Mediator, a hulking transcription machine comprising 25 to 30 different proteins and tipping the scales at more than 1 MDa, is hard to miss. Still, its structural details have not been well understood, much to the frustration of biochemists, who are eager to figure out how Mediator works—as well as how it fails to work, since Mediator’s failures likely give rise to myriad calamities, from cancer to inherited disorders.
To understand these intricacies, scientists have needed an accurate 3D model of its architecture, including the locations of all its subunit proteins and a description of the different conformations Mediator can adopt to influence interactions between other components of the transcription machinery. Mediator, however, is so large, so complex, and so flexible that it is makes it a poor candidate for high-resolution imaging methods such as X-ray crystallography or nuclear magnetic resonance spectroscopy.
Nonetheless, researchers at The Scripps Research Institute report that they have overcome biochemical and image analysis hurdles to obtain accurate electron microscopy structures of yeast and human mediators. The researchers first presented these structures June 5 in Cell, in an article entitled “Subunit Architecture and Functional Modular Rearrangements of the Transcriptional Mediator Complex.”
“Subunit localization experiments, docking of partial X-ray structures, and biochemical analyses resulted in comprehensive mapping of yeast Mediator subunits and a complete reinterpretation of our previous Mediator organization model,” wrote the authors. “Large-scale Mediator rearrangements depend on changes at the interfaces between previously described Mediator modules, which appear to be facilitated by factors conducive to transcription initiation.”
To determine the full structure clearly, the researchers, led by Francisco J. Asturias, Ph.D., began by producing highly pure quantities of a standard yeast version of Mediator—the purification process itself being a major challenge. They then used this collection of Mediator particles to record roughly 85,000 EM images, which they categorized according to conformation. Averaging these yielded the clearest 3D model yet of the Mediator structure, to a resolution of about 18 Angstroms (1.8 billionths of a meter).
Using various other biochemical analyses, including the subtraction of different protein subunits to see how the EM images changed, the scientists were able to identify the precise locations of yeast Mediator’s 25 protein subunits.
This mapping resulted in a comprehensive revision of the old rough model of Mediator’s head-middle-tail structure. “After we located all the protein subunits, we realized that the head module is at the top of Mediator, not the bottom as had been thought,” said Kuang-Lei Tsai, Ph.D., a co-author who works in the Asturias Laboratory. “These new data have helped us make sense of many previous biochemical observations.”
The researchers found that yeast Mediator and human Mediator share the same broad architecture, implying that this structure has been, for the most part, conserved throughout the billion years of evolution that separate yeast and humans. “Basically the two Mediators have similar overall structure,” said Dr. Tsai.
In the last part of the study, Drs. Asturias and Tsai used the new structural data to show how Mediator likely changes its conformation as it interacts with RNA polymerase on the one hand, and various transcription regulators on the other.
“This study has given us a fairly definitive picture of the Mediator architecture and how the different subunits are organized, so we can start to work toward an atomic resolution model,” Dr. Asturias said. “We also want to understand better how Mediator interacts with all those other proteins to actually carry out transcription in a regulated manner.”