerase remains at the promoter site synthesizing short pieces of RNA, which fall out, and then the process starts over again. This can happen hundreds of times before the process moves on and a long RNA transcript is synthesized.

From the structural information and other experiments, the researchers now know that abortive initiation happens because of a competition between RNA and the linker peptide. If the peptide wins, the RNA falls out and transcription starts over.

"Every once in a while the RNA manages to push the peptide all the way out, and once it’s long enough to push it all the way out, it’s done, and there’s no more abortive initiation," says Darst.

The structure of the holoenzyme bound to the promoter fragment, Darst says, "provides fuel for a model to understand the steps in bacterial transcription."

The holoenzyme specifically recognizes the sequence of the promoter and forms a structure called a closed complex — closed because the DNA is not melted. Once the sequence is recognized, the holoenzyme goes through a series of steps and ultimately forms the open complex, where it actually melts DNA.

"From this structure and others, we made a model of the closed complex and the open complex, kind of a beginning and an end," says Darst. "What we’re really interested in now is the in-between steps.

"We think the models are pretty good, but they don’t provide a lot of insight into how it gets from A to B."

A better understanding of transcription in bacteria could lead to new drug targets to treat bacterial infections, particularly those that are resistant to current antibiotics. Previous research from Darst’s lab showed that rifampicin, one of the two drugs most effective against tuberculosis, kills the microbe that causes TB by physically blocking that bacterium’s RNA polymerase.

The Rockefeller scientists who conducted this research are postdoctoral fellow Katsuhiko S.Murakami, research scientist Shoko Masuda, postdoctoral fellow Elizabeth A. Campbell and research scientist Oriana Muzzin.

This research was supported in part by the National Institute of General Medical Sciences, part of the federal government’s National Institutes of Health, the Norman and Rosita Winston Foundation and the Human Frontiers Sciences Program.