Life’s instructions are written in DNA, but it is the enzyme RNA polymerase II (Pol II) that reads the script, transcribing RNA in eukaryotic cells and eventually giving rise to proteins. Scientists know that Pol II must advance down the gene in perfect sync with other biological processes; aberrations in the movement of this enzyme have been linked to cancer and aging. But technical hurdles prevented them from precisely determining how this important molecular machine moves along DNA, and what governs its pauses and accelerations.
A new study fills in many of those knowledge gaps. In a paper published in Nature Structural & Molecular Biology, researchers used a single-molecule platform to watch individual mammalian transcription complexes in action. The result is a clear view of how this molecular engine accelerates, pauses, and shifts gears as it transcribes genetic information.
"What's really striking is how this machine functions almost like a finely tuned automobile," says Shixin Liu, head of the Laboratory of Nanoscale Biophysics and Biochemistry. "It has the equivalent of multiple gears, or speed modes, each controlled by the binding of different regulatory proteins. We figured out, for the first time, how each gear is controlled."
"We're finally seeing where Pol II is, in both time and space, during the process," adds Joel E. Cohen, head of the Laboratory of Populations. "Our platform allowed us to objectively assess when this machine shifts gears, and how fast it goes."
New tool provides new answers
First discovered more than fifty years ago by Rockefeller's Robert Roeder, Pol II moves step-by-step along the DNA molecule, constructing a matching RNA strand that will ultimately give rise to proteins. But Pol II does not travel along DNA at a steady clip, especially in higher organisms like humans.
After initiation, it slows and often pauses near the start of a gene before regulatory proteins such as P-TEFb and PAF1C propel it into rapid transcription mode. As it nears the end of a gene, the enzyme decelerates again to finish cleanly. This pacing is crucial: too fast or too slow, and RNA molecules cannot be properly processed or coordinated with other vital cellular events. Missteps in Pol II's speed control have been linked to aging and a myriad of diseases including cancer.
"A lot of people aren’t aware of these links. In fact, I'm often asked: as long as it can make RNA and we know how it does it, do we really care about the speed of the machine, or whether it pauses?" Liu says. "We care precisely because we know that the kinetics of transcription are important for proper gene expression and are linked to various diseases."
Technical limitations explain much of why prior studies struggled to unambiguously illuminate how Pol II regulates its motion. Techniques that measured averages across many molecules blurred the contributions of individual proteins, while single-molecule studies in simpler organisms such as yeast did not fully represent the intricate regulatory mechanisms of mammalian cells. Liu, an expert in single-molecule methods, realized that he could overcome these barriers only by rebuilding a mammalian transcription system in vitro, piece by piece from purified proteins, and pairing it with advanced imaging techniques and computational algorithms.
The idea gained momentum through a chance encounter with Cohen at Rockefeller’s Bass Dining Commons. "We sat together, and Joel asked me if I was working on anything interesting," Liu recalls. "I sort of pitched him this project, we started looking into it together, and we've been collaborating since then. I'm so glad we sat at the same table that day."
"Our work is collaborative across different disciplines, our two labs, and also across national boundaries," Cohen adds, referring to coauthors from China in the group of Yanhui Xu at Fudan University who made critical contributions to the research. "If we want science to prosper, we need to continue to enable collaboration across boundaries."
A molecular gearbox
Working together, the researchers developed a platform that merges biochemistry, single-molecule imaging, and computation to reveal Pol II at work in unprecedented detail.
By rebuilding the transcription machinery piece by piece from purified mammalian proteins and tracking its motion in real time—and using a computational framework and structural modeling to pinpoint exactly when the enzyme shifts gears—the team showed that several key regulatory proteins govern how Pol II moves. P-TEFb, a sort of master switch, phosphorylates both Pol II and a protein complex, DSIF, to unlock its full activity. (DSIF turned out to be more complicated than expected—depending on its state, it could either push Pol II forward or hold it back.) Next, the protein PAF1C emerged as Pol II's main accelerator, snapping transcription into motion the moment it bound DNA. The SPT6 protein played the role of stabilizer, making sure PAF1C stayed securely attached so that the machine could keep running smoothly.
Once PAF1C was in place, it enabled yet another factor, RTF1, to bind. RTF1 provided an additional boost in transcription speed and switched Pol II to high gear—a step that strictly required PAF1C but not DSIF. This suggested a functional nexus among PAF1C, RTF1, and DSIF in mammalian cells that is not yet seen in yeast, underscoring the evolutionary sophistication of the system.
"This is the first time we've been able to see mammalian Pol II move at a physiological speed in real time and, because we labeled the various elongation factors, we were also able to measure their binding kinetics during active transcription," says Yukun Wang, a postdoctoral fellow in the Liu lab. "Sometimes, to see is to understand. The power of real-time visualization is that you can really see the molecule—you can watch and learn."
The findings also shed light on how disruptions in this control may contribute to cancer and aging by providing new insight into factors such as P-TEFb, which is considered a promising drug target for leukemia and solid tumors.
"P-TEFb has proven notoriously difficult to inhibit without toxic side effects," Liu says. "Our work may offer clues for designing more specific therapeutics. At the same time, insight into how Pol II is regulated through each of the stages in healthy cells should give us an even broader understanding of what could go wrong in disease."
The study’s greatest impact, however, may be the platform itself. By proving that single-molecule visualization is possible in a fully reconstituted mammalian system, the researchers have created a tool that can now be used to tackle longstanding questions in biology. Work is already underway to improve the platform by adding in nucleosomes—the basic units of DNA packaging in eukaryotic chromosomes—to better understand how Pol II moves through templates more akin to its natural environment.
And the potential applications for the platform's computational component may also be far-reaching. "Anything that involves navigation in space and changes in speed could potentially use this software.” Cohen says.
Journal
Nature Structural & Molecular Biology